Analysis of Algorithms | Set 1 (Asymptotic Analysis)

Why performance analysis?

There are many important things that should be taken care of, like user friendliness, modularity, security, maintainability, etc. Why to worry about
performance?
The answer to this is simple, we can have all the above things only if we have performance. So performance is like currency through which we can
buy all the above things. Another reason for studying performance is speed is fun!
Given two algorithms for a task, how do we find out which one is better?
One naive way of doing this is implement both the algorithms and run the two programs on your computer for different inputs and see which one
takes less time. There are many problems with this approach for analysis of algorithms.
1) It might be possible that for some inputs, first algorithm performs better than the second. And for some inputs second performs better.
2) It might also be possible that for some inputs, first algorithm perform better on one machine and the second works better on other machine for
some other inputs.
Asymptotic Analysis is the big idea that handles above issues in analyzing algorithms. In Asymptotic Analysis, we evaluate the performance of an
algorithm in terms of input size (we dont measure the actual running time). We calculate, how does the time (or space) taken by an algorithm
increases with the input size.
For example, let us consider the search problem (searching a given item) in a sorted array. One way to search is Linear Search (order of growth is
linear) and other way is Binary Search (order of growth is logarithmic). To understand how Asymptotic Analysis solves the above mentioned
problems in analyzing algorithms, let us say we run the Linear Search on a fast computer and Binary Search on a slow computer. For small values
of input array size n, the fast computer may take less time. But, after certain value of input array size, the Binary Search will definitely start taking
less time compared to the Linear Search even though the Binary Search is being run on a slow machine. The reason is the order of growth of
Binary Search with respect to input size logarithmic while the order of growth of Linear Search is linear. So the machine dependent constants can
always be ignored after certain values of input size.
Does Asymptotic Analysis always work?
Asymptotic Analysis is not perfect, but thats the best way available for analyzing algorithms. For example, say there are two sorting algorithms that
take 1000nLogn and 2nLogn time respectively on a machine. Both of these algorithms are asymptotically same (order of growth is nLogn). So,
With Asymptotic Analysis, we cant judge which one is better as we ignore constants in Asymptotic Analysis.
Also, in Asymptotic analysis, we always talk about input sizes larger than a constant value. It might be possible that those large inputs are never
given to your software and an algorithm which is asymptotically slower, always performs better for your particular situation. So, you may end up
choosing an algorithm that is Asymptotically slower but faster for your software.
We will covering more on analysis of algorithms in some more posts on this topic.
References:
MITs Video lecture 1 on Introduction to Algorithms.

Analysis of Algorithms | Set 2 (Worst, Average and Best Cases)

In the previous post, we discussed how Asymptotic analysis overcomes the problems of naive way of analyzing algorithms. In this post, we will
take an example of Linear Search and analyze it using Asymptotic analysis.
We can have three cases to analyze an algorithm:
1) Worst Case
2) Average Case
3) Best Case
Let us consider the following implementation of Linear Search.
#include <stdio.h>
// Linearly search x in arr[]. If x is present then return the index,
// otherwise return -1
int search(int arr[], int n, int x)
{
int i;
for (i=0; i<n; i++)
{
if (arr[i] == x)
return i;
}
return -1;
}
/* Driver program to test above functions*/
int main()
{
int arr[] = {1, 10, 30, 15};
int x = 30;
int n = sizeof(arr)/sizeof(arr[0]);
printf("%d is present at index %d", x, search(arr, n, x));
getchar();
return 0;
}

Worst Case Analysis (Usually Done)

In the worst case analysis, we calculate upper bound on running time of an algorithm. We must know the case that causes maximum number of
operations to be executed. For Linear Search, the worst case happens when the element to be searched (x in the above code) is not present in the
array. When x is not present, the search() functions compares it with all the elements of arr[] one by one. Therefore, the worst case time
complexity of linear search would be ?(n).
Average Case Analysis (Sometimes done)
In average case analysis, we take all possible inputs and calculate computing time for all of the inputs. Sum all the calculated values and divide the
sum by total number of inputs. We must know (or predict) distribution of cases. For the linear search problem, let us assume that all cases are
uniformly distributed (including the case of x not being present in array). So we sum all the cases and divide the sum by (n+1). Following is the
value of average case time complexity.
Average Case Time =
=
= ?(n)

Best Case Analysis (Bogus)

In the best case analysis, we calculate lower bound on running time of an algorithm. We must know the case that causes minimum number of
operations to be executed. In the linear search problem, the best case occurs when x is present at the first location. The number of operations in
the best case is constant (not dependent on n). So time complexity in the best case would be ?(1)

Most of the times, we do worst case analysis to analyze algorithms. In the worst analysis, we guarantee an upper bound on the running time of an
algorithm which is good information.
The average case analysis is not easy to do in most of the practical cases and it is rarely done. In the average case analysis, we must know (or
predict) the mathematical distribution of all possible inputs.
The Best Case analysis is bogus. Guaranteeing a lower bound on an algorithm doesnt provide any information as in the worst case, an algorithm
may take years to run.

For some algorithms, all the cases are asymptotically same, i.e., there are no worst and best cases. For example, Merge Sort. Merge Sort does ?
(nLogn) operations in all cases. Most of the other sorting algorithms have worst and best cases. For example, in the typical implementation of
Quick Sort (where pivot is chosen as a corner element), the worst occurs when the input array is already sorted and the best occur when the pivot
elements always divide array in two halves. For insertion sort, the worst case occurs when the array is reverse sorted and the best case occurs
when the array is sorted in the same order as output.
References:
MITs Video lecture 1 on Introduction to Algorithms.

Analysis of Algorithms | Set 3 (Asymptotic Notations)

We have discussed Asymptotic Analysis, and Worst, Average and Best Cases of Algorithms. The main idea of asymptotic analysis is to have a
measure of efficiency of algorithms that doesnt depend on machine specific constants, and doesnt require algorithms to be implemented and time
taken by programs to be compared. Asymptotic notations are mathematical tools to represent time complexity of algorithms for asymptotic
analysis. The following 3 asymptotic notations are mostly used to represent time complexity of algorithms.

1) ? Notation: The theta notation bounds a functions from above and below, so it defines exact asymptotic behavior.
A simple way to get Theta notation of an expression is to drop low order terms and ignore leading constants. For example, consider the following
expression.
3n3 + 6n2 + 6000 = ?(n3)
Dropping lower order terms is always fine because there will always be a n0 after which ?(n3) beats ?n2) irrespective of the constants involved.
For a given function g(n), we denote ?(g(n)) is following set of functions.
?((g(n)) = {f(n): there exist positive constants c1, c2 and n0 such that
0 <= c1*g(n) <= f(n) <= c2*g(n) for all n >= n0}

The above definition means, if f(n) is theta of g(n), then the value f(n) is always between c1*g(n) and c2*g(n) for large values of n (n >= n0). The
definition of theta also requires that f(n) must be non-negative for values of n greater than n0.

2) Big O Notation: The Big O notation defines an upper bound of an algorithm, it bounds a function only from
above. For example, consider the case of Insertion Sort. It takes linear time in best case and quadratic time in worst case. We can safely say that
the time complexity of Insertion sort is O(n^2). Note that O(n^2) also covers linear time.
If we use ? notation to represent time complexity of Insertion sort, we have to use two statements for best and worst cases:
1. The worst case time complexity of Insertion Sort is ?(n^2).
2. The best case time complexity of Insertion Sort is ?(n).
The Big O notation is useful when we only have upper bound on time complexity of an algorithm. Many times we easily find an upper bound by
simply looking at the algorithm.
O(g(n)) = { f(n): there exist positive constants c and n0 such that
0 <= f(n) <= cg(n) for all n >= n0}

3) ? Notation: Just as Big O notation provides an asymptotic upper bound on a function, ? notation provides an
asymptotic lower bound.
? Notation< can be useful when we have lower bound on time complexity of an algorithm. As discussed in the previous post, the best case
performance of an algorithm is generally not useful, the Omega notation is the least used notation among all three.
For a given function g(n), we denote by ?(g(n)) the set of functions.

? (g(n)) = {f(n): there exist positive constants c and n0 such that

0 <= cg(n) <= f(n) for all n >= n0}.

Let us consider the same Insertion sort example here. The time complexity of Insertion Sort can be written as ?(n), but it is not a very useful
information about insertion sort, as we are generally interested in worst case and sometimes in average case.
Exercise:
Which of the following statements is/are valid?
1. Time Complexity of QuickSort is ?(n^2)
2. Time Complexity of QuickSort is O(n^2)
3. For any two functions f(n) and g(n), we have f(n) = ?(g(n)) if and only if f(n) = O(g(n)) and f(n) = ?(g(n)).
4. Time complexity of all computer algorithms can be written as ?(1)
References:
Lec 1 | MIT (Introduction to Algorithms)
Introduction to Algorithms 3rd Edition by Clifford Stein, Thomas H. Cormen, Charles E. Leiserson, Ronald L. Rivest

Analysis of Algorithms | Set 4 (Analysis of Loops)

We have discussed Asymptotic Analysis, Worst, Average and Best Cases and Asymptotic Notations in previous posts. In this post, analysis of
iterative programs with simple examples is discussed.
1) O(1): Time complexity of a function (or set of statements) is considered as O(1) if it doesnt contain loop, recursion and call to any other nonconstant time function.
// set of non-recursive and non-loop statements

For example swap() function has O(1) time complexity.

A loop or recursion that runs a constant number of times is also considered as O(1). For example the following loop is O(1).
// Here c is a constant
for (int i = 1; i <= c; i++) {
// some O(1) expressions
}

2) O(n): Time Complexity of a loop is considered as O(n) if the loop variables is incremented / decremented by a constant amount. For example
following functions have O(n) time complexity.
// Here c is a positive integer constant
for (int i = 1; i <= n; i += c) {
// some O(1) expressions
}
for (int i = n; i > 0; i -= c) {
// some O(1) expressions
}

3) O(nc): Time complexity of nested loops is equal to the number of times the innermost statement is executed. For example the following sample
loops have O(n2) time complexity
for (int i = 1; i <=n; i += c) {
for (int j = 1; j <=n; j += c) {
// some O(1) expressions
}
}
for (int i = n; i > 0; i += c) {
for (int j = i+1; j <=n; j += c) {
// some O(1) expressions
}

For example Selection sort and Insertion Sort have O(n2) time complexity.
4) O(Logn) Time Complexity of a loop is considered as O(Logn) if the loop variables is divided / multiplied by a constant amount.
for (int i = 1; i <=n; i *= c) {
// some O(1) expressions
}
for (int i = n; i > 0; i /= c) {
// some O(1) expressions
}

For example Binary Search(refer iterative implementation) has O(Logn) time complexity.
5) O(LogLogn) Time Complexity of a loop is considered as O(LogLogn) if the loop variables is reduced / increased exponentially by a constant
amount.
// Here c is a constant greater than 1
for (int i = 2; i <=n; i = pow(i, c)) {
// some O(1) expressions
}
//Here fun is sqrt or cuberoot or any other constant root
for (int i = n; i > 0; i = fun(i)) {
// some O(1) expressions
}

See this for more explanation.

How to combine time complexities of consecutive loops?

When there are consecutive loops, we calculate time complexity as sum of time complexities of individual loops.
for (int i = 1; i <=m; i += c) {
// some O(1) expressions
}
for (int i = 1; i <=n; i += c) {
// some O(1) expressions
}
Time complexity of above code is O(m) + O(n) which is O(m+n)
If m == n, the time complexity becomes O(2n) which is O(n).

How to calculate time complexity when there are many if, else statements inside loops?
As discussed here, worst case time complexity is the most useful among best, average and worst. Therefore we need to consider worst case. We
evaluate the situation when values in if-else conditions cause maximum number of statements to be executed.
For example consider the linear search function where we consider the case when element is present at the end or not present at all.
When the code is too complex to consider all if-else cases, we can get an upper bound by ignoring if else and other complex control statements.
How to calculate time complexity of recursive functions?
Time complexity of a recursive function can be written as a mathematical recurrence relation. To calculate time complexity, we must know how to
solve recurrences. We will soon be discussing recurrence solving techniques as a separate post.
Quiz on Analysis of Algorithms

Analysis of Algorithm | Set 4 (Solving Recurrences)

In the previous post, we discussed analysis of loops. Many algorithms are recursive in nature. When we analyze them, we geta recurrence relation
for time complexity. We get running time on an input of size n as a function of n and the running time on inputs of smaller sizes. For example in
Merge Sort, to sort a given array, we divide it in two halves and recursively repeat the process for the two halves. Finally we merge the results.
Time complexity of Merge Sort can be written as T(n) = 2T(n/2) + cn. There are many other algorithms like Binary Search, Tower of Hanoi, etc.
There are mainly three ways for solving recurrences.
1) Substitution Method: We make a guess for the solution and then we use mathematical induction to prove the the guess is correct or incorrect.
For example consider the recurrence T(n) = 2T(n/2) + n
We guess the solution as T(n) = O(nLogn). Now we use induction
to prove our guess.
We need to prove that T(n) <= cnLogn. We can assume that it is true
for values smaller than n.
T(n) =
<=
=
=
<=

2T(n/2) + n
cn/2Log(n/2) + n
cnLogn - cnLog2 + n
cnLogn - cn + n
cnLogn

2) Recurrence Tree Method: In this method, we draw a recurrence tree and calculate the time taken by every level of tree. Finally, we sum the
work done at all levels. To draw the recurrence tree, we start from the given recurrence and keep drawing till we find a pattern among levels. The
pattern is typically a arithmetic or geometric series.
For example consider the recurrence relation
T(n) = T(n/4) + T(n/2) + cn2
cn2
/
T(n/4)

\
T(n/2)

If we further break down the expression T(n/4) and T(n/2),

we get following recursion tree.
cn2
/
\
c(n2)/16
c(n2)/4
/
\
/
\
T(n/16)
T(n/8) T(n/8)
T(n/4)
Breaking down further gives us following
cn2
/
\
c(n2)/16
c(n2)/4
/
\
/
\
c(n2)/256 c(n2)/64 c(n2)/64
c(n2)/16
/
\
/
\
/
\
/
\
To know the value of T(n), we need to calculate sum of tree
nodes level by level. If we sum the above tree level by level,
we get the following series
T(n) = c(n^2 + 5(n^2)/16 + 25(n^2)/256) + ....
The above series is geometrical progression with ratio 5/16.
To get an upper bound, we can sum the infinite series.
We get the sum as (n2)/(1 - 5/16) which is O(n2)

3) Master Method:
Master Method is a direct way to get the solution. The master method works only for following type of recurrences or for recurrences that can be
transformed to following type.
T(n) = aT(n/b) + f(n) where a >= 1 and b > 1

There are following three cases:

1. If f(n) = ?(nc) where c < Logba then T(n) = ?(nLogba)
2. If f(n) = ?(nc) where c = Logba then T(n) = ?(ncLog n)
3.If f(n) = ?(nc) where c > Logba then T(n) = ?(f(n))

How does this work?

Master method is mainly derived from recurrence tree method. If we draw recurrence tree of T(n) = aT(n/b) + f(n), we can see that the work done
at root is f(n) and work done at all leaves is ?(nc) where c is Logba. And the height of recurrence tree is Logbn

In recurrence tree method, we calculate total work done. If the work done at leaves is polynomially more, then leaves are the dominant part, and
our result becomes the work done at leaves (Case 1). If work done at leaves and root is asymptotically same, then our result becomes height
multiplied by work done at any level (Case 2). If work done at root is asymptotically more, then our result becomes work done at root (Case 3).
Examples of some standard algorithms whose time complexity can be evaluated using Master Method
Merge Sort: T(n) = 2T(n/2) + ?(n). It falls in case 2 as c is 1 and Logba] is also 1. So the solution is ?(n Logn)
Binary Search: T(n) = T(n/2) + ?(1). It also falls in case 2 as c is 0 and Logba is also 0. So the solution is ?(Logn)
Notes:
1) It is not necessary that a recurrence of the form T(n) = aT(n/b) + f(n) can be solved using Master Theorem. The given three cases have some
gaps between them. For example, the recurrence T(n) = 2T(n/2) + n/Logn cannot be solved using master method.
2) Case 2 can be extended for f(n) = ?(ncLogkn)
If f(n) = ?(ncLogkn) for some constant k >= 0 and c = Logba, then T(n) = ?(ncLogk+1n)
Practice Problems and Solutions on Master Theorem.
References:
http://en.wikipedia.org/wiki/Master_theorem
MIT Video Lecture on Asymptotic Notation | Recurrences | Substitution, Master Method
Introduction to Algorithms 3rd Edition by Clifford Stein, Thomas H. Cormen, Charles E. Leiserson, Ronald L. Rivest

Analysis of Algorithm | Set 5 (Amortized Analysis Introduction)

Amortized Analysis is used for algorithms where an occasional operation is very slow, but most of the other operations are faster. In Amortized
Analysis, we analyze a sequence of operations and guarantee a worst case average time which is lower than the worst case time of a particular
expensive operation.
The example data structures whose operations are analyzed using Amortized Analysis are Hash Tables, Disjoint Sets and Splay Trees.
Let us consider an example of a simple hash table insertions. How do we decide table size? There is a trade-off between space and time, if we
make hash-table size big, search time becomes fast, but space required becomes high.

The solution to this trade-off problem is to use Dynamic Table (or Arrays). The idea is to increase size of table whenever it becomes full. Following
are the steps to follow when table becomes full.
1) Allocate memory for a larger table of size, typically twice the old table.
2) Copy the contents of old table to new table.
3) Free the old table.
If the table has space available, we simply insert new item in available space.
What is the time complexity of n insertions using the above scheme?
If we use simple analysis, the worst case cost of an insertion is O(n). Therefore, worst case cost of n inserts is n * O(n) which is O(n2). This
analysis gives an upper bound, but not a tight upper bound for n insertions as all insertions dont take ?(n) time.

So using Amortized Analysis, we could prove that the Dynamic Table scheme has O(1) insertion time which is a great result used in hashing. Also,
the concept of dynamic table is used in vectors in C++, ArrayList in Java.
Following are few important notes.
1) Amortized cost of a sequence of operations can be seen as expenses of a salaried person. The average monthly expense of the person is less
than or equal to the salary, but the person can spend more money in a particular month by buying a car or something. In other months, he or she
saves money for the expensive month.
2) The above Amortized Analysis done for Dynamic Array example is called Aggregate Method. There are two more powerful ways to do
Amortized analysis called Accounting Method and Potential Method. We will be discussing the other two methods in separate posts.
3) The amortized analysis doesnt involve probability. There is also another different notion of average case running time where algorithms use
randomization to make them faster and expected running time is faster than the worst case running time. These algorithms are analyzed using

Randomized Analysis. Examples of these algorithms are Randomized Quick Sort, Quick Select and Hashing. We will soon be covering
Randomized analysis in a different post.
Sources:
Berkeley Lecture 35: Amortized Analysis
MIT Lecture 13: Amortized Algorithms, Table Doubling, Potential Method
http://www.cs.cornell.edu/courses/cs3110/2011sp/lectures/lec20-amortized/amortized.htm

What does Space Complexity mean?

Space Complexity:
The term Space Complexity is misused for Auxiliary Space at many places. Following are the correct definitions of Auxiliary Space and Space
Complexity.
Auxiliary Space is the extra space or temporary space used by an algorithm.
Space Complexity of an algorithm is total space taken by the algorithm with respect to the input size. Space complexity includes both Auxiliary
space and space used by input.
For example, if we want to compare standard sorting algorithms on the basis of space, then Auxiliary Space would be a better criteria than Space
Complexity. Merge Sort uses O(n) auxiliary space, Insertion sort and Heap Sort use O(1) auxiliary space. Space complexity of all these sorting
algorithms is O(n) though.

NP-Completeness | Set 1 (Introduction)

We have been writing about efficient algorithms to solve complex problems, like shortest path, Euler graph, minimum spanning tree, etc. Those
were all success stories of algorithm designers. In this post, failure stories of computer science are discussed.
Can all computational problems be solved by a computer? There are computational problems that can not be solved by algorithms even with
unlimited time. For example Turing Halting problem (Given a program and an input, whether the program will eventually halt when run with that
input, or will run forever). Alan Turing proved that general algorithm to solve the halting problem for all possible program-input pairs cannot exist.
A key part of the proof is, Turing machine was used as a mathematical definition of a computer and program (Source Halting Problem).
Status of NP Complete problems is another failure story, NP complete problems are problems whose status is unknown. No polynomial time
algorithm has yet been discovered for any NP complete problem, nor has anybody yet been able to prove that no polynomial-time algorithm exist
for any of them. The interesting part is, if any one of the NP complete problems can be solved in polynomial time, then all of them can be solved.
What are NP, P, NP-complete and NP-Hard problems?
P is set of problems that can be solved by a deterministic Turing machine in Polynomial time.
NP is set of decision problems that can be solved by a Non-deterministic Turing Machine in Polynomial time. P is subset of NP (any problem that
can be solved by deterministic machine in polynomial time can also be solved by non-deterministic machine in polynomial time).
Informally, NP is set of decision problems which can be solved by a polynomial time via a Lucky Algorithm, a magical algorithm that always makes
a right guess among the given set of choices (Source Ref 1).
NP-complete problems are the hardest problems in NP set. A decision problem L is NP-complete if:

1) L is in NP (Any given solution for NP-complete problems can be verified quickly, but there is no efficient known solution).
2) Every problem in NP is reducible to L in polynomial time (Reduction is defined below).
A problem is NP-Hard if it follows property 2 mentioned above, doesnt need to follow property 1. Therefore, NP-Complete set is also a subset
of NP-Hard set.

Decision vs Optimization Problems

NP-completeness applies to the realm of decision problems. It was set up this way because its easier to compare the difficulty of decision
problems than that of optimization problems. In reality, though, being able to solve a decision problem in polynomial time will often permit us to
solve the corresponding optimization problem in polynomial time (using a polynomial number of calls to the decision problem). So, discussing the
difficulty of decision problems is often really equivalent to discussing the difficulty of optimization problems.(Source Ref 2).
For example, consider the vertex cover problem (Given a graph, find out the minimum sized vertex set that covers all edges). It is an optimization
problem. Corresponding decision problem is, given undirected graph G and k, is there a vertex cover of size k?
What is Reduction?

Let L1 and L2 be two decision problems. Suppose algorithm A2 solves L2. That is, if y is an input for L2 then algorithm A2 will answer Yes or No
depending upon whether y belongs to L2 or not.
The idea is to find a transformation from L1 to L2 so that the algorithm A2 can be part of an algorithm A1 to solve L1.

Learning reduction in general is very important. For example, if we have library functions to solve certain problem and if we can reduce a new
problem to one of the solved problems, we save a lot of time. Consider the example of a problem where we have to find minimum product path in
a given directed graph where product of path is multiplication of weights of edges along the path. If we have code for Dijkstras algorithm to find
shortest path, we can take log of all weights and use Dijkstras algorithm to find the minimum product path rather than writing a fresh code for this
new problem.

How to prove that a given problem is NP complete?

From the definition of NP-complete, it appears impossible toprove that a problem L is NP-Complete. By definition, it requires us to that show
every problem in NP is polynomial time reducible to L. Fortunately, there is an alternate way to prove it. The idea is to take a known NPComplete problem and reduce it to L. If polynomial time reduction is possible, we can prove that L is NP-Complete by transitivity of reduction (If
a NP-Complete problem is reducible to L in polynomial time, then all problems are reducible to L in polynomial time).
What was the first problem proved as NP-Complete?
There must be some first NP-Complete problem proved by definition of NP-Complete problems. SAT (Boolean satisfiability problem) is the first
NP-Complete problem proved by Cook (See CLRS book for proof).
It is always useful to know about NP-Completeness even for engineers. Suppose you are asked to write an efficient algorithm to solve an
extremely important problem for your company. After a lot of thinking, you can only come up exponential time approach which is impractical. If
you dont know about NP-Completeness, you can only say that I could not come with an efficient algorithm. If you know about NP-Completeness
and prove that the problem as NP-complete, you can proudly say that the polynomial time solution is unlikely to exist. If there is a polynomial time
solution possible, then that solution solves a big problem of computer science many scientists have been trying for years.
We will soon be discussing more NP-Complete problems and their proof for NP-Completeness.
References:
MIT Video Lecture on Computational Complexity
Introduction to Algorithms 3rd Edition by Clifford Stein, Thomas H. Cormen, Charles E. Leiserson, Ronald L. Rivest
http://www.ics.uci.edu/~eppstein/161/960312.html

A Time Complexity Question

What is the time complexity of following function fun()? Assume that log(x) returns log value in base 2.
void fun()
{
int i, j;
for (i=1; i<=n; i++)
for (j=1; j<=log(i); j++)
printf("GeeksforGeeks");
}

Time Complexity of the above function can be written as ?(log 1) + ?(log 2) + ?(log 3) + . . . . + ?(log n) which is ? (log n!)
Order of growth of log n! and n log n is same for large values of n, i.e., ? (log n!) = ?(n log n). So time complexity of fun() is ?(n log n).
The expression ?(log n!) = ?(n log n) can be easily derived from following Stirlings approximation (or Stirlings formula).
log n! = n log n - n + O(log(n))

Sources:
http://en.wikipedia.org/wiki/Stirling%27s_approximation

Time Complexity of building a heap

Consider the following algorithm for building a Heap of an input array A.
BUILD-HEAP(A)
heapsize := size(A);
for i := floor(heapsize/2) downto 1
do HEAPIFY(A, i);
end for
END

What is the worst case time complexity of the above algo?

Although the worst case complexity looks like O(nLogn), upper bound of time complexity is O(n). See following links for the proof of time
complexity.
http://www.cse.iitk.ac.in/users/sbaswana/Courses/ESO211/heap.pdf/
http://www.cs.sfu.ca/CourseCentral/307/petra/2009/SLN_2.pdf

Time Complexity where loop variable is incremented by 1, 2, 3, 4 ..

What is the time complexity of below code?
void fun(int n)
{
int j = 1, i = 0;
while (i < n)
{
// Some O(1) task
i = i + j;
j++;
}
}

The loop variable i is incremented by 1, 2, 3, 4, until i becomes greater than or equal to n.

The value of i is x(x+1)/2 after x iterations. So if loop runs x times, then x(x+1)/2 < n. Therefore time complexity can be written as ?(?n).

Time Complexity of Loop with Powers

What is the time complexity of below function?
void fun(int n, int k)
{
for (int i=1; i<=n; i++)
{
int p = pow(i, k);
for (int j=1; j<=p; j++)
{
// Some O(1) work
}
}
}

Time complexity of above function can be written as 1k + 2k + 3k + n1k.

Let us try few examples:
k=1
Sum =
=
=

1 + 2 + 3 ... n
n(n+1)/2
n2 + n/2

k=2
Sum =
=
=

12 + 22 + 32 + ... n12.
n(n+1)(2n+1)/6
n3/3 + n2/2 + n/6

k=3
Sum =
=
=

13 + 23 + 33 + ... n13.
n2(n+1)2/4
n4/4 + n3/2 + n2/4

In general, asymptotic value can be written as (nk+1)/(k+1) + ?(nk)

Note that, in asymptotic notations like ? we can always ignore lower order terms. So the time complexity is ?(nk+1 / (k+1))

Binary Search
Given a sorted array arr[] of n elements, write a function to search a given element x in arr[].
A simple approach is to do linear search, i.e., start from the leftmost element of arr[] and one by one compare x with each element of arr[], if x
matches with an element, return the index. If x doesnt match with any of elements, return -1.

C/C++
// Linearly search x in arr[]. If x is present then return its
// location, otherwise return -1
int search(int arr[], int n, int x)
{
int i;
for (i=0; i<n; i++)
if (arr[i] == x)
return i;
return -1;
}

Python
#
#
#
#
#

Searching an element in a list/array in python

can be simply done using 'in' operator
Example:
if x in arr:
print arr.index(x)

# If you want to implement Linear Search in python

# Linearly search x in arr[]
# If x is present then return its location
# else return -1
def search(arr, x):
for i in range(len(arr)):
if arr[i] == x:
return i
return -1

The idea of binary search is to use the information that the array is sorted and reduce the time complexity to O(Logn). We basically ignore half of
the elements just after one comparison.
1) Compare x with the middle element.
2) If x matches with middle element, we return the mid index.
3) Else If x is greater than the mid element, then x can only lie in right half subarray after the mid element. So we recur for right half.
4) Else (x is smaller) recur for the left half.
Following is Recursive implementation of Binary Search.

C/C++
#include <stdio.h>
// A recursive binary search function. It returns location of x in
// given array arr[l..r] is present, otherwise -1
int binarySearch(int arr[], int l, int r, int x)
{
if (r >= l)
{
int mid = l + (r - l)/2;
// If the element is present at the middle itself
if (arr[mid] == x) return mid;
// If element is smaller than mid, then it can only be present
// in left subarray
if (arr[mid] > x) return binarySearch(arr, l, mid-1, x);
// Else the element can only be present in right subarray

return binarySearch(arr, mid+1, r, x);

}
// We reach here when element is not present in array
return -1;
}
int main(void)
{
int arr[] = {2, 3, 4, 10, 40};
int n = sizeof(arr)/ sizeof(arr[0]);
int x = 10;
int result = binarySearch(arr, 0, n-1, x);
(result == -1)? printf("Element is not present in array")
: printf("Element is present at index %d", result);
return 0;
}

Python
# Python Program for recursive binary search.
# Returns index of x in arr if present, else -1
def binarySearch (arr, l, r, x):
# Check base case
if r >= l:
mid = l + (r - l)/2
# If element is present at the middle itself
if arr[mid] == x:
return mid
# If element is smaller than mid, then it can only
# be present in left subarray
elif arr[mid] > x:
return binarySearch(arr, l, mid-1, x)
# Else the element can only be present in right subarray
else:
return binarySearch(arr, mid+1, r, x)
else:
# Element is not present in the array
return -1
# Test array
arr = [ 2, 3, 4, 10, 40 ]
x = 10
# Function call
result = binarySearch(arr, 0, len(arr)-1, x)
if result != -1:
print "Element is present at index %d" % result
else:
print "Element is not present in array"

Element is present at index 3

Following is Iterative implementation of Binary Search.

C/C++
#include <stdio.h>
// A iterative binary search function. It returns location of x in
// given array arr[l..r] if present, otherwise -1
int binarySearch(int arr[], int l, int r, int x)
{
while (l <= r)
{
int m = l + (r-l)/2;
if (arr[m] == x) return m; // Check if x is present at mid

if (arr[m] < x) l = m + 1; // If x greater, ignore left half

else r = m - 1; // If x is smaller, ignore right half
}
return -1; // if we reach here, then element was not present
}
int main(void)
{
int arr[] = {2, 3, 4, 10, 40};
int n = sizeof(arr)/ sizeof(arr[0]);
int x = 10;
int result = binarySearch(arr, 0, n-1, x);
(result == -1)? printf("Element is not present in array")
: printf("Element is present at index %d", result);
return 0;
}

Python
# Iterative Binary Search Function
# It returns location of x in given array arr if present,
# else returns -1
def binarySearch(arr, l, r, x):
while l<=r:
mid = l + (r - l)/2;
# Check if x is present at mid
if arr[mid] == x:
return mid
# If x is greater, ignore left half
elif arr[mid] < x:
l = mid + 1
# If x is smaller, ignore right half
else:
r = mid - 1
# If we reach here, then the element was not present
return -1
# Test array
arr = [ 2, 3, 4, 10, 40 ]
x = 10
# Function call
result = binarySearch(arr, 0, len(arr)-1, x)
if result != -1:
print "Element is present at index %d" % result
else:
print "Element is not present in array"

Element is present at index 3

Time Complexity:
The time complexity of Binary Search can be written as
T(n) = T(n/2) + c

The above recurrence can be solved either using Recurrence T ree method or Master method. It falls in case II of Master Method and solution of
the recurrence is
.
Auxiliary Space: O(1) in case of iterative implementation. In case of recursive implementation, O(Logn) recursion call stack space.
Algorithmic Paradigm: Divide and Conquer
Following are some interesting articles based on Binary Search.
The Ubiquitous Binary Search
Interpolation search vs Binary search
Find the minimum element in a sorted and rotated array

Find a peak element

Find a Fixed Point in a given array
Count the number of occurrences in a sorted array
Median of two sorted arrays
Floor and Ceiling in a sorted array
Find the maximum element in an array which is first increasing and then decreasing
GATE CornerQuiz Corner

Selection Sort
The selection sort algorithm sorts an array by repeatedly finding the minimum element (considering ascending order) from unsorted part and putting
it at the beginning. The algorithm maintains two subarrays in a given array.
1) The subarray which is already sorted.
2) Remaining subarray which is unsorted.
In every iteration of selection sort, the minimum element (considering ascending order) from the unsorted subarray is picked and moved to the
sorted subarray.
Following example explains the above steps:
arr[] = 64 25 12 22 11
// Find the minimum element in arr[0...4] and place it at beginning
11 25 12 22 64
// Find the minimum element in arr[1...4] and
// place it at beginning of arr[1...4]
11 12 25 22 64
// Find the minimum element in arr[2...4] and
// place it at beginning of arr[2...4]
11 12 22 25 64
// Find the minimum element in arr[3...4] and
// place it at beginning of arr[3...4]
11 12 22 25 64
// C program for implementation of selection sort
#include <stdio.h>
void swap(int *xp, int *yp)
{
int temp = *xp;
*xp = *yp;
*yp = temp;
}
void selectionSort(int arr[], int n)
{
int i, j, min_idx;
// One by one move boundary of unsorted subarray
for (i = 0; i < n-1; i++)
{
// Find the minimum element in unsorted array
min_idx = i;
for (j = i+1; j < n; j++)
if (arr[j] < arr[min_idx])
min_idx = j;
// Swap the found minimum element with the first element
swap(&arr[min_idx], &arr[i]);
}
}
/* Function to print an array */
void printArray(int arr[], int size)
{
int i;
for (i=0; i < size; i++)
printf("%d ", arr[i]);
printf("\n");
}
// Driver program to test above functions
int main()
{
int arr[] = {64, 25, 12, 22, 11};
int n = sizeof(arr)/sizeof(arr[0]);
selectionSort(arr, n);
printf("Sorted array: \n");
printArray(arr, n);
return 0;
}

Output:

Sorted array:
11 12 22 25 64

Time Complexity: O(n*n) as there are two nested loops.

Auxiliary Space: O(1)
The good thing about selection sort is it never makes more than O(n) swaps and can be useful when memory write is a costly operation.
Other Sorting Algorithms on GeeksforGeeks/GeeksQuiz:
Bubble Sort
Insertion Sort
Merge Sort
Heap Sort
QuickSort
Radix Sort
Counting Sort
Bucket Sort
ShellSort

Bubble Sort
Bubble Sort is the simplest sorting algorithm that works by repeatedly swapping the adjacent elements if they are in wrong order.
Example:
First Pass:
( 5 1 4 2 8 ) > ( 1 5 4 2 8 ), Here, algorithm compares the first two elements, and swaps since 5 > 1.
( 1 5 4 2 8 ) > ( 1 4 5 2 8 ), Swap since 5 > 4
( 1 4 5 2 8 ) > ( 1 4 2 5 8 ), Swap since 5 > 2
( 1 4 2 5 8 ) > ( 1 4 2 5 8 ), Now, since these elements are already in order (8 > 5), algorithm does not swap them.
Second Pass:
(14258)>(14258)
( 1 4 2 5 8 ) > ( 1 2 4 5 8 ), Swap since 4 > 2
(12458)>(12458)
(12458)>(12458)
Now, the array is already sorted, but our algorithm does not know if it is completed. The algorithm needs one whole pass without any swap to
know it is sorted.
Third Pass:
(12458)>(12458)
(12458)>(12458)
(12458)>(12458)
(12458)>(12458)
Following is C implementation of Bubble Sort.
// C program for implementation of Bubble sort
#include <stdio.h>
void swap(int *xp, int *yp)
{
int temp = *xp;
*xp = *yp;
*yp = temp;
}
// A function to implement bubble sort
void bubbleSort(int arr[], int n)
{
int i, j;
for (i = 0; i < n-1; i++)
for (j = 0; j < n-i-1; j++) //Last i elements are already in place
if (arr[j] > arr[j+1])
swap(&arr[j], &arr[j+1]);
}
/* Function to print an array */
void printArray(int arr[], int size)
{
int i;
for (i=0; i < size; i++)
printf("%d ", arr[i]);
printf("\n");
}
// Driver program to test above functions
int main()
{
int arr[] = {64, 34, 25, 12, 22, 11, 90};
int n = sizeof(arr)/sizeof(arr[0]);
bubbleSort(arr, n);
printf("Sorted array: \n");
printArray(arr, n);
return 0;
}

Output:
Sorted array:
11 12 22 25 34 64 90

Optimized Implementation:
The above function always runs O(n^2) time even if the array is sorted. It can be optimized by stopping the algorithm if inner loop didnt cause any

swap.
// Optimized implementation of Bubble sort
#include <stdio.h>
void swap(int *xp, int *yp)
{
int temp = *xp;
*xp = *yp;
*yp = temp;
}
// An optimized version of Bubble Sort
void bubbleSort(int arr[], int n)
{
int i, j;
bool swapped;
for (i = 0; i < n-1; i++)
{
swapped = false;
for (j = 0; j < n-i-1; j++)
{
if (arr[j] > arr[j+1])
{
swap(&arr[j], &arr[j+1]);
swapped = true;
}
}
// IF no two elements were swapped by inner loop, then break
if (swapped == false)
break;
}
}
/* Function to print an array */
void printArray(int arr[], int size)
{
int i;
for (i=0; i < size; i++)
printf("%d ", arr[i]);
printf("\n");
}
// Driver program to test above functions
int main()
{
int arr[] = {64, 34, 25, 12, 22, 11, 90};
int n = sizeof(arr)/sizeof(arr[0]);
bubbleSort(arr, n);
printf("Sorted array: \n");
printArray(arr, n);
return 0;
}

Output:
Sorted array:
11 12 22 25 34 64 90

Worst and Average Case Time Complexity: O(n*n). Worst case occurs when array is reverse sorted.
Best Case Time Complexity: O(n). Best case occurs when array is already sorted.
Auxiliary Space: O(1)
Boundary Cases: Bubble sort takes minimum time (Order of n) when elements are already sorted.
Sorting In Place: Yes
Stable: Yes
Due to its simplicity, bubble sort is often used to introduce the concept of a sorting algorithm.
In computer graphics it is popular for its capability to detect a very small error (like swap of just two elements) in almost-sorted arrays and fix it
with just linear complexity (2n). For example, it is used in a polygon filling algorithm, where bounding lines are sorted by their x coordinate at a
specific scan line (a line parallel to x axis) and with incrementing y their order changes (two elements are swapped) only at intersections of two lines
(Source: Wikipedia)

Other Sorting Algorithms on GeeksforGeeks/GeeksQuiz:

Selection Sort
Insertion Sort
Merge Sort
Heap Sort
QuickSort
Radix Sort
Counting Sort
Bucket Sort
ShellSort

Insertion Sort
Insertion sort is a simple sorting algorithm that works the way we sort playing cards in our hands.

Algorithm
// Sort an arr[] of size n
insertionSort(arr, n)
Loop from i = 1 to n-1.
a) Pick element arr[i] and insert it into sorted sequence arr[0i-1]
Example:
12, 11, 13, 5, 6
Let us loop for i = 1 (second element of the array) to 5 (Size of input array)
i = 1. Since 11 is smaller than 12, move 12 and insert 11 before 12
11, 12, 13, 5, 6
i = 2. 13 will remain at its position as all elements in A[0..I-1] are smaller than 13
11, 12, 13, 5, 6
i = 3. 5 will move to the beginning and all other elements from 11 to 13 will move one position ahead of their current position.
5, 11, 12, 13, 6
i = 4. 6 will move to position after 5, and elements from 11 to 13 will move one position ahead of their current position.
5, 6, 11, 12, 13
// C program for insertion sort
#include <stdio.h>
#include <math.h>
/* Function to sort an array using insertion sort*/
void insertionSort(int arr[], int n)
{
int i, key, j;
for (i = 1; i < n; i++)
{
key = arr[i];
j = i-1;
/* Move elements of arr[0..i-1], that are
greater than key, to one position ahead
of their current position */
while (j >= 0 && arr[j] > key)
{
arr[j+1] = arr[j];
j = j-1;
}
arr[j+1] = key;
}
}
// A utility function ot print an array of size n
void printArray(int arr[], int n)
{
int i;
for (i=0; i < n; i++)
printf("%d ", arr[i]);
printf("\n");

/* Driver program to test insertion sort */

int main()
{
int arr[] = {12, 11, 13, 5, 6};
int n = sizeof(arr)/sizeof(arr[0]);
insertionSort(arr, n);
printArray(arr, n);
return 0;
}

Output:
5 6 11 12 13

Time Complexity: O(n*n)

Auxiliary Space: O(1)
Boundary Cases: Insertion sort takes maximum time to sort if elements are sorted in reverse order. And it takes minimum time (Order of n) when
elements are already sorted.
Algorithmic Paradigm: Incremental Approach
Sorting In Place: Yes
Stable: Yes
Online: Yes
Uses: Insertion sort is uses when number of elements is small. It can also be useful when input array is almost sorted, only few elements are
misplaced in complete big array.
Quizzes: Sorting Questions
Other Sorting Algorithms on GeeksforGeeks/GeeksQuiz:
Selection Sort
Bubble Sort
Merge Sort
Heap Sort
QuickSort
Radix Sort
Counting Sort
Bucket Sort
ShellSort

Merge Sort
MergeSort is a Divide and Conquer algorithm. It divides input array in two halves, calls itself for the two halves and then merges the two sorted
halves. The merg() function is used for merging two halves. The merge(arr, l, m, r) is key process that assumes that arr[l..m] and arr[m+1..r] are
sorted and merges the two sorted sub-arrays into one. See following C implementation for details.
MergeSort(arr[], l, r)
If r > l
1. Find the middle point to divide the array into two halves:
middle m = (l+r)/2
2. Call mergeSort for first half:
Call mergeSort(arr, l, m)
3. Call mergeSort for second half:
Call mergeSort(arr, m+1, r)
4. Merge the two halves sorted in step 2 and 3:
Call merge(arr, l, m, r)

The following diagram from wikipedia shows the complete merge sort process for an example array {38, 27, 43, 3, 9, 82, 10}. If we take a closer
look at the diagram, we can see that the array is recursively divided in two halves till the size becomes 1. Once the size becomes 1, the merge
processes comes into action and starts merging arrays back till the complete array is merged.

/* C program for merge sort */

#include<stdlib.h>
#include<stdio.h>
/* Function to merge the two haves arr[l..m] and arr[m+1..r] of array arr[] */
void merge(int arr[], int l, int m, int r)
{
int i, j, k;
int n1 = m - l + 1;
int n2 = r - m;
/* create temp arrays */
int L[n1], R[n2];
/* Copy data to temp arrays L[] and R[] */
for(i = 0; i < n1; i++)
L[i] = arr[l + i];
for(j = 0; j < n2; j++)
R[j] = arr[m + 1+ j];
/* Merge the temp arrays back into arr[l..r]*/
i = 0;
j = 0;
k = l;
while (i < n1 && j < n2)
{
if (L[i] <= R[j])
{
arr[k] = L[i];
i++;

}
else
{
arr[k] = R[j];
j++;
}
k++;
}
/* Copy the remaining elements of L[], if there are any */
while (i < n1)
{
arr[k] = L[i];
i++;
k++;
}
/* Copy the remaining elements of R[], if there are any */
while (j < n2)
{
arr[k] = R[j];
j++;
k++;
}
}
/* l is for left index and r is right index of the sub-array
of arr to be sorted */
void mergeSort(int arr[], int l, int r)
{
if (l < r)
{
int m = l+(r-l)/2; //Same as (l+r)/2, but avoids overflow for large l and h
mergeSort(arr, l, m);
mergeSort(arr, m+1, r);
merge(arr, l, m, r);
}
}
/* UITLITY FUNCTIONS */
/* Function to print an array */
void printArray(int A[], int size)
{
int i;
for (i=0; i < size; i++)
printf("%d ", A[i]);
printf("\n");
}
/* Driver program to test above functions */
int main()
{
int arr[] = {12, 11, 13, 5, 6, 7};
int arr_size = sizeof(arr)/sizeof(arr[0]);
printf("Given array is \n");
printArray(arr, arr_size);
mergeSort(arr, 0, arr_size - 1);
printf("\nSorted array is \n");
printArray(arr, arr_size);
return 0;
}

Output:
Given array is
12 11 13 5 6 7
Sorted array is
5 6 7 11 12 13

Time Complexity: Sorting arrays on different machines. Merge Sort is a recursive algorithm and time complexity can be expressed as following
recurrence relation.
T(n) = 2T(n/2) +
The above recurrence can be solved either using Recurrence Tree method or Master method. It falls in case II of Master Method and solution of
the recurrence is
.
Time complexity of Merge Sort is
in all 3 cases (worst, average and best) as merge sort always divides the array in two halves and

take linear time to merge two halves.

Auxiliary Space: O(n)
Algorithmic Paradigm: Divide and Conquer
Sorting In Place: No in a typical implementation
Stable: Yes
Applications of Merge Sort
1) Merge Sort is useful for sorting linked lists in O(nLogn) time. Other nlogn algorithms like Heap Sort, Quick Sort (average case nLogn) cannot
be applied to linked lists.
2) Inversion Count Problem
3) Used in External Sorting
Other Sorting Algorithms on GeeksforGeeks/GeeksQuiz:
Selection Sort
Bubble Sort
Insertion Sort
Merge Sort
Heap Sort
QuickSort
Bucket Sort
ShellSort
Radix Sort
Counting Sort

Heap Sort
Heap sort is a comparison based sorting technique based on Binary Heap data structure. It is similar to selection sort where we first find the
maximum element and place the maximum element at the end. We repeat the same process for remaining element.
What is Binary Heap?
Let us first define a Complete Binary Tree. A complete binary tree is a binary tree in which every level, except possibly the last, is completely filled,
and all nodes are as far left as possible (Source Wikipedia)
A Binary Heap is a Complete Binary Tree where items are stored in a special order such that value in a parent node is greater(or smaller) than the
values in its two children nodes. The former is called as max heap and the latter is called min heap. The heap can be represented by binary tree or
array.
Why array based representation for Binary Heap?
Since a Binary Heap is a Complete Binary Tree, it can be easily represented as array and array based representation is space efficient. If the
parent node is stored at index I, the left child can be calculated by 2 * I + 1 and right child by 2 * I + 2.
Heap Sort Algorithm for sorting in increasing order:
1. Build a max heap from the input data.
2. At this point, the largest item is stored at the root of the heap. Replace it with the last item of the heap followed by reducing the size of heap by
1. Finally, heapify the root of tree.
3. Repeat above steps until size of heap is greater than 1.
How to build the heap?
Heapify procedure can be applied to a node only if its children nodes are heapified. So the heapification must be performed in the bottom up
order.
Lets understand with the help of an example:
Input data: 4, 10, 3, 5, 1
4(0)
/ \
10(1) 3(2)
/ \
5(3)
1(4)
The numbers in bracket represent the indices in the array
representation of data.
Applying heapify procedure to index 1:
4(0)
/ \
10(1)
3(2)
/ \
5(3)
1(4)
Applying heapify procedure to index 0:
10(0)
/ \
5(1) 3(2)
/ \
4(3)
1(4)
The heapify procedure calls itself recursively to build heap
in top down manner.
// C implementation of Heap Sort
#include <stdio.h>
#include <stdlib.h>
// A heap has current size and array of elements
struct MaxHeap
{
int size;
int* array;
};
// A utility function to swap to integers
void swap(int* a, int* b) { int t = *a; *a = *b; *b = t; }
// The main function to heapify a Max Heap. The function
// assumes that everything under given root (element at
// index idx) is already heapified
void maxHeapify(struct MaxHeap* maxHeap, int idx)
{
int largest = idx; // Initialize largest as root
int left = (idx << 1) + 1; // left = 2*idx + 1

int right = (idx + 1) << 1; // right = 2*idx + 2

// See if left child of root exists and is greater than
// root
if (left < maxHeap->size &&
maxHeap->array[left] > maxHeap->array[largest])
largest = left;
// See if right child of root exists and is greater than
// the largest so far
if (right < maxHeap->size &&
maxHeap->array[right] > maxHeap->array[largest])
largest = right;
// Change root, if needed
if (largest != idx)
{
swap(&maxHeap->array[largest], &maxHeap->array[idx]);
maxHeapify(maxHeap, largest);
}
}
// A utility function to create a max heap of given capacity
struct MaxHeap* createAndBuildHeap(int *array, int size)
{
int i;
struct MaxHeap* maxHeap =
(struct MaxHeap*) malloc(sizeof(struct MaxHeap));
maxHeap->size = size; // initialize size of heap
maxHeap->array = array; // Assign address of first element of array
// Start from bottommost and rightmost internal mode and heapify all
// internal modes in bottom up way
for (i = (maxHeap->size - 2) / 2; i >= 0; --i)
maxHeapify(maxHeap, i);
return maxHeap;
}
// The main function to sort an array of given size
void heapSort(int* array, int size)
{
// Build a heap from the input data.
struct MaxHeap* maxHeap = createAndBuildHeap(array, size);
// Repeat following steps while heap size is greater than 1.
// The last element in max heap will be the minimum element
while (maxHeap->size > 1)
{
// The largest item in Heap is stored at the root. Replace
// it with the last item of the heap followed by reducing the
// size of heap by 1.
swap(&maxHeap->array[0], &maxHeap->array[maxHeap->size - 1]);
--maxHeap->size; // Reduce heap size
// Finally, heapify the root of tree.
maxHeapify(maxHeap, 0);
}
}
// A utility function to print a given array of given size
void printArray(int* arr, int size)
{
int i;
for (i = 0; i < size; ++i)
printf("%d ", arr[i]);
}
/* Driver program to test above functions */
int main()
{
int arr[] = {12, 11, 13, 5, 6, 7};
int size = sizeof(arr)/sizeof(arr[0]);
printf("Given array is \n");
printArray(arr, size);
heapSort(arr, size);
printf("\nSorted array is \n");
printArray(arr, size);
return 0;

Output:
Given array is
12 11 13 5 6 7
Sorted array is
5 6 7 11 12 13

Notes:
Heap sort is an in-place algorithm.
Its typical implementation is not stable, but can be made stable (See this)
Time Complexity: Time complexity of heapify is O(Logn). Time complexity of createAndBuildHeap() is O(n) and overall time complexity of
Heap Sort is O(nLogn).
Applications of HeapSort
1. Sort a nearly sorted (or K sorted) array
2. k largest(or smallest) elements in an array
Heap sort algorithm has limited uses because Quicksort and Mergesort are better in practice. Nevertheless, the Heap data structure itself is
enormously used. See Applications of Heap Data Structure
Other Sorting Algorithms on GeeksforGeeks/GeeksQuiz:
Selection Sort
Bubble Sort
Insertion Sort
Merge Sort
QuickSort
Radix Sort
Counting Sort
Bucket Sort
ShellSort

QuickSort
Like Merge Sort, QuickSort is a Divide and Conquer algorithm. It picks an element as pivot and partitions the given array around the picked
pivot. There are many different versions of quickSort that pick pivot in different ways.
1) Always pick first element as pivot.
2) Always pick last element as pivot (implemented below)
3) Pick a random element as pivot.
4) Pick median as pivot.
The key process in quickSort is partition(). Target of partitions is, given an array and an element x of array as pivot, put x at its correct position in
sorted array and put all smaller elements (smaller than x) before x, and put all greater elements (greater than x) after x. All this should be done in
linear time.
Partition Algorithm
There can be many ways to do partition, following code adopts the method given in CLRS book. The logic is simple, we start from the leftmost
element and keep track of index of smaller (or equal to) elements as i. While traversing, if we find a smaller element, we swap current element with
arr[i]. Otherwise we ignore current element.
Implementation:
Following is C++ implementation of QuickSort.
/* A typical recursive implementation of quick sort */
#include<stdio.h>
// A utility function to swap two elements
void swap(int* a, int* b)
{
int t = *a;
*a = *b;
*b = t;
}
/* This function takes last element as pivot, places the pivot element at its
correct position in sorted array, and places all smaller (smaller than pivot)
to left of pivot and all greater elements to right of pivot */
int partition (int arr[], int l, int h)
{
int x = arr[h];
// pivot
int i = (l - 1); // Index of smaller element
for (int j = l; j <= h- 1; j++)
{
// If current element is smaller than or equal to pivot
if (arr[j] <= x)
{
i++;
// increment index of smaller element
swap(&arr[i], &arr[j]); // Swap current element with index
}
}
swap(&arr[i + 1], &arr[h]);
return (i + 1);
}
/* arr[] --> Array to be sorted, l --> Starting index, h --> Ending index */
void quickSort(int arr[], int l, int h)
{
if (l < h)
{
int p = partition(arr, l, h); /* Partitioning index */
quickSort(arr, l, p - 1);
quickSort(arr, p + 1, h);
}
}
/* Function to print an array */
void printArray(int arr[], int size)
{
int i;
for (i=0; i < size; i++)
printf("%d ", arr[i]);
printf("\n");
}
// Driver program to test above functions
int main()
{
int arr[] = {10, 7, 8, 9, 1, 5};

int n = sizeof(arr)/sizeof(arr[0]);
quickSort(arr, 0, n-1);
printf("Sorted array: \n");
printArray(arr, n);
return 0;
}

Output:
Sorted array:
1 5 7 8 9 10

Analysis of QuickSort
Time taken by QuickSort in general can be written as following.
T(n) = T(k) + T(n-k-1) +

(n)

The first two terms are for two recursive calls, the last term is for the partition process. k is the number of elements which are smaller than pivot.
The time taken by QuickSort depends upon the input array and partition strategy. Following are three cases.
Worst Case: The worst case occurs when the partition process always picks greatest or smallest element as pivot. If we consider above partition
strategy where last element is always picked as pivot, the worst case would occur when the array is already sorted in increasing or decreasing
order. Following is recurrence for worst case.
T(n) = T(0) + T(n-1) +
which is equivalent to
T(n) = T(n-1) + (n)

(n)

The solution of above recurrence is (n2).

Best Case: The best case occurs when the partition process always picks the middle element as pivot. Following is recurrence for best case.
T(n) = 2T(n/2) +

(n)

The solution of above recurrence is (nLogn). It can be solved using case 2 of Master Theorem.
Average Case:
To do average case analysis, we need to consider all possible permutation of array and calculate time taken by every permutation which doesnt
look easy.
We can get an idea of average case by considering the case when partition puts O(n/9) elements in one set and O(9n/10) elements in other set.
Following is recurrence for this case.
T(n) = T(n/9) + T(9n/10) +

(n)

Solution of above recurrence is also O(nLogn)

Although the worst case time complexity of QuickSort is O(n2) which is more than many other sorting algorithms like Merge Sort and Heap Sort,
QuickSort is faster in practice, because its inner loop can be efficiently implemented on most architectures, and in most real-world data. QuickSort
can be implemented in different ways by changing the choice of pivot, so that the worst case rarely occurs for a given type of data. However,
merge sort is generally considered better when data is huge and stored in external storage.
References:
http://en.wikipedia.org/wiki/Quicksort
Other Sorting Algorithms on GeeksforGeeks/GeeksQuiz:
Selection Sort
Bubble Sort
Insertion Sort
Merge Sort
Heap Sort
Radix Sort
Counting Sort
Bucket Sort
ShellSort

Radix Sort
The lower bound for Comparison based sorting algorithm (Merge Sort, Heap Sort, Quick-Sort .. etc) is ?(nLogn), i.e., they cannot do better than
nLogn.
Counting sort is a linear time sorting algorithm that sort in O(n+k) time when elements are in range from 1 to k.
What if the elements are in range from 1 to n2?
We cant use counting sort because counting sort will take O(n2) which is worse than comparison based sorting algorithms. Can we sort such an
array in linear time?
Radix Sort is the answer. The idea of Radix Sort is to do digit by digit sort starting from least significant digit to most significant digit. Radix sort
uses counting sort as a subroutine to sort.
The Radix Sort Algorithm
1) Do following for each digit i where i varies from least significant digit to the most significant digit.
.a) Sort input array using counting sort (or any stable sort) according to the ith digit.
Example:
Original, unsorted list:
170, 45, 75, 90, 802, 24, 2, 66
Sorting by least significant digit (1s place) gives: [*Notice that we keep 802 before 2, because 802 occurred before 2 in the original list, and
similarly for pairs 170 & 90 and 45 & 75.]
170, 90, 802,2, 24, 45, 75, 66
Sorting by next digit (10s place) gives: [*Notice that 802 again comes before 2 as 802 comes before 2 in the previous list.]
802, 2,24,45,66, 170,75,90
Sorting by most significant digit (100s place) gives:
2, 24, 45, 66, 75, 90,170,802
What is the running time of Radix Sort?
Let there be d digits in input integers. Radix Sort takes O(d*(n+b)) time where b is the base for representing numbers, for example, for decimal
system, b is 10. What is the value of d? If k is the maximum possible value, then d would be O(logb(k)). So overall time complexity is O((n+b) *
logb(k)). Which looks more than the time complexity of comparison based sorting algorithms for a large k. Let us first limit k. Let k <= nc where c
is a constant. In that case, the complexity becomes O(nLogb(n)). But it still doesnt beat comparison based sorting algorithms.
What if we make value of b larger?. What should be the value of b to make the time complexity linear? If we set b as n, we get the time complexity
as O(n). In other words, we can sort an array of integers with range from 1 to nc if the numbers are represented in base n (or every digit takes
log2(n) bits).
Is Radix Sort preferable to Comparison based sorting algorithms like Quick-Sort?
If we have log2n bits for every digit, the running time of Radix appears to be better than Quick Sort for a wide range of input numbers. The
constant factors hidden in asymptotic notation are higher for Radix Sort and Quick-Sort uses hardware caches more effectively. Also, Radix sort
uses counting sort as a subroutine and counting sort takes extra space to sort numbers.
Implementation of Radix Sort
Following is a simple C++ implementation of Radix Sort. For simplicity, the value of d is assumed to be 10. We recommend you to see Counting
Sort for details of countSort() function in below code.

C/C++
// C++ implementation of Radix Sort
#include<iostream>
using namespace std;
// A utility function to get maximum value in arr[]
int getMax(int arr[], int n)
{
int mx = arr[0];
for (int i = 1; i < n; i++)
if (arr[i] > mx)
mx = arr[i];
return mx;
}

// A function to do counting sort of arr[] according to

// the digit represented by exp.
void countSort(int arr[], int n, int exp)
{
int output[n]; // output array
int i, count[10] = {0};
// Store count of occurrences in count[]
for (i = 0; i < n; i++)
count[ (arr[i]/exp)%10 ]++;
// Change count[i] so that count[i] now contains actual
// position of this digit in output[]
for (i = 1; i < 10; i++)
count[i] += count[i - 1];
// Build the output array
for (i = n - 1; i >= 0; i--)
{
output[count[ (arr[i]/exp)%10 ] - 1] = arr[i];
count[ (arr[i]/exp)%10 ]--;
}
// Copy the output array to arr[], so that arr[] now
// contains sorted numbers according to current digit
for (i = 0; i < n; i++)
arr[i] = output[i];
}
// The main function to that sorts arr[] of size n using
// Radix Sort
void radixsort(int arr[], int n)
{
// Find the maximum number to know number of digits
int m = getMax(arr, n);
// Do counting sort for every digit. Note that instead
// of passing digit number, exp is passed. exp is 10^i
// where i is current digit number
for (int exp = 1; m/exp > 0; exp *= 10)
countSort(arr, n, exp);
}
// A utility function to print an array
void print(int arr[], int n)
{
for (int i = 0; i < n; i++)
cout << arr[i] << " ";
}
// Driver program to test above functions
int main()
{
int arr[] = {170, 45, 75, 90, 802, 24, 2, 66};
int n = sizeof(arr)/sizeof(arr[0]);
radixsort(arr, n);
print(arr, n);
return 0;
}

Java
// Radix sort Java implementation
import java.io.*;
import java.util.*;
class Radix {
// A utility function to get maximum value in arr[]
static int getMax(int arr[], int n)
{
int mx = arr[0];
for (int i = 1; i < n; i++)
if (arr[i] > mx)
mx = arr[i];
return mx;
}
// A function to do counting sort of arr[] according to
// the digit represented by exp.

static void countSort(int arr[], int n, int exp)

{
int output[] = new int[n]; // output array
int i;
int count[] = new int[10];
Arrays.fill(count,0);
// Store count of occurrences in count[]
for (i = 0; i < n; i++)
count[ (arr[i]/exp)%10 ]++;
// Change count[i] so that count[i] now contains
// actual position of this digit in output[]
for (i = 1; i < 10; i++)
count[i] += count[i - 1];
// Build the output array
for (i = n - 1; i >= 0; i--)
{
output[count[ (arr[i]/exp)%10 ] - 1] = arr[i];
count[ (arr[i]/exp)%10 ]--;
}
// Copy the output array to arr[], so that arr[] now
// contains sorted numbers according to curent digit
for (i = 0; i < n; i++)
arr[i] = output[i];
}
// The main function to that sorts arr[] of size n using
// Radix Sort
static void radixsort(int arr[], int n)
{
// Find the maximum number to know number of digits
int m = getMax(arr, n);
// Do counting sort for every digit. Note that instead
// of passing digit number, exp is passed. exp is 10^i
// where i is current digit number
for (int exp = 1; m/exp > 0; exp *= 10)
countSort(arr, n, exp);
}
// A utility function to print an array
static void print(int arr[], int n)
{
for (int i=0; i<n; i++)
System.out.print(arr[i]+" ");
}
/*Driver function to check for above function*/
public static void main (String[] args)
{
int arr[] = {170, 45, 75, 90, 802, 24, 2, 66};
int n = arr.length;
radixsort(arr, n);
print(arr, n);
}
}
/* This code is contributed by Devesh Agrawal */

2 24 45 66 75 90 170 802

Other Sorting Algorithms on GeeksforGeeks/GeeksQuiz:

Selection Sort
Bubble Sort
Insertion Sort
Merge Sort
Heap Sort
QuickSort
Counting Sort
Bucket Sort
ShellSort
References:

http://en.wikipedia.org/wiki/Radix_sort
http://alg12.wikischolars.columbia.edu/file/view/RADIX.pdf
MIT Video Lecture
Introduction to Algorithms 3rd Edition by Clifford Stein, Thomas H. Cormen, Charles E. Leiserson, Ronald L. Rivest

Counting Sort
Counting sort is a sorting technique based on keys between a specific range. It works by counting the number of objects having distinct key values
(kind of hashing). Then doing some arithmetic to calculate the position of each object in the output sequence.
Let us understand it with the help of an example.
For simplicity, consider the data in
Input data: 1, 4, 1, 2, 7, 5, 2
1) Take a count array to store the
Index:
0 1 2 3 4 5 6 7
Count:
0 2 2 0 1 1 0 1

the range 0 to 9.
count of each unique object.
8 9
0 0

2) Modify the count array such that each element at each index
stores the sum of previous counts.
Index:
0 1 2 3 4 5 6 7 8 9
Count:
0 2 4 4 5 6 6 7 7 7
The modified count array indicates the position of each object in
the output sequence.
3) Output each object from the input sequence followed by
decreasing its count by 1.
Process the input data: 1, 4, 1, 2, 7, 5, 2. Position of 1 is 2.
Put data 1 at index 2 in output. Decrease count by 1 to place
next data 1 at an index 1 smaller than this index.

Following is C implementation of counting sort.

// C Program for counting sort
#include <stdio.h>
#include <string.h>
#define RANGE 255
// The main function that sort the given string str in alphabatical order
void countSort(char *str)
{
// The output character array that will have sorted str
char output[strlen(str)];
// Create a count array to store count of inidividul characters and
// initialize count array as 0
int count[RANGE + 1], i;
memset(count, 0, sizeof(count));
// Store count of each character
for(i = 0; str[i]; ++i)
++count[str[i]];
// Change count[i] so that count[i] now contains actual position of
// this character in output array
for (i = 1; i <= RANGE; ++i)
count[i] += count[i-1];
// Build the output character array
for (i = 0; str[i]; ++i)
{
output[count[str[i]]-1] = str[i];
--count[str[i]];
}
// Copy the output array to str, so that str now
// contains sorted characters
for (i = 0; str[i]; ++i)
str[i] = output[i];
}
// Driver program to test above function
int main()
{
char str[] = "geeksforgeeks";//"applepp";
countSort(str);
printf("Sorted string is %s\n", str);
return 0;
}

Output:

Sorted character array is eeeefggkkorss

Time Complexity: O(n+k) where n is the number of elements in input array and k is the range of input.
Auxiliary Space: O(n+k)
Points to be noted:
1. Counting sort is efficient if the range of input data is not significantly greater than the number of objects to be sorted. Consider the situation
where the input sequence is between range 1 to 10K and the data is 10, 5, 10K, 5K.
2. It is not a comparison based sorting. It running time complexity is O(n) with space proportional to the range of data.
3. It is often used as a sub-routine to another sorting algorithm like radix sort.
4. Counting sort uses a partial hashing to count the occurrence of the data object in O(1).
5. Counting sort can be extended to work for negative inputs also.
Exercise:
1. Modify above code to sort the input data in the range from M to N.
2. Modify above code to sort negative input data.
3. Is counting sort stable and online?
4. Thoughts on parallelizing the counting sort algorithm.
Other Sorting Algorithms on GeeksforGeeks/GeeksQuiz:
Selection Sort
Bubble Sort
Insertion Sort
Merge Sort
Heap Sort
QuickSort
Radix Sort
Bucket Sort
ShellSort

Bucket Sort
Bucket sort is mainly useful when input is uniformly distributed over a range. For example, consider the following problem.
Sort a large set of floating point numbers which are in range from 0.0 to 1.0 and are uniformly distributed across the range. How do we
sort the numbers efficiently?
A simple way is to apply a comparison based sorting algorithm. The lower bound for Comparison based sorting algorithm (Merge Sort, Heap
Sort, Quick-Sort .. etc) is ?(n Log n), i.e., they cannot do better than nLogn.
Can we sort the array in linear time? Counting sort can not be applied here as we use keys as index in counting sort. Here keys are floating point
numbers.
The idea is to use bucket sort. Following is bucket algorithm.
bucketSort(arr[], n)
1) Create n empty buckets (Or lists).
2) Do following for every array element arr[i].
.......a) Insert arr[i] into bucket[n*array[i]]
3) Sort individual buckets using insertion sort.
4) Concatenate all sorted buckets.

Following diagram (taken from CLRS book) demonstrates working of bucket sort.

Time Complexity: If we assume that insertion in a bucket takes O(1) time then steps 1 and 2 of the above algorithm clearly take O(n) time. The
O(1) is easily possible if we use a linked list to represent a bucket (In the following code, C++ vector is used for simplicity). Step 4 also takes
O(n) time as there will be n items in all buckets.
The main step to analyze is step 3. This step also takes O(n) time on average if all numbers are uniformly distributed (please refer CLRS book for
more details)
Following is C++ implementation of the above algorithm.
// C++ program to sort an array using bucket sort
#include <iostream>
#include <algorithm>
#include <vector>
using namespace std;
// Function to sort arr[] of size n using bucket sort
void bucketSort(float arr[], int n)
{
// 1) Create n empty buckets
vector<float> b[n];
// 2) Put array elements in different buckets
for (int i=0; i<n; i++)
{
int bi = n*arr[i]; // Index in bucket
b[bi].push_back(arr[i]);
}
// 3) Sort individual buckets
for (int i=0; i<n; i++)
sort(b[i].begin(), b[i].end());
// 4) Concatenate all buckets into arr[]
int index = 0;
for (int i = 0; i < n; i++)
for (int j = 0; j < b[i].size(); j++)
arr[index++] = b[i][j];
}
/* Driver program to test above funtion */
int main()

{
float arr[] = {0.897, 0.565, 0.656, 0.1234, 0.665, 0.3434};
int n = sizeof(arr)/sizeof(arr[0]);
bucketSort(arr, n);
cout << "Sorted array is \n";
for (int i=0; i<n; i++)
cout << arr[i] << " ";
return 0;
}

Output:
Sorted array is
0.1234 0.3434 0.565 0.656 0.665 0.897

References:
Introduction to Algorithms 3rd Edition by Clifford Stein, Thomas H. Cormen, Charles E. Leiserson, Ronald L. Rivest
http://en.wikipedia.org/wiki/Bucket_sort
Other Sorting Algorithms on GeeksforGeeks/GeeksQuiz:
Selection Sort
Bubble Sort
Insertion Sort
Merge Sort
Heap Sort
QuickSort
Radix Sort
Counting Sort
ShellSort

ShellSort
ShellSort is mainly a variation of Insertion Sort. In insertion sort, we move elements only one position ahead. When an element has to be moved far
ahead, many movements are involved. The idea of shellSort is to allow exchange of far items. In shellSort, we make the array h-sorted for a large
value of h. We keep reducing the value of h until it becomes 1. An array is said to be h-sorted if all sublists of every hth element is sorted.
Following is C++ implementation of ShellSort.
#include <iostream>
using namespace std;
/* function to sort arr using shellSort */
int shellSort(int arr[], int n)
{
// Start with a big gap, then reduce the gap
for (int gap = n/2; gap > 0; gap /= 2)
{
// Do a gapped insertion sort for this gap size.
// The first gap elements a[0..gap-1] are already in gapped order
// keep adding one more element until the entire array is
// gap sorted
for (int i = gap; i < n; i += 1)
{
// add a[i] to the elements that have been gap sorted
// save a[i] in temp and make a hole at position i
int temp = arr[i];
// shift earlier gap-sorted elements up until the correct
// location for a[i] is found
int j;
for (j = i; j >= gap && arr[j - gap] > temp; j -= gap)
arr[j] = arr[j - gap];
// put temp (the original a[i]) in its correct location
arr[j] = temp;
}
}
return 0;
}
void printArray(int arr[], int n)
{
for (int i=0; i<n; i++)
cout << arr[i] << " ";
}
int main()
{
int arr[] = {12, 34, 54, 2, 3}, i;
int n = sizeof(arr)/sizeof(arr[0]);
cout << "Array before sorting: \n";
printArray(arr, n);
shellSort(arr, n);
cout << "\nArray after sorting: \n";
printArray(arr, n);
return 0;
}

Output:
Array before sorting:
12 34 54 2 3
Array after sorting:
2 3 12 34 54

Time Complexity: Time complexity of above implementation of shellsort is O(n2). In the above implementation gap is reduce by half in every
iteration. There are many other ways to reduce gap which lead to better time complexity. See this for more details.
References:
https://www.youtube.com/watch?v=pGhazjsFW28
http://en.wikipedia.org/wiki/Shellsort
Other Sorting Algorithms on GeeksforGeeks/GeeksQuiz:

Interpolation search vs Binary search

Interpolation search works better than Binary Search for a sorted and uniformly distributed array.
On average the interpolation search makes about log(log(n)) comparisons (if the elements are uniformly distributed), where n is the number of
elements to be searched. In the worst case (for instance where the numerical values of the keys increase exponentially) it can make up to O(n)
comparisons.
Sources:
http://en.wikipedia.org/wiki/Interpolation_search

Stability in sorting algorithms

A sorting algorithm is said to be stable if two objects with equal keys appear in the same order in sorted output as they appear in the input
unsorted array. Some sorting algorithms are stable by nature like Insertion sort, Merge Sort, Bubble Sort, etc. And some sorting algorithms are
not, like Heap Sort, Quick Sort, etc.
However, any given sorting algo which is not stable can be modified to be stable. There can be sorting algo specific ways to make it stable, but in
general, any comparison based sorting algorithm which is not stable by nature can be modified to be stable by changing the key comparison
operation so that the comparison of two keys considers position as a factor for objects with equal keys.
References:
http://www.math.uic.edu/~leon/cs-mcs401-s08/handouts/stability.pdf
http://en.wikipedia.org/wiki/Sorting_algorithm#Stability

When does the worst case of Quicksort occur?

The answer depends on strategy for choosing pivot. In early versions of Quick Sort where leftmost (or rightmost) element is chosen as pivot, the
worst occurs in following cases.
1) Array is already sorted in same order.
2) Array is already sorted in reverse order.
3) All elements are same (special case of case 1 and 2)
Since these cases are very common use cases, the problem was easily solved by choosing either a random index for the pivot, choosing the middle
index of the partition or (especially for longer partitions) choosing the median of the first, middle and last element of the partition for the pivot. With
these modifications, the worst case of Quick sort has less chances to occur, but worst case can still occur if the input array is such that the
maximum (or minimum) element is always chosen as pivot.
References:
http://en.wikipedia.org/wiki/Quicksort

Lower bound for comparison based sorting algorithms

The problem of sorting can be viewed as following.
Input: A sequence of n numbers <a1, a2, . . . , an>.
Output: A permutation (reordering) <a1, a2, . . . , an> of the input sequence such that a1 <= a2 .. <= an.
A sorting algorithm is comparison based if it uses comparison operators to find the order between two numbers. Comparison sorts can be viewed
abstractly in terms of decision trees. A decision tree is a full binary tree that represents the comparisons between elements that are performed by a
particular sorting algorithm operating on an input of a given size. The execution of the sorting algorithm corresponds to tracing a path from the root
of the decision tree to a leaf. At each internal node, a comparison ai aj is made. The left subtree then dictates subsequent comparisons for ai aj,
and the right subtree dictates subsequent comparisons for ai > aj. When we come to a leaf, the sorting algorithm has established the ordering. So
we can say following about the decison tree.
1) Each of the n! permutations on n elements must appear as one of the leaves of the decision tree for the sorting algorithm to sort properly.
2) Let x be the maximum number of comparisons in a sorting algorithm. The maximum height of the decison tree would be x. A tree with maximum
height x has at most 2^x leaves.
After combining the above two facts, we get following relation.
n! <= 2^x
Taking Log on both sides.
log2(n!) <= x
Since log2(n!) = ?(nLogn), we can say
x = ?(nLog2n)

Therefore, any comparison based sorting algorithm must make at least nLog2n comparisons to sort the input array, and Heapsort and merge sort
are asymptotically optimal comparison sorts.
References:
Introduction to Algorithms, by Thomas H. Cormen, Charles E. Leiserson, Ronald L. Rivest and Clifford Stein

Which sorting algorithm makes minimum number of memory writes?

Minimizing the number of writes is useful when making writes to some huge data set is very expensive, such as with EEPROMs or Flash memory,
where each write reduces the lifespan of the memory.
Among the sorting algorithms that we generally study in our data structure and algorithm courses, Selection Sort makes least number of writes (it
makes O(n) swaps). But, Cycle Sort almost always makes less number of writes compared to Selection Sort. In Cycle Sort, each value is either
written zero times, if its already in its correct position, or written one time to its correct position. This matches the minimal number of overwrites
required for a completed in-place sort.
Sources:
http://en.wikipedia.org/wiki/Cycle_sort
http://en.wikipedia.org/wiki/Selection_sort

Find the Minimum length Unsorted Subarray, sorting which makes the complete
array sorted
Given an unsorted array arr[0..n-1] of size n, find the minimum length subarray arr[s..e] such that sorting this subarray makes the whole array
sorted.
Examples:
1) If the input array is [10, 12, 20, 30, 25, 40, 32, 31, 35, 50, 60], your program should be able to find that the subarray lies between the indexes
3 and 8.
2) If the input array is [0, 1, 15, 25, 6, 7, 30, 40, 50], your program should be able to find that the subarray lies between the indexes 2 and 5.
Solution:
1) Find the candidate unsorted subarray
a) Scan from left to right and find the first element which is greater than the next element. Let s be the index of such an element. In the above
example 1, s is 3 (index of 30).
b) Scan from right to left and find the first element (first in right to left order) which is smaller than the next element (next in right to left order). Let e
be the index of such an element. In the above example 1, e is 7 (index of 31).
2) Check whether sorting the candidate unsorted subarray makes the complete array sorted or not. If not, then include more
elements in the subarray.
a) Find the minimum and maximum values in arr[s..e]. Let minimum and maximum values be min and max. min and max for [30, 25, 40, 32, 31]
are 25 and 40 respectively.
b) Find the first element (if there is any) in arr[0..s-1] which is greater than min, change s to index of this element. There is no such element in
above example 1.
c) Find the last element (if there is any) in arr[e+1..n-1] which is smaller than max, change e to index of this element. In the above example 1, e is
changed to 8 (index of 35)
3) Print s and e.

Implementation:
#include<stdio.h>
void printUnsorted(int arr[], int n)
{
int s = 0, e = n-1, i, max, min;
// step 1(a) of above algo
for (s = 0; s < n-1; s++)
{
if (arr[s] > arr[s+1])
break;
}
if (s == n-1)
{
printf("The complete array is sorted");
return;
}
// step 1(b) of above algo
for(e = n - 1; e > 0; e--)
{
if(arr[e] < arr[e-1])
break;
}
// step 2(a) of above algo
max = arr[s]; min = arr[s];
for(i = s + 1; i <= e; i++)
{
if(arr[i] > max)
max = arr[i];
if(arr[i] < min)
min = arr[i];
}
// step 2(b) of above algo
for( i = 0; i < s; i++)
{
if(arr[i] > min)

{
s = i;
break;
}
}
// step 2(c) of above algo
for( i = n -1; i >= e+1; i--)
{
if(arr[i] < max)
{
e = i;
break;
}
}
// step 3 of above algo
printf(" The unsorted subarray which makes the given array "
" sorted lies between the indees %d and %d", s, e);
return;
}
int main()
{
int arr[] = {10, 12, 20, 30, 25, 40, 32, 31, 35, 50, 60};
int arr_size = sizeof(arr)/sizeof(arr[0]);
printUnsorted(arr, arr_size);
getchar();
return 0;
}

Time Complexity: O(n)

Merge Sort for Linked Lists

Merge sort is often preferred for sorting a linked list. The slow random-access performance of a linked list makes some other algorithms (such as
quicksort) perform poorly, and others (such as heapsort) completely impossible.
Let head be the first node of the linked list to be sorted and headRef be the pointer to head. Note that we need a reference to head in MergeSort()
as the below implementation changes next links to sort the linked lists (not data at the nodes), so head node has to be changed if the data at original
head is not the smallest value in linked list.
MergeSort(headRef)
1) If head is NULL or there is only one element in the Linked List
then return.
2) Else divide the linked list into two halves.
FrontBackSplit(head, &a, &b); /* a and b are two halves */
3) Sort the two halves a and b.
MergeSort(a);
MergeSort(b);
4) Merge the sorted a and b (using SortedMerge() discussed here)
and update the head pointer using headRef.
*headRef = SortedMerge(a, b);
#include<stdio.h>
#include<stdlib.h>
/* Link list node */
struct node
{
int data;
struct node* next;
};
/* function prototypes */
struct node* SortedMerge(struct node* a, struct node* b);
void FrontBackSplit(struct node* source,
struct node** frontRef, struct node** backRef);
/* sorts the linked list by changing next pointers (not data) */
void MergeSort(struct node** headRef)
{
struct node* head = *headRef;
struct node* a;
struct node* b;
/* Base case -- length 0 or 1 */
if ((head == NULL) || (head->next == NULL))
{
return;
}
/* Split head into 'a' and 'b' sublists */
FrontBackSplit(head, &a, &b);
/* Recursively sort the sublists */
MergeSort(&a);
MergeSort(&b);
/* answer = merge the two sorted lists together */
*headRef = SortedMerge(a, b);
}
/* See http://geeksforgeeks.org/?p=3622 for details of this
function */
struct node* SortedMerge(struct node* a, struct node* b)
{
struct node* result = NULL;
/* Base cases */
if (a == NULL)
return(b);
else if (b==NULL)
return(a);
/* Pick either a or b, and recur */
if (a->data <= b->data)
{
result = a;
result->next = SortedMerge(a->next, b);
}
else

{
result = b;
result->next = SortedMerge(a, b->next);
}
return(result);
}
/* UTILITY FUNCTIONS */
/* Split the nodes of the given list into front and back halves,
and return the two lists using the reference parameters.
If the length is odd, the extra node should go in the front list.
Uses the fast/slow pointer strategy. */
void FrontBackSplit(struct node* source,
struct node** frontRef, struct node** backRef)
{
struct node* fast;
struct node* slow;
if (source==NULL || source->next==NULL)
{
/* length < 2 cases */
*frontRef = source;
*backRef = NULL;
}
else
{
slow = source;
fast = source->next;
/* Advance 'fast' two nodes, and advance 'slow' one node */
while (fast != NULL)
{
fast = fast->next;
if (fast != NULL)
{
slow = slow->next;
fast = fast->next;
}
}
/* 'slow' is before the midpoint in the list, so split it in two
at that point. */
*frontRef = source;
*backRef = slow->next;
slow->next = NULL;
}
}
/* Function to print nodes in a given linked list */
void printList(struct node *node)
{
while(node!=NULL)
{
printf("%d ", node->data);
node = node->next;
}
}
/* Function to insert a node at the beginging of the linked list */
void push(struct node** head_ref, int new_data)
{
/* allocate node */
struct node* new_node =
(struct node*) malloc(sizeof(struct node));
/* put in the data */
new_node->data = new_data;
/* link the old list off the new node */
new_node->next = (*head_ref);
/* move the head to point to the new node */
(*head_ref)
= new_node;
}
/* Drier program to test above functions*/
int main()
{
/* Start with the empty list */
struct node* res = NULL;
struct node* a = NULL;

/* Let us create a unsorted linked lists to test the functions

Created lists shall be a: 2->3->20->5->10->15 */
push(&a, 15);
push(&a, 10);
push(&a, 5);
push(&a, 20);
push(&a, 3);
push(&a, 2);
/* Sort the above created Linked List */
MergeSort(&a);
printf("\n Sorted Linked List is: \n");
printList(a);
getchar();
return 0;
}

Time Complexity: O(nLogn)

Sources:
http://en.wikipedia.org/wiki/Merge_sort
http://cslibrary.stanford.edu/105/LinkedListProblems.pdf

Sort a nearly sorted (or K sorted) array

Given an array of n elements, where each element is at most k away from its target position, devise an algorithm that sorts in O(n log k) time.
For example, let us consider k is 2, an element at index 7 in the sorted array, can be at indexes 5, 6, 7, 8, 9 in the given array.
Source: Nearly sorted algorithm
We can use Insertion Sort to sort the elements efficiently. Following is the C code for standard Insertion Sort.
/* Function to sort an array using insertion sort*/
void insertionSort(int A[], int size)
{
int i, key, j;
for (i = 1; i < size; i++)
{
key = A[i];
j = i-1;
/* Move elements of A[0..i-1], that are greater than key, to one
position ahead of their current position.
This loop will run at most k times */
while (j >= 0 && A[j] > key)
{
A[j+1] = A[j];
j = j-1;
}
A[j+1] = key;
}
}

The inner loop will run at most k times. To move every element to its correct place, at most k elements need to be moved. So overall complexity
will be O(nk)
We can sort such arrays more efficiently with the help of Heap data structure. Following is the detailed process that uses Heap.
1) Create a Min Heap of size k+1 with first k+1 elements. This will take O(k) time (See this GFact)
2) One by one remove min element from heap, put it in result array, and add a new element to heap from remaining elements.
Removing an element and adding a new element to min heap will take Logk time. So overall complexity will be O(k) + O((n-k)*logK)
#include<iostream>
using namespace std;
// Prototype of a utility function to swap two integers
void swap(int *x, int *y);
// A class for Min Heap
class MinHeap
{
int *harr; // pointer to array of elements in heap
int heap_size; // size of min heap
public:
// Constructor
MinHeap(int a[], int size);
// to heapify a subtree with root at given index
void MinHeapify(int );
// to get index of left child of node at index i
int left(int i) { return (2*i + 1); }
// to get index of right child of node at index i
int right(int i) { return (2*i + 2); }
// to remove min (or root), add a new value x, and return old root
int replaceMin(int x);
// to extract the root which is the minimum element
int extractMin();
};
// Given an array of size n, where every element is k away from its target
// position, sorts the array in O(nLogk) time.
int sortK(int arr[], int n, int k)
{
// Create a Min Heap of first (k+1) elements from
// input array
int *harr = new int[k+1];
for (int i = 0; i<=k && i<n; i++) // i < n condition is needed when k > n

harr[i] = arr[i];
MinHeap hp(harr, k+1);
// i is index for remaining elements in arr[] and ti
// is target index of for cuurent minimum element in
// Min Heapm 'hp'.
for(int i = k+1, ti = 0; ti < n; i++, ti++)
{
// If there are remaining elements, then place
// root of heap at target index and add arr[i]
// to Min Heap
if (i < n)
arr[ti] = hp.replaceMin(arr[i]);
// Otherwise place root at its target index and
// reduce heap size
else
arr[ti] = hp.extractMin();
}
}
// FOLLOWING ARE IMPLEMENTATIONS OF STANDARD MIN HEAP METHODS FROM CORMEN BOOK
// Constructor: Builds a heap from a given array a[] of given size
MinHeap::MinHeap(int a[], int size)
{
heap_size = size;
harr = a; // store address of array
int i = (heap_size - 1)/2;
while (i >= 0)
{
MinHeapify(i);
i--;
}
}
// Method to remove minimum element (or root) from min heap
int MinHeap::extractMin()
{
int root = harr[0];
if (heap_size > 1)
{
harr[0] = harr[heap_size-1];
heap_size--;
MinHeapify(0);
}
return root;
}
// Method to change root with given value x, and return the old root
int MinHeap::replaceMin(int x)
{
int root = harr[0];
harr[0] = x;
if (root < x)
MinHeapify(0);
return root;
}
// A recursive method to heapify a subtree with root at given index
// This method assumes that the subtrees are already heapified
void MinHeap::MinHeapify(int i)
{
int l = left(i);
int r = right(i);
int smallest = i;
if (l < heap_size && harr[l] < harr[i])
smallest = l;
if (r < heap_size && harr[r] < harr[smallest])
smallest = r;
if (smallest != i)
{
swap(&harr[i], &harr[smallest]);
MinHeapify(smallest);
}
}
// A utility function to swap two elements
void swap(int *x, int *y)
{
int temp = *x;
*x = *y;

*y = temp;
}
// A utility function to print array elements
void printArray(int arr[], int size)
{
for (int i=0; i < size; i++)
cout << arr[i] << " ";
cout << endl;
}
// Driver program to test above functions
int main()
{
int k = 3;
int arr[] = {2, 6, 3, 12, 56, 8};
int n = sizeof(arr)/sizeof(arr[0]);
sortK(arr, n, k);
cout << "Following is sorted array\n";
printArray (arr, n);
return 0;
}

Output:
Following is sorted array
2 3 6 8 12 56

The Min Heap based method takes O(nLogk) time and uses O(k) auxiliary space.
We can also use a Balanced Binary Search Tree instead of Heap to store K+1 elements. The insert and delete operations on Balanced BST
also take O(Logk) time. So Balanced BST based method will also take O(nLogk) time, but the Heap bassed method seems to be more efficient
as the minimum element will always be at root. Also, Heap doesnt need extra space for left and right pointers.

Iterative Quick Sort

Following is a typical recursive implementation of Quick Sort that uses last element as pivot.
/* A typical recursive implementation of quick sort */
/* This function takes last element as pivot, places the pivot element at its
correct position in sorted array, and places all smaller (smaller than pivot)
to left of pivot and all greater elements to right of pivot */
int partition (int arr[], int l, int h)
{
int x = arr[h];
int i = (l - 1);
for (int j = l; j <= h- 1; j++)
{
if (arr[j] <= x)
{
i++;
swap (&arr[i], &arr[j]);
}
}
swap (&arr[i + 1], &arr[h]);
return (i + 1);
}
/* A[] --> Array to be sorted, l --> Starting index, h --> Ending index */
void quickSort(int A[], int l, int h)
{
if (l < h)
{
int p = partition(A, l, h); /* Partitioning index */
quickSort(A, l, p - 1);
quickSort(A, p + 1, h);
}
}

The above implementation can be optimized in many ways

1) The above implementation uses last index as pivot. This causes worst-case behavior on already sorted arrays, which is a commonly occurring
case. The problem can be solved by choosing either a random index for the pivot, or choosing the middle index of the partition or choosing the
median of the first, middle and last element of the partition for the pivot. (See this for details)
2) To reduce the recursion depth, recur first for the smaller half of the array, and use a tail call to recurse into the other.
3) Insertion sort works better for small subarrays. Insertion sort can be used for invocations on such small arrays (i.e. where the length is less than
a threshold t determined experimentally). For example, this library implementation of qsort uses insertion sort below size 7.
Despite above optimizations, the function remains recursive and uses function call stack to store intermediate values of l and h. The function call
stack stores other bookkeeping information together with parameters. Also, function calls involve overheads like storing activation record of the
caller function and then resuming execution.
The above function can be easily converted to iterative version with the help of an auxiliary stack. Following is an iterative implementation of the
above recursive code.
// An iterative implementation of quick sort
#include <stdio.h>
// A utility function to swap two elements
void swap ( int* a, int* b )
{
int t = *a;
*a = *b;
*b = t;
}
/* This function is same in both iterative and recursive*/
int partition (int arr[], int l, int h)
{
int x = arr[h];
int i = (l - 1);
for (int j = l; j <= h- 1; j++)
{
if (arr[j] <= x)
{
i++;

swap (&arr[i], &arr[j]);

}
}
swap (&arr[i + 1], &arr[h]);
return (i + 1);
}
/* A[] --> Array to be sorted, l --> Starting index, h --> Ending index */
void quickSortIterative (int arr[], int l, int h)
{
// Create an auxiliary stack
int stack[ h - l + 1 ];
// initialize top of stack
int top = -1;
// push initial values of l and h to stack
stack[ ++top ] = l;
stack[ ++top ] = h;
// Keep popping from stack while is not empty
while ( top >= 0 )
{
// Pop h and l
h = stack[ top-- ];
l = stack[ top-- ];
// Set pivot element at its correct position in sorted array
int p = partition( arr, l, h );
// If there are elements on left side of pivot, then push left
// side to stack
if ( p-1 > l )
{
stack[ ++top ] = l;
stack[ ++top ] = p - 1;
}
// If there are elements on right side of pivot, then push right
// side to stack
if ( p+1 < h )
{
stack[ ++top ] = p + 1;
stack[ ++top ] = h;
}
}
}
// A utility function to print contents of arr
void printArr( int arr[], int n )
{
int i;
for ( i = 0; i < n; ++i )
printf( "%d ", arr[i] );
}
// Driver program to test above functions
int main()
{
int arr[] = {4, 3, 5, 2, 1, 3, 2, 3};
int n = sizeof( arr ) / sizeof( *arr );
quickSortIterative( arr, 0, n - 1 );
printArr( arr, n );
return 0;
}

Output:
1 2 2 3 3 3 4 5

The above mentioned optimizations for recursive quick sort can also be applied to iterative version.
1) Partition process is same in both recursive and iterative. The same techniques to choose optimal pivot can also be applied to iterative version.
2) To reduce the stack size, first push the indexes of smaller half.
3) Use insertion sort when the size reduces below a experimentally calculated threshold.
References:
http://en.wikipedia.org/wiki/Quicksort

QuickSort on Singly Linked List

QuickSort on Doubly Linked List is discussed here. QuickSort on Singly linked list was given as an exercise. Following is C++ implementation for
same. The important things about implementation are, it changes pointers rather swapping data and time complexity is same as the implementation
for Doubly Linked List.
In partition(), we consider last element as pivot. We traverse through the current list and if a node has value greater than pivot, we move it after
tail. If the node has smaller value, we keep it at its current position.
In QuickSortRecur(), we first call partition() which places pivot at correct position and returns pivot. After pivot is placed at correct position, we
find tail node of left side (list before pivot) and recur for left list. Finally, we recur for right list.
// C++ program for Quick Sort on Singly Linled List
#include <iostream>
#include <cstdio>
using namespace std;
/* a node of the singly linked list */
struct node
{
int data;
struct node *next;
};
/* A utility function to insert a node at the beginning of linked list */
void push(struct node** head_ref, int new_data)
{
/* allocate node */
struct node* new_node = new node;
/* put in the data */
new_node->data = new_data;
/* link the old list off the new node */
new_node->next = (*head_ref);
/* move the head to point to the new node */
(*head_ref)
= new_node;
}
/* A utility function to print linked list */
void printList(struct node *node)
{
while (node != NULL)
{
printf("%d ", node->data);
node = node->next;
}
printf("\n");
}
// Returns the last node of the list
struct node *getTail(struct node *cur)
{
while (cur != NULL && cur->next != NULL)
cur = cur->next;
return cur;
}
// Partitions the list taking
struct node *partition(struct
struct
{
struct node *pivot = end;
struct node *prev = NULL,

the last element as the pivot

node *head, struct node *end,
node **newHead, struct node **newEnd)
*cur = head, *tail = pivot;

// During partition, both the head and end of the list might change
// which is updated in the newHead and newEnd variables
while (cur != pivot)
{
if (cur->data < pivot->data)
{
// First node that has a value less than the pivot - becomes
// the new head
if ((*newHead) == NULL)
(*newHead) = cur;
prev = cur;
cur = cur->next;
}
else // If cur node is greater than pivot

{
// Move cur node to next of tail, and change tail
if (prev)
prev->next = cur->next;
struct node *tmp = cur->next;
cur->next = NULL;
tail->next = cur;
tail = cur;
cur = tmp;
}
}
// If the pivot data is the smallest element in the current list,
// pivot becomes the head
if ((*newHead) == NULL)
(*newHead) = pivot;
// Update newEnd to the current last node
(*newEnd) = tail;
// Return the pivot node
return pivot;
}
//here
struct
{
//
if

the sorting happens exclusive of the end node

node *quickSortRecur(struct node *head, struct node *end)
base condition
(!head || head == end)
return head;

node *newHead = NULL, *newEnd = NULL;

// Partition the list, newHead and newEnd will be updated
// by the partition function
struct node *pivot = partition(head, end, &newHead, &newEnd);
// If pivot is the smallest element - no need to recur for
// the left part.
if (newHead != pivot)
{
// Set the node before the pivot node as NULL
struct node *tmp = newHead;
while (tmp->next != pivot)
tmp = tmp->next;
tmp->next = NULL;
// Recur for the list before pivot
newHead = quickSortRecur(newHead, tmp);
// Change next of last node of the left half to pivot
tmp = getTail(newHead);
tmp->next = pivot;
}
// Recur for the list after the pivot element
pivot->next = quickSortRecur(pivot->next, newEnd);
return newHead;
}
// The main function for quick sort. This is a wrapper over recursive
// function quickSortRecur()
void quickSort(struct node **headRef)
{
(*headRef) = quickSortRecur(*headRef, getTail(*headRef));
return;
}
// Driver program to test above functions
int main()
{
struct node *a = NULL;
push(&a, 5);
push(&a, 20);
push(&a, 4);
push(&a, 3);
push(&a, 30);
cout << "Linked List before sorting \n";

printList(a);
quickSort(&a);
cout << "Linked List after sorting \n";
printList(a);
return 0;
}

Output:
Linked List before sorting
30 3 4 20 5
Linked List after sorting
3 4 5 20 30

QuickSort on Doubly Linked List

Following is a typical recursive implementation of QuickSort for arrays. The implementation uses last element as pivot.
/* A typical recursive implementation of Quicksort for array*/
/* This function takes last element as pivot, places the pivot element at its
correct position in sorted array, and places all smaller (smaller than
pivot) to left of pivot and all greater elements to right of pivot */
int partition (int arr[], int l, int h)
{
int x = arr[h];
int i = (l - 1);
for (int j = l; j <= h- 1; j++)
{
if (arr[j] <= x)
{
i++;
swap (&arr[i], &arr[j]);
}
}
swap (&arr[i + 1], &arr[h]);
return (i + 1);
}
/* A[] --> Array to be sorted, l --> Starting index, h --> Ending index */
void quickSort(int A[], int l, int h)
{
if (l < h)
{
int p = partition(A, l, h); /* Partitioning index */
quickSort(A, l, p - 1);
quickSort(A, p + 1, h);
}
}

Can we use same algorithm for Linked List?

Following is C++ implementation for doubly linked list. The idea is simple, we first find out pointer to last node. Once we have pointer to last node,
we can recursively sort the linked list using pointers to first and last nodes of linked list, similar to the above recursive function where we pass
indexes of first and last array elements. The partition function for linked list is also similar to partition for arrays. Instead of returning index of the
pivot element, it returns pointer to the pivot element. In the following implementation, quickSort() is just a wrapper function, the main recursive
function is _quickSort() which is similar to quickSort() for array implementation.

// A C++ program to sort a linked list using Quicksort

#include <iostream>
#include <stdio.h>
using namespace std;
/* a node of the doubly linked list */
struct node
{
int data;
struct node *next;
struct node *prev;
};
/* A utility function to swap two elements */
void swap ( int* a, int* b )
{ int t = *a;
*a = *b;
*b = t; }
// A utility function to find last node of linked list
struct node *lastNode(node *root)
{
while (root && root->next)
root = root->next;
return root;
}
/* Considers last element as pivot, places the pivot element at its
correct position in sorted array, and places all smaller (smaller than
pivot) to left of pivot and all greater elements to right of pivot */

node* partition(node *l, node *h)

{
// set pivot as h element
int x = h->data;
// similar to i = l-1 for array implementation
node *i = l->prev;
// Similar to "for (int j = l; j <= h- 1; j++)"
for (node *j = l; j != h; j = j->next)
{
if (j->data <= x)
{
// Similar to i++ for array
i = (i == NULL)? l : i->next;
swap(&(i->data), &(j->data));
}
}
i = (i == NULL)? l : i->next; // Similar to i++
swap(&(i->data), &(h->data));
return i;
}
/* A recursive implementation of quicksort for linked list */
void _quickSort(struct node* l, struct node *h)
{
if (h != NULL && l != h && l != h->next)
{
struct node *p = partition(l, h);
_quickSort(l, p->prev);
_quickSort(p->next, h);
}
}
// The main function to sort a linked list. It mainly calls _quickSort()
void quickSort(struct node *head)
{
// Find last node
struct node *h = lastNode(head);
// Call the recursive QuickSort
_quickSort(head, h);
}
// A utility function to print contents of arr
void printList(struct node *head)
{
while (head)
{
cout << head->data << " ";
head = head->next;
}
cout << endl;
}
/* Function to insert a node at the beginging of the Doubly Linked List */
void push(struct node** head_ref, int new_data)
{
struct node* new_node = new node;
/* allocate node */
new_node->data = new_data;
/* since we are adding at the begining, prev is always NULL */
new_node->prev = NULL;
/* link the old list off the new node */
new_node->next = (*head_ref);
/* change prev of head node to new node */
if ((*head_ref) != NULL) (*head_ref)->prev = new_node ;
/* move the head to point to the new node */
(*head_ref)
= new_node;
}
/* Driver program to test above function */
int main()
{
struct node *a = NULL;
push(&a, 5);
push(&a, 20);

push(&a, 4);
push(&a, 3);
push(&a, 30);
cout << "Linked List before sorting \n";
printList(a);
quickSort(a);
cout << "Linked List after sorting \n";
printList(a);
return 0;
}

Output :
Linked List before sorting
30 3 4 20 5
Linked List after sorting
3 4 5 20 30

Time Complexity: Time complexity of the above implementation is same as time complexity of QuickSort() for arrays. It takes O(n^2) time in
worst case and O(nLogn) in average and best cases. The worst case occurs when the linked list is already sorted.
Can we implement random quick sort for linked list?
Quicksort can be implemented for Linked List only when we can pick a fixed point as pivot (like last element in above implementation). Random
QuickSort cannot be efficiently implemented for Linked Lists by picking random pivot.
Exercise:
The above implementation is for doubly linked list. Modify it for singly linked list. Note that we dont have prev pointer in singly linked list.
Refer QuickSort on Singly Linked List for solution.

Find k closest elements to a given value

Given a sorted array arr[] and a value X, find the k closest elements to X in arr[].
Examples:
Input: K = 4, X = 35
arr[] = {12, 16, 22, 30, 35, 39, 42,
45, 48, 50, 53, 55, 56}
Output: 30 39 42 45

Note that if the element is present in array, then it should not be in output, only the other closest elements are required.
In the following solutions, it is assumed that all elements of array are distinct.
A simple solution is to do linear search for k closest elements.
1) Start from the first element and search for the crossover point (The point before which elements are smaller than or equal to X and after which
elements are greater). This step takes O(n) time.
2) Once we find the crossover point, we can compare elements on both sides of crossover point to print k closest elements. This step takes O(k)
time.
The time complexity of the above solution is O(n).
An Optimized Solution is to find k elements in O(Logn + k) time. The idea is to use Binary Search to find the crossover point. Once we find
index of crossover point, we can print k closest elements in O(k) time.
#include<stdio.h>
/* Function to find the cross over point (the point before
which elements are smaller than or equal to x and after
which greater than x)*/
int findCrossOver(int arr[], int low, int high, int x)
{
// Base cases
if (arr[high] <= x) // x is greater than all
return high;
if (arr[low] > x) // x is smaller than all
return low;
// Find the middle point
int mid = (low + high)/2; /* low + (high - low)/2 */
/* If x is same as middle element, then return mid */
if (arr[mid] <= x && arr[mid+1] > x)
return mid;
/* If x is greater than arr[mid], then either arr[mid + 1]
is ceiling of x or ceiling lies in arr[mid+1...high] */
if(arr[mid] < x)
return findCrossOver(arr, mid+1, high, x);
return findCrossOver(arr, low, mid - 1, x);
}
// This function prints k closest elements to x in arr[].
// n is the number of elements in arr[]
void printKclosest(int arr[], int x, int k, int n)
{
// Find the crossover point
int l = findCrossOver(arr, 0, n-1, x); // le
int r = l+1; // Right index to search
int count = 0; // To keep track of count of elements already printed
// If x is present in arr[], then reduce left index
// Assumption: all elements in arr[] are distinct
if (arr[l] == x) l--;
// Compare elements on left and right of crossover
// point to find the k closest elements
while (l >= 0 && r < n && count < k)
{
if (x - arr[l] < arr[r] - x)
printf("%d ", arr[l--]);
else
printf("%d ", arr[r++]);
count++;
}

// If there are no more elements on right side, then

// print left elements
while (count < k && l >= 0)
printf("%d ", arr[l--]), count++;
// If there are no more elements on left side, then
// print right elements
while (count < k && r < n)
printf("%d ", arr[r++]), count++;
}
/* Driver program to check above functions */
int main()
{
int arr[] ={12, 16, 22, 30, 35, 39, 42,
45, 48, 50, 53, 55, 56};
int n = sizeof(arr)/sizeof(arr[0]);
int x = 35, k = 4;
printKclosest(arr, x, 4, n);
return 0;
}

Output:
39 30 42 45

The time complexity of this method is O(Logn + k).

Exercise: Extend the optimized solution to work for duplicates also, i.e., to work for arrays where elements dont have to be distinct.

Sort n numbers in range from 0 to n^2 1 in linear time

Given an array of numbers of size n. It is also given that the array elements are in range from 0 to n2 1. Sort the given array in linear time.
Examples:
Since there are 5 elements, the elements can be from 0 to 24.
Input: arr[] = {0, 23, 14, 12, 9}
Output: arr[] = {0, 9, 12, 14, 23}
Since there are 3 elements, the elements can be from 0 to 8.
Input: arr[] = {7, 0, 2}
Output: arr[] = {0, 2, 7}

Solution: If we use Counting Sort, it would take O(n^2) time as the given range is of size n^2. Using any comparison based sorting like Merge
Sort, Heap Sort, .. etc would take O(nLogn) time.
Now question arises how to do this in 0(n)? Firstly, is it possible? Can we use data given in question? n numbers in range from 0 to n2 1?
The idea is to use Radix Sort. Following is standard Radix Sort algorithm.
1) Do following for each digit i where i varies from least
significant digit to the most significant digit.
..a) Sort input array using counting sort (or any stable
sort) according to the ith digit

Let there be d digits in input integers. Radix Sort takes O(d*(n+b)) time where b is the base for representing numbers, for example, for decimal
system, b is 10. Since n2-1 is the maximum possible value, the value of d would be O(logb(n)). So overall time complexity is O((n+b)*O(logb(n)).
Which looks more than the time complexity of comparison based sorting algorithms for a large k. The idea is to change base b. If we set b as n,
the value of O(logb(n)) becomes O(1) and overall time complexity becomes O(n).
arr[] = {0, 10, 13, 12, 7}
Let us consider the elements in base 5. For example 13 in
base 5 is 23, and 7 in base 5 is 12.
arr[] = {00(0), 20(10), 23(13), 22(12), 12(7)}
After first iteration (Sorting according to the last digit in
base 5), we get.
arr[] = {00(0), 20(10), 12(7), 22(12), 23(13)}
After second iteration, we get
arr[] = {00(0), 12(7), 20(10), 22(12), 23(13)}

Following is C++ implementation to sort an array of size n where elements are in range from 0 to n2 1.
#include<iostream>
using namespace std;
// A function to do counting sort of arr[] according to
// the digit represented by exp.
int countSort(int arr[], int n, int exp)
{
int output[n]; // output array
int i, count[n] ;
for (int i=0; i < n; i++)
count[i] = 0;
// Store count of occurrences in count[]
for (i = 0; i < n; i++)
count[ (arr[i]/exp)%n ]++;
// Change count[i] so that count[i] now contains actual
// position of this digit in output[]
for (i = 1; i < n; i++)
count[i] += count[i - 1];
// Build the output array
for (i = n - 1; i >= 0; i--)
{
output[count[ (arr[i]/exp)%n] - 1] = arr[i];
count[(arr[i]/exp)%n]--;
}
// Copy the output array to arr[], so that arr[] now
// contains sorted numbers according to curent digit
for (i = 0; i < n; i++)
arr[i] = output[i];
}

// The main function to that sorts arr[] of size n using Radix Sort
void sort(int arr[], int n)
{
// Do counting sort for first digit in base n. Note that
// instead of passing digit number, exp (n^0 = 0) is passed.
countSort(arr, n, 1);
// Do counting sort for second digit in base n. Note that
// instead of passing digit number, exp (n^1 = n) is passed.
countSort(arr, n, n);
}
// A utility function to print an array
void printArr(int arr[], int n)
{
for (int i = 0; i < n; i++)
cout << arr[i] << " ";
}
// Driver program to test above functions
int main()
{
// Since array size is 7, elements should be from 0 to 48
int arr[] = {40, 12, 45, 32, 33, 1, 22};
int n = sizeof(arr)/sizeof(arr[0]);
cout << "Given array is \n";
printArr(arr, n);
sort(arr, n);
cout << "\nSorted array is \n";
printArr(arr, n);
return 0;
}

Output:
Given array is
40 12 45 32 33 1 22
Sorted array is
1 12 22 32 33 40 45

How to sort if range is from 1 to n2?

If range is from 1 to n n2, the above process can not be directly applied, it must be changed. Consider n = 100 and range from 1 to 10000. Since
the base is 100, a digit must be from 0 to 99 and there should be 2 digits in the numbers. But the number 10000 has more than 2 digits. So to sort
numbers in a range from 1 to n2, we can use following process.
1) Subtract all numbers by 1.
2) Since the range is now 0 to n2, do counting sort twice as done in the above implementation.
3) After the elements are sorted, add 1 to all numbers to obtain the original numbers.
How to sort if range is from 0 to n^3 -1?
Since there can be 3 digits in base n, we need to call counting sort 3 times.

A Problem in Many Binary Search Implementations

Consider the following C implementation of Binary Search function, is there anything wrong in this?
// A iterative binary search function. It returns location of x in
// given array arr[l..r] if present, otherwise -1
int binarySearch(int arr[], int l, int r, int x)
{
while (l <= r)
{
// find index of middle element
int m = (l+r)/2;
// Check if x is present at mid
if (arr[m] == x) return m;
// If x greater, ignore left half
if (arr[m] < x) l = m + 1;
// If x is smaller, ignore right half
else r = m - 1;
}
// if we reach here, then element was not present
return -1;
}

The above looks fine except one subtle thing, the expression m = (l+r)/2?. It fails for large values of l and r. Specifically, it fails if the sum of low
and high is greater than the maximum positive int value (231 1). The sum overflows to a negative value, and the value stays negative when divided
by two. In C this causes an array index out of bounds with unpredictable results.
What is the way to resolve this problem?
Following is one way:
int mid = low + ((high - low) / 2);

Probably faster, and arguably as clear is (works only in Java, refer this):
int mid = (low + high) >>> 1;

In C and C++ (where you dont have the >>> operator), you can do this:
mid = ((unsigned int)low + (unsigned int)high)) >> 1

The similar problem appears in Merge Sort as well.

The above content is taken from google reasearch blog.
Please refer this as well, it points out that the above solutions may not always work.
The above problem occurs when array length is 230 or greater and the search repeatedly moves to second half of the array. This much size of array
is not likely to appear most of the time. For example, when we try the below program with 32 bit Code Blocks compiler, we get compiler error.
int main()
{
int arr[1<<30];
return 0;
}

Output:
error: size of array 'arr' is too large

Even when we try boolean array, the program compiles fine, but crashes when run in Windows 7.0 and Code Blocks 32 bit compiler
#include <stdbool.h>
int main()
{
bool arr[1<<30];
return 0;
}

Output: No compiler error, but crashes at run time.

Sources:
http://googleresearch.blogspot.in/2006/06/extra-extra-read-all-about-it-nearly.html

http://locklessinc.com/articles/binary_search/

Search in an almost sorted array

Given an array which is sorted, but after sorting some elements are moved to either of the adjacent positions, i.e., arr[i] may be present at arr[i+1]
or arr[i-1]. Write an efficient function to search an element in this array. Basically the element arr[i] can only be swapped with either arr[i+1] or
arr[i-1].
For example consider the array {2, 3, 10, 4, 40}, 4 is moved to next position and 10 is moved to previous position.
Example:
Input: arr[] = {10, 3, 40, 20, 50, 80, 70}, key = 40
Output: 2
Output is index of 40 in given array
Input: arr[] = {10, 3, 40, 20, 50, 80, 70}, key = 90
Output: -1
-1 is returned to indicate element is not present

A simple solution is to linearly search the given key in given array. Time complexity of this solution is O(n). We cab modify binary search to do it in
O(Logn) time.
The idea is to compare the key with middle 3 elements, if present then return the index. If not present, then compare the key with middle element
to decide whether to go in left half or right half. Comparing with middle element is enough as all the elements after mid+2 must be greater than
element mid and all elements before mid-2 must be smaller than mid element.
Following is C++ implementation of this approach.
// C++ program to find an element in an almost sorted array
#include <stdio.h>
// A recursive binary search based function. It returns index of x in
// given array arr[l..r] is present, otherwise -1
int binarySearch(int arr[], int l, int r, int x)
{
if (r >= l)
{
int mid = l + (r - l)/2;
//
if
if
if

If the element is present at one of the middle 3 positions

(arr[mid] == x) return mid;
(mid > l && arr[mid-1] == x) return (mid - 1);
(mid < r && arr[mid+1] == x) return (mid + 1);

// If element is smaller than mid, then it can only be present

// in left subarray
if (arr[mid] > x) return binarySearch(arr, l, mid-2, x);
// Else the element can only be present in right subarray
return binarySearch(arr, mid+2, r, x);
}
// We reach here when element is not present in array
return -1;
}
// Driver program to test above function
int main(void)
{
int arr[] = {3, 2, 10, 4, 40};
int n = sizeof(arr)/ sizeof(arr[0]);
int x = 4;
int result = binarySearch(arr, 0, n-1, x);
(result == -1)? printf("Element is not present in array")
: printf("Element is present at index %d", result);
return 0;
}

Output:
Element is present at index 3

Time complexity of the above function is O(Logn).

Sort an array in wave form

Given an unsorted array of integers, sort the array into a wave like array. An array arr[0..n-1] is sorted in wave form if arr[0] >= arr[1] <= arr[2]
>= arr[3] <= arr[4] >= ..
Examples:
Input: arr[] = {10, 5, 6, 3, 2, 20, 100, 80}
Output: arr[] = {10, 5, 6, 2, 20, 3, 100, 80} OR
{20, 5, 10, 2, 80, 6, 100, 3} OR
any other array that is in wave form
Input: arr[] = {20, 10, 8, 6, 4, 2}
Output: arr[] = {20, 8, 10, 4, 6, 2} OR
{10, 8, 20, 2, 6, 4} OR
any other array that is in wave form
Input: arr[] = {2, 4, 6, 8, 10, 20}
Output: arr[] = {4, 2, 8, 6, 20, 10} OR
any other array that is in wave form
Input: arr[] = {3, 6, 5, 10, 7, 20}
Output: arr[] = {6, 3, 10, 5, 20, 7} OR
any other array that is in wave form

A Simple Solution is to use sorting. First sort the input array, then swap all adjacent elements.
For example, let the input array be {3, 6, 5, 10, 7, 20}. After sorting, we get {3, 5, 6, 7, 10, 20}. After swapping adjacent elements, we get {5,
3, 7, 6, 20, 10}.
Below are implementations of this simple approach.

C++
// A C++ program to sort an array in wave form using a sorting function
#include<iostream>
#include<algorithm>
using namespace std;
// A utility method to swap two numbers.
void swap(int *x, int *y)
{
int temp = *x;
*x = *y;
*y = temp;
}
// This function sorts arr[0..n-1] in wave form, i.e.,
// arr[0] >= arr[1] <= arr[2] >= arr[3] <= arr[4] >= arr[5]..
void sortInWave(int arr[], int n)
{
// Sort the input array
sort(arr, arr+n);
// Swap adjacent elements
for (int i=0; i<n-1; i += 2)
swap(&arr[i], &arr[i+1]);
}
// Driver program to test above function
int main()
{
int arr[] = {10, 90, 49, 2, 1, 5, 23};
int n = sizeof(arr)/sizeof(arr[0]);
sortInWave(arr, n);
for (int i=0; i<n; i++)
cout << arr[i] << " ";
return 0;
}

Python
# Python function to sort the array arr[0..n-1] in wave form,
# i.e., arr[0] >= arr[1] <= arr[2] >= arr[3] <= arr[4] >= arr[5]
def sortInWave(arr, n):
#sort the array

arr.sort()
# Swap adjacent elements
for i in range(0,n-1,2):
arr[i], arr[i+1] = arr[i+1], arr[i]
# Driver progrM
arr = [10, 90, 49, 2, 1, 5, 23]
sortInWave(arr, len(arr))
for i in range(0,len(arr)):
print arr[i],
# This code is contributed by __Devesh Agrawal__

2 1 10 5 49 23 90

The time complexity of the above solution is O(nLogn) if a O(nLogn) sorting algorithm like Merge Sort, Heap Sort, .. etc is used.
This can be done in O(n) time by doing a single traversal of given array. The idea is based on the fact that if we make sure that all even
positioned (at index 0, 2, 4, ..) elements are greater than their adjacent odd elements, we dont need to worry about odd positioned element.
Following are simple steps.
1) Traverse all even positioned elements of input array, and do following.
.a) If current element is smaller than previous odd element, swap previous and current.
.b) If current element is smaller than next odd element, swap next and current.
Below are implementations of above simple algorithm.

C++
// A O(n) program to sort an input array in wave form
#include<iostream>
using namespace std;
// A utility method to swap two numbers.
void swap(int *x, int *y)
{
int temp = *x;
*x = *y;
*y = temp;
}
// This function sorts arr[0..n-1] in wave form, i.e., arr[0] >=
// arr[1] <= arr[2] >= arr[3] <= arr[4] >= arr[5] ....
void sortInWave(int arr[], int n)
{
// Traverse all even elements
for (int i = 0; i < n; i+=2)
{
// If current even element is smaller than previous
if (i>0 && arr[i-1] > arr[i] )
swap(&arr[i], &arr[i-1]);
// If current even element is smaller than next
if (i<n-1 && arr[i] < arr[i+1] )
swap(&arr[i], &arr[i + 1]);
}
}
// Driver program to test above function
int main()
{
int arr[] = {10, 90, 49, 2, 1, 5, 23};
int n = sizeof(arr)/sizeof(arr[0]);
sortInWave(arr, n);
for (int i=0; i<n; i++)
cout << arr[i] << " ";
return 0;
}

Python
# Python function to sort the array arr[0..n-1] in wave form,
# i.e., arr[0] >= arr[1] <= arr[2] >= arr[3] <= arr[4] >= arr[5]
def sortInWave(arr, n):

# Traverse all even elements

for i in range(0, n, 2):
# If current even element is smaller than previous
if (i> 0 and arr[i] < arr[i-1]):
arr[i],arr[i-1]=arr[i-1],arr[i]
# If current even element is smaller than next
if (i < n-1 and arr[i] < arr[i+1]):
arr[i],arr[i+1]=arr[i+1],arr[i]
# Driver program
arr = [10, 90, 49, 2, 1, 5, 23]
sortInWave(arr, len(arr))
for i in range(0,len(arr)):
print arr[i],
# This code is contributed by __Devesh Agrawal__

Output:
90 10 49 1 5 2 23

Why is Binary Search preferred over Ternary Search?

The following is a simple recursive Binary Search function in C++ taken from here.
// A recursive binary search function. It returns location of x in
// given array arr[l..r] is present, otherwise -1
int binarySearch(int arr[], int l, int r, int x)
{
if (r >= l)
{
int mid = l + (r - l)/2;
// If the element is present at the middle itself
if (arr[mid] == x) return mid;
// If element is smaller than mid, then it can only be present
// in left subarray
if (arr[mid] > x) return binarySearch(arr, l, mid-1, x);
// Else the element can only be present in right subarray
return binarySearch(arr, mid+1, r, x);
}
// We reach here when element is not present in array
return -1;
}

The following is a simple recursive Ternary Search function in C++.

// A recursive ternary search function. It returns location of x in
// given array arr[l..r] is present, otherwise -1
int ternarySearch(int arr[], int l, int r, int x)
{
if (r >= l)
{
int mid1 = l + (r - l)/3;
int mid2 = mid1 + (r - l)/3;
// If x is present at the mid1
if (arr[mid1] == x) return mid1;
// If x is present at the mid2
if (arr[mid2] == x) return mid2;
// If x is present in left one-third
if (arr[mid1] > x) return ternarySearch(arr, l, mid1-1, x);
// If x is present in right one-third
if (arr[mid2] < x) return ternarySearch(arr, mid2+1, r, x);
// If x is present in middle one-third
return ternarySearch(arr, mid1+1, mid2-1, x);
}
// We reach here when element is not present in array
return -1;
}

Which of the above two does less comparisons in worst case?

From the first look, it seems the ternary search does less number of comparisons as it makes Log3n recursive calls, but binary search makes Log2n
recursive calls. Let us take a closer look.
The following is recursive formula for counting comparisons in worst case of Binary Search.
T(n) = T(n/2) + 2, T(1) = 1

The following is recursive formula for counting comparisons in worst case of Ternary Search.
T(n) = T(n/3) + 4, T(1) = 1

In binary search, there are 2Log2n + 1 comparisons in worst case. In ternary search, there are 4Log3n + 1 comparisons in worst case.
Therefore, the comparison of Ternary and Binary Searches boils down the comparison of expressions 2Log3n and Log2n . The value of 2Log3n
can be written as (2 / Log23) * Log2n . Since the value of (2 / Log23) is more than one, Ternary Search does more comparisons than Binary
Search in worst case.
Exercise:
Why Merge Sort divides input array in two halves, why not in three or more parts?

Kth Smallest/Largest Element in Unsorted Array | Set 2 (Expected Linear Time)

We recommend to read following post as a prerequisite of this post.
Kth Smallest/Largest Element in Unsorted Array | Set 1
Given an array and a number k where k is smaller than size of array, we need to find the kth smallest element in the given array. It is given that ll
array elements are distinct.
Examples:
Input: arr[] = {7, 10, 4, 3, 20, 15}
k = 3
Output: 7
Input: arr[] = {7, 10, 4, 3, 20, 15}
k = 4
Output: 10

We have discussed three different solutions here.

In this post method 4 is discussed which is mainly an extension of method 3 (QuickSelect) discussed in the previous post. The idea is to randomly
pick a pivot element. To implement randomized partition, we use a random function, rand() to generate index between l and r, swap the element at
randomly generated index with the last element, and finally call the standard partition process which uses last element as pivot.
Following is C++ implementation of above Randomized QuickSelect.
// C++ implementation of randomized quickSelect
#include<iostream>
#include<climits>
#include<cstdlib>
using namespace std;
int randomPartition(int arr[], int l, int r);
// This function returns k'th smallest element in arr[l..r] using
// QuickSort based method. ASSUMPTION: ALL ELEMENTS IN ARR[] ARE DISTINCT
int kthSmallest(int arr[], int l, int r, int k)
{
// If k is smaller than number of elements in array
if (k > 0 && k <= r - l + 1)
{
// Partition the array around a random element and
// get position of pivot element in sorted array
int pos = randomPartition(arr, l, r);
// If position is same as k
if (pos-l == k-1)
return arr[pos];
if (pos-l > k-1) // If position is more, recur for left subarray
return kthSmallest(arr, l, pos-1, k);
// Else recur for right subarray
return kthSmallest(arr, pos+1, r, k-pos+l-1);
}
// If k is more than number of elements in array
return INT_MAX;
}
void swap(int *a, int *b)
{
int temp = *a;
*a = *b;
*b = temp;
}
// Standard partition process of QuickSort(). It considers the last
// element as pivot and moves all smaller element to left of it and
// greater elements to right. This function is used by randomPartition()
int partition(int arr[], int l, int r)
{
int x = arr[r], i = l;
for (int j = l; j <= r - 1; j++)
{
if (arr[j] <= x)
{
swap(&arr[i], &arr[j]);

i++;
}
}
swap(&arr[i], &arr[r]);
return i;
}
// Picks a random pivot element between l and r and partitions
// arr[l..r] arount the randomly picked element using partition()
int randomPartition(int arr[], int l, int r)
{
int n = r-l+1;
int pivot = rand() % n;
swap(&arr[l + pivot], &arr[r]);
return partition(arr, l, r);
}
// Driver program to test above methods
int main()
{
int arr[] = {12, 3, 5, 7, 4, 19, 26};
int n = sizeof(arr)/sizeof(arr[0]), k = 3;
cout << "K'th smallest element is " << kthSmallest(arr, 0, n-1, k);
return 0;
}

Output:
K'th smallest element is 5

Time Complexity:
The worst case time complexity of the above solution is still O(n2). In worst case, the randomized function may always pick a corner element. The
expected time complexity of above randomized QuickSelect is ?(n), see CLRS book or MIT video lecture for proof. The assumption in the
analysis is, random number generator is equally likely to generate any number in the input range.
Sources:
MIT Video Lecture on Order Statistics, Median
Introduction to Algorithms by Clifford Stein, Thomas H. Cormen, Charles E. Leiserson, Ronald L.

Kth Smallest/Largest Element in Unsorted Array | Set 2 (Expected Linear Time)

We recommend to read following post as a prerequisite of this post.
Kth Smallest/Largest Element in Unsorted Array | Set 1
Given an array and a number k where k is smaller than size of array, we need to find the kth smallest element in the given array. It is given that ll
array elements are distinct.
Examples:
Input: arr[] = {7, 10, 4, 3, 20, 15}
k = 3
Output: 7
Input: arr[] = {7, 10, 4, 3, 20, 15}
k = 4
Output: 10

We have discussed three different solutions here.

In this post method 4 is discussed which is mainly an extension of method 3 (QuickSelect) discussed in the previous post. The idea is to randomly
pick a pivot element. To implement randomized partition, we use a random function, rand() to generate index between l and r, swap the element at
randomly generated index with the last element, and finally call the standard partition process which uses last element as pivot.
Following is C++ implementation of above Randomized QuickSelect.
// C++ implementation of randomized quickSelect
#include<iostream>
#include<climits>
#include<cstdlib>
using namespace std;
int randomPartition(int arr[], int l, int r);
// This function returns k'th smallest element in arr[l..r] using
// QuickSort based method. ASSUMPTION: ALL ELEMENTS IN ARR[] ARE DISTINCT
int kthSmallest(int arr[], int l, int r, int k)
{
// If k is smaller than number of elements in array
if (k > 0 && k <= r - l + 1)
{
// Partition the array around a random element and
// get position of pivot element in sorted array
int pos = randomPartition(arr, l, r);
// If position is same as k
if (pos-l == k-1)
return arr[pos];
if (pos-l > k-1) // If position is more, recur for left subarray
return kthSmallest(arr, l, pos-1, k);
// Else recur for right subarray
return kthSmallest(arr, pos+1, r, k-pos+l-1);
}
// If k is more than number of elements in array
return INT_MAX;
}
void swap(int *a, int *b)
{
int temp = *a;
*a = *b;
*b = temp;
}
// Standard partition process of QuickSort(). It considers the last
// element as pivot and moves all smaller element to left of it and
// greater elements to right. This function is used by randomPartition()
int partition(int arr[], int l, int r)
{
int x = arr[r], i = l;
for (int j = l; j <= r - 1; j++)
{
if (arr[j] <= x)
{
swap(&arr[i], &arr[j]);

i++;
}
}
swap(&arr[i], &arr[r]);
return i;
}
// Picks a random pivot element between l and r and partitions
// arr[l..r] arount the randomly picked element using partition()
int randomPartition(int arr[], int l, int r)
{
int n = r-l+1;
int pivot = rand() % n;
swap(&arr[l + pivot], &arr[r]);
return partition(arr, l, r);
}
// Driver program to test above methods
int main()
{
int arr[] = {12, 3, 5, 7, 4, 19, 26};
int n = sizeof(arr)/sizeof(arr[0]), k = 3;
cout << "K'th smallest element is " << kthSmallest(arr, 0, n-1, k);
return 0;
}

Output:
K'th smallest element is 5

Time Complexity:
The worst case time complexity of the above solution is still O(n2). In worst case, the randomized function may always pick a corner element. The
expected time complexity of above randomized QuickSelect is ?(n), see CLRS book or MIT video lecture for proof. The assumption in the
analysis is, random number generator is equally likely to generate any number in the input range.
Sources:
MIT Video Lecture on Order Statistics, Median
Introduction to Algorithms by Clifford Stein, Thomas H. Cormen, Charles E. Leiserson, Ronald L.

Kth Smallest/Largest Element in Unsorted Array | Set 3 (Worst Case Linear Time)
We recommend to read following posts as a prerequisite of this post.
Kth Smallest/Largest Element in Unsorted Array | Set 1
Kth Smallest/Largest Element in Unsorted Array | Set 2 (Expected Linear Time)
Given an array and a number k where k is smaller than size of array, we need to find the kth smallest element in the given array. It is given that ll
array elements are distinct.
Examples:
Input: arr[] = {7, 10, 4, 3, 20, 15}
k = 3
Output: 7
Input: arr[] = {7, 10, 4, 3, 20, 15}
k = 4
Output: 10

In previous post, we discussed an expected linear time algorithm. In this post, a worst case linear time method is discussed. The idea in this new
method is similar to quickSelect(), we get worst case linear time by selecting a pivot that divides array in a balanced way (there are not
very few elements on one side and many on other side). After the array is divided in a balanced way, we apply the same steps as used in
quickSelect() to decide whether to go left or right of pivot.
Following is complete algorithm.
kthSmallest(arr[0..n-1], k)
1) Divide arr[] into ?n/5rceil; groups where size of each group is 5
except possibly the last group which may have less than 5 elements.
2) Sort the above created ?n/5? groups and find median
of all groups. Create an auxiliary array 'median[]' and store medians
of all ?n/5? groups in this median array.
// Recursively call this method to find median of median[0..?n/5?-1]
3) medOfMed = kthSmallest(median[0..?n/5?-1], ?n/10?)
4) Partition arr[] around medOfMed and obtain its position.
pos = partition(arr, n, medOfMed)
5) If pos == k return medOfMed
6) If pos < k return kthSmallest(arr[l..pos-1], k)
7) If poa > k return kthSmallest(arr[pos+1..r], k-pos+l-1)

In above algorithm, last 3 steps are same as algorithm in previous post. The first four steps are used to obtain a good point for partitioning the array
(to make sure that there are not too many elements either side of pivot).
Following is C++ implementation of above algorithm.
// C++ implementation of worst case linear time algorithm
// to find k'th smallest element
#include<iostream>
#include<algorithm>
#include<climits>
using namespace std;
int partition(int arr[], int l, int r, int k);
// A simple function to find median of arr[]. This is called
// only for an array of size 5 in this program.
int findMedian(int arr[], int n)
{
sort(arr, arr+n); // Sort the array
return arr[n/2]; // Return middle element
}
// Returns k'th smallest element in arr[l..r] in worst case
// linear time. ASSUMPTION: ALL ELEMENTS IN ARR[] ARE DISTINCT
int kthSmallest(int arr[], int l, int r, int k)
{
// If k is smaller than number of elements in array
if (k > 0 && k <= r - l + 1)
{
int n = r-l+1; // Number of elements in arr[l..r]
// Divide arr[] in groups of size 5, calculate median

// of every group and store it in median[] array.

int i, median[(n+4)/5]; // There will be floor((n+4)/5) groups;
for (i=0; i<n/5; i++)
median[i] = findMedian(arr+l+i*5, 5);
if (i*5 < n) //For last group with less than 5 elements
{
median[i] = findMedian(arr+l+i*5, n%5);
i++;
}
// Find median of all medians using recursive call.
// If median[] has only one element, then no need
// of recursive call
int medOfMed = (i == 1)? median[i-1]:
kthSmallest(median, 0, i-1, i/2);
// Partition the array around a random element and
// get position of pivot element in sorted array
int pos = partition(arr, l, r, medOfMed);
// If position is same as k
if (pos-l == k-1)
return arr[pos];
if (pos-l > k-1) // If position is more, recur for left
return kthSmallest(arr, l, pos-1, k);
// Else recur for right subarray
return kthSmallest(arr, pos+1, r, k-pos+l-1);
}
// If k is more than number of elements in array
return INT_MAX;
}
void swap(int *a, int *b)
{
int temp = *a;
*a = *b;
*b = temp;
}
// It searches for x in arr[l..r], and partitions the array
// around x.
int partition(int arr[], int l, int r, int x)
{
// Search for x in arr[l..r] and move it to end
int i;
for (i=l; i<r; i++)
if (arr[i] == x)
break;
swap(&arr[i], &arr[r]);
// Standard partition algorithm
i = l;
for (int j = l; j <= r - 1; j++)
{
if (arr[j] <= x)
{
swap(&arr[i], &arr[j]);
i++;
}
}
swap(&arr[i], &arr[r]);
return i;
}
// Driver program to test above methods
int main()
{
int arr[] = {12, 3, 5, 7, 4, 19, 26};
int n = sizeof(arr)/sizeof(arr[0]), k = 3;
cout << "K'th smallest element is "
<< kthSmallest(arr, 0, n-1, k);
return 0;
}

Output:
K'th smallest element is 5

Time Complexity:

The worst case time complexity of the above algorithm is O(n). Let us analyze all steps.
The steps 1) and 2) take O(n) time as finding median of an array of size 5 takes O(1) time and there are n/5 arrays of size 5.
The step 3) takes T(n/5) time. The step 4 is standard partition and takes O(n) time.
The interesting steps are 6) and 7). At most, one of them is executed. These are recursive steps. What is the worst case size of these recursive
calls. The answer is maximum number of elements greater than medOfMed (obtained in step 3) or maximum number of elements smaller than
medOfMed.
How many elements are greater than medOfMed and how many are smaller?
At least half of the medians found in step 2 are greater than or equal to medOfMed. Thus, at least half of the n/5 groups contribute 3 elements that
are greater than medOfMed, except for the one group that has fewer than 5 elements. Therefore, the number of elements greater than medOfMed
is at least.

Similarly, the number of elements that are less than medOfMed is at least 3n/10 6. In the worst case, the function recurs for at most n (3n/10 6)
which is 7n/10 + 6 elements.
Note that 7n/10 + 6 < n for n > 20 and that any input of 80 or fewer elements requires O(1) time. We can therefore obtain the recurrence

We show that the running time is linear by substitution. Assume that T(n) cn for some constant c and all n > 80. Substituting this inductive
hypothesis into the right-hand side of the recurrence yields
T(n) <= cn/5 + c(7n/10 + 6) + O(n)
<= cn/5 + c + 7cn/10 + 6c + O(n)
<= 9cn/10 + 7c + O(n)
<= cn,

since we can pick c large enough so that c(n/10 - 7) is larger than the function described by the O(n) term for all n > 80. The worst-case running
time of is therefore linear (Source: http://staff.ustc.edu.cn/~csli/graduate/algorithms/book6/chap10.htm ).
Note that the above algorithm is linear in worst case, but the constants are very high for this algorithm. Therefore, this algorithm doesn't work well
in practical situations, randomized quickSelect works much better and preferred.
Sources:
MIT Video Lecture on Order Statistics, Median
Introduction to Algorithms by Clifford Stein, Thomas H. Cormen, Charles E. Leiserson, Ronald L.
http://staff.ustc.edu.cn/~csli/graduate/algorithms/book6/chap10.htm

Find the closest pair from two sorted arrays

Given two sorted arrays and a number x, find the pair whose sum is closest to x and the pair has an element from each array.
We are given two arrays ar1[0m-1] and ar2[0..n-1] and a number x, we need to find the pair ar1[i] + ar2[j] such that absolute value of (ar1[i] +
ar2[j] x) is minimum.
Example:
Input: ar1[] = {1, 4, 5, 7};
ar2[] = {10, 20, 30, 40};
x = 32
Output: 1 and 30
Input: ar1[] = {1, 4, 5, 7};
ar2[] = {10, 20, 30, 40};
x = 50
Output: 7 and 40

A Simple Solution is to run two loops. The outer loop considers every element of first array and inner loop checks for the pair in second array.
We keep track of minimum difference between ar1[i] + ar2[j] and x.
We can do it in O(n) time using following steps.
1) Merge given two arrays into an auxiliary array of size m+n using merge process of merge sort. While merging keep another boolean array of
size m+n to indicate whether the current element in merged array is from ar1[] or ar2[].
2) Consider the merged array and use the linear time algorithm to find the pair with sum closest to x. One extra thing we need to consider only
those pairs which have one element from ar1[] and other from ar2[], we use the boolean array for this purpose.
Can we do it in a single pass and O(1) extra space?
The idea is to start from left side of one array and right side of another array, and use the algorithm same as step 2 of above approach. Following is
detailed algorithm.
1) Initialize a variable diff as infinite (Diff is used to store the
difference between pair and x). We need to find the minimum diff.
2) Initialize two index variables l and r in the given sorted array.
(a) Initialize first to the leftmost index in ar1: l = 0
(b) Initialize second the rightmost index in ar2: r = n-1
3) Loop while l < m and r >= 0
(a) If abs(ar1[l] + ar2[r] - sum) < diff then
update diff and result
(b) Else if(ar1[l] + ar2[r] < sum ) then l++
(c) Else r-4) Print the result.

Following is C++ implementation of this approach.

// C++ program to find the pair from two sorted arays such
// that the sum of pair is closest to a given number x
#include <iostream>
#include <climits>
#include <cstdlib>
using namespace std;
// ar1[0..m-1] and ar2[0..n-1] are two given sorted arrays
// and x is given number. This function prints the pair from
// both arrays such that the sum of the pair is closest to x.
void printClosest(int ar1[], int ar2[], int m, int n, int x)
{
// Initialize the diff between pair sum and x.
int diff = INT_MAX;
// res_l and res_r are result indexes from ar1[] and ar2[]
// respectively
int res_l, res_r;
// Start from left side of ar1[] and right side of ar2[]
int l = 0, r = n-1;
while (l<m && r>=0)
{
// If this pair is closer to x than the previously
// found closest, then update res_l, res_r and diff
if (abs(ar1[l] + ar2[r] - x) < diff)
{
res_l = l;
res_r = r;
diff = abs(ar1[l] + ar2[r] - x);

}
// If sum of this pair is more than x, move to smaller
// side
if (ar1[l] + ar2[r] > x)
r--;
else // move to the greater side
l++;
}
// Print the result
cout << "The closest pair is [" << ar1[res_l] << ", "
<< ar2[res_r] << "] \n";
}
// Driver program to test above functions
int main()
{
int ar1[] = {1, 4, 5, 7};
int ar2[] = {10, 20, 30, 40};
int m = sizeof(ar1)/sizeof(ar1[0]);
int n = sizeof(ar2)/sizeof(ar2[0]);
int x = 38;
printClosest(ar1, ar2, m, n, x);
return 0;
}

Output:
The closest pair is [7, 30]

Find common elements in three sorted arrays

Given three arrays sorted in non-decreasing order, print all common elements in these arrays.
Examples:
ar1[] =
ar2[] =
ar3[] =
Output:

{1,
{6,
{3,
20,

5, 10, 20, 40, 80}

7, 20, 80, 100}
4, 15, 20, 30, 70, 80, 120}
80

ar1[] =
ar2[] =
ar3[] =
Outptu:

{1, 5, 5}
{3, 4, 5, 5, 10}
{5, 5, 10, 20}
5, 5

A simple solution is to first find intersection of two arrays and store the intersection in a temporary array, then find the intersection of third array
and temporary array. Time complexity of this solution is O(n1 + n2 + n3) where n1, n2 and n3 are sizes of ar1[], ar2[] and ar3[] respectively.
The above solution requires extra space and two loops, we can find the common elements using a single loop and without extra space. The idea is
similar to intersection of two arrays. Like two arrays loop, we run a loop and traverse three arrays.
Let the current element traversed in ar1[] be x, in ar2[] be y and in ar3[] be z. We can have following cases inside the loop.
1) If x, y and z are same, we can simply print any of them as common element and move ahead in all three arrays.
2) Else If x < y, we can move ahead in ar1[] as x cannot be a common element 3) Else If y < z, we can move ahead in ar2[] as y cannot be a
common element 4) Else (We reach here when x > y and y > z), we can simply move ahead in ar3[] as z cannot be a common element.
Following are implementations of the above idea.

C++
// C++ program to print common elements in three arrays
#include <iostream>
using namespace std;
// This function prints common elements in ar1
int findCommon(int ar1[], int ar2[], int ar3[], int n1, int n2, int n3)
{
// Initialize starting indexes for ar1[], ar2[] and ar3[]
int i = 0, j = 0, k = 0;
// Iterate through three arrays while all arrays have elements
while (i < n1 && j < n2 && k < n3)
{
// If x = y and y = z, print any of them and move ahead
// in all arrays
if (ar1[i] == ar2[j] && ar2[j] == ar3[k])
{ cout << ar1[i] << " "; i++; j++; k++; }
// x < y
else if (ar1[i] < ar2[j])
i++;
// y < z
else if (ar2[j] < ar3[k])
j++;
// We reach here when x > y and z < y, i.e., z is smallest
else
k++;
}
}
// Driver program to test above function
int main()
{
int ar1[] = {1, 5, 10, 20, 40, 80};
int ar2[] = {6, 7, 20, 80, 100};
int ar3[] = {3, 4, 15, 20, 30, 70, 80, 120};
int n1 = sizeof(ar1)/sizeof(ar1[0]);
int n2 = sizeof(ar2)/sizeof(ar2[0]);
int n3 = sizeof(ar3)/sizeof(ar3[0]);
cout << "Common Elements are ";
findCommon(ar1, ar2, ar3, n1, n2, n3);
return 0;
}

Python
# Python function to print common elements in three sorted arrays
def findCommon(ar1, ar2, ar3, n1, n2, n3):
# Initialize starting indexes for ar1[], ar2[] and ar3[]
i, j, k = 0, 0, 0
# Iterate through three arrays while all arrays have elements
while (i < n1 and j < n2 and k< n3):
# If x = y and y = z, print any of them and move ahead
# in all arrays
if (ar1[i] == ar2[j] and ar2[j] == ar3[k]):
print ar1[i],
i += 1
j += 1
k += 1
# x < y
elif ar1[i] < ar2[j]:
i += 1
# y < z
elif ar2[j] < ar3[k]:
j += 1
# We reach here when x > y and z < y, i.e., z is smallest
else:
k += 1
#Driver program to check above function
ar1 = [1, 5, 10, 20, 40, 80]
ar2 = [6, 7, 20, 80, 100]
ar3 = [3, 4, 15, 20, 30, 70, 80, 120]
n1 = len(ar1)
n2 = len(ar2)
n3 = len(ar3)
print "Common elements are",
findCommon(ar1, ar2, ar3, n1, n2, n3)
# This code is contributed by __Devesh Agrawal__

Common Elements are 20 80

Time complexity of the above solution is O(n1 + n2 + n3). In worst case, the largest sized array may have all small elements and middle sized array
has all middle elements.

Given a sorted array and a number x, find the pair in array whose sum is closest to x
Given a sorted array and a number x, find a pair in array whose sum is closest to x.
Examples:
Input: arr[] = {10, 22, 28, 29, 30, 40}, x = 54
Output: 22 and 30
Input: arr[] = {1, 3, 4, 7, 10}, x = 15
Output: 4 and 10

A simple solution is to consider every pair and keep track of closest pair (absolute difference between pair sum and x is minimum). Finally print the
closest pair. Time complexity of this solution is O(n2)
An efficient solution can find the pair in O(n) time. The idea is similar to method 2 of this post. Following is detailed algorithm.
1) Initialize a variable diff as infinite (Diff is used to store the
difference between pair and x). We need to find the minimum diff.
2) Initialize two index variables l and r in the given sorted array.
(a) Initialize first to the leftmost index: l = 0
(b) Initialize second the rightmost index: r = n-1
3) Loop while l < r.
(a) If abs(arr[l] + arr[r] - sum) < diff then
update diff and result
(b) Else if(arr[l] + arr[r] < sum ) then l++
(c) Else r--

Following is C++ implementation of above algorithm.

C++
// Simple C++ program to find the pair with sum closest to a given no.
#include <iostream>
#include <climits>
#include <cstdlib>
using namespace std;
// Prints the pair with sum closest to x
void printClosest(int arr[], int n, int x)
{
int res_l, res_r; // To store indexes of result pair
// Initialize left and right indexes and difference between
// pair sum and x
int l = 0, r = n-1, diff = INT_MAX;
// While there are elements between l and r
while (r > l)
{
// Check if this pair is closer than the closest pair so far
if (abs(arr[l] + arr[r] - x) < diff)
{
res_l = l;
res_r = r;
diff = abs(arr[l] + arr[r] - x);
}
// If this pair has more sum, move to smaller values.
if (arr[l] + arr[r] > x)
r--;
else // Move to larger values
l++;
}
cout <<" The closest pair is " << arr[res_l] << " and " << arr[res_r];
}
// Driver program to test above functions
int main()
{
int arr[] = {10, 22, 28, 29, 30, 40}, x = 54;
int n = sizeof(arr)/sizeof(arr[0]);
printClosest(arr, n, x);
return 0;
}

Java
// Java program to find pair with sum closest to x
import java.io.*;
import java.util.*;
import java.lang.Math;
class CloseSum {
// Prints the pair with sum cloest to x
static void printClosest(int arr[], int n, int x)
{
int res_l=0, res_r=0; // To store indexes of result pair
// Initialize left and right indexes and difference between
// pair sum and x
int l = 0, r = n-1, diff = Integer.MAX_VALUE;
// While there are elements between l and r
while (r > l)
{
// Check if this pair is closer than the closest pair so far
if (Math.abs(arr[l] + arr[r] - x) < diff)
{
res_l = l;
res_r = r;
diff = Math.abs(arr[l] + arr[r] - x);
}
// If this pair has more sum, move to smaller values.
if (arr[l] + arr[r] > x)
r--;
else // Move to larger values
l++;
}
System.out.println(" The closest pair is "+arr[res_l]+" and "+ arr[res_r]);
}
// Driver program to test above function
public static void main(String[] args)
{
int arr[] = {10, 22, 28, 29, 30, 40}, x = 54;
int n = arr.length;
printClosest(arr, n, x);
}
}
/*This code is contributed by Devesh Agrawal*/

The closest pair is 22 and 30

Count 1s in a sorted binary array

Given a binary array sorted in non-increasing order, count the number of 1s in it.
Examples:
Input: arr[] = {1, 1, 0, 0, 0, 0, 0}
Output: 2
Input: arr[] = {1, 1, 1, 1, 1, 1, 1}
Output: 7
Input: arr[] = {0, 0, 0, 0, 0, 0, 0}
Output: 0

A simple solution is to linearly traverse the array. The time complexity of the simple solution is O(n). We can use Binary Search to find count in
O(Logn) time. The idea is to look for last occurrence of 1 using Binary Search. Once we find the index last occurrence, we return index + 1 as
count.
The following is C++ implementation of above idea.

C++
// C++ program to count one's in a boolean array
#include <iostream>
using namespace std;
/* Returns counts of 1's in arr[low..high]. The array is
assumed to be sorted in non-increasing order */
int countOnes(bool arr[], int low, int high)
{
if (high >= low)
{
// get the middle index
int mid = low + (high - low)/2;
// check if the element at middle index is last 1
if ( (mid == high || arr[mid+1] == 0) && (arr[mid] == 1))
return mid+1;
// If element is not last 1, recur for right side
if (arr[mid] == 1)
return countOnes(arr, (mid + 1), high);
// else recur for left side
return countOnes(arr, low, (mid -1));
}
return 0;
}
/* Driver program to test above functions */
int main()
{
bool arr[] = {1, 1, 1, 1, 0, 0, 0};
int n = sizeof(arr)/sizeof(arr[0]);
cout << "Count of 1's in given array is " << countOnes(arr, 0, n-1);
return 0;
}

Python
# Python program to count one's in a boolean array
# Returns counts of 1's in arr[low..high]. The array is
# assumed to be sorted in non-increasing order
def countOnes(arr,low,high):
if high>=low:
# get the middle index
mid = low + (high-low)/2
# check if the element at middle index is last 1
if ((mid == high or arr[mid+1]==0) and (arr[mid]==1)):
return mid+1
# If element is not last 1, recur for right side

if arr[mid]==1:
return countOnes(arr, (mid+1), high)
# else recur for left side
return countOnes(arr, low, mid-1)
return 0
# Driver function
arr=[1, 1, 1, 1, 0, 0, 0]
print "Count of 1's in given array is",countOnes(arr, 0 , len(arr)-1)
# This code is contributed by __Devesh Agrawal__

Output:
Count of 1's in given array is 4

Time complexity of the above solution is O(Logn)

Binary Insertion Sort

We can use binary search to reduce the number of comparisons in normal insertion sort. Binary Insertion Sort find use binary search to find the
proper location to insert the selected item at each iteration.
In normal insertion, sort it takes O(i) (at ith iteration) in worst case. we can reduce it to O(logi) by using binary search.
// C program for implementation of binary insertion sort
#include <stdio.h>
// A binary search based function to find the position
// where item should be inserted in a[low..high]
int binarySearch(int a[], int item, int low, int high)
{
if (high <= low)
return (item > a[low])? (low + 1): low;
int mid = (low + high)/2;
if(item == a[mid])
return mid+1;
if(item > a[mid])
return binarySearch(a, item, mid+1, high);
return binarySearch(a, item, low, mid-1);
}
// Function to sort an array a[] of size 'n'
void insertionSort(int a[], int n)
{
int i, loc, j, k, selected;
for (i = 1; i < n; ++i)
{
j = i - 1;
selected = a[i];
// find location where selected sould be inseretd
loc = binarySearch(a, selected, 0, j);
// Move all elements after location to create space
while (j >= loc)
{
a[j+1] = a[j];
j--;
}
a[j+1] = selected;
}
}
// Driver program to test above function
int main()
{
int a[] = {37, 23, 0, 17, 12, 72, 31,
46, 100, 88, 54};
int n = sizeof(a)/sizeof(a[0]), i;
insertionSort(a, n);
printf("Sorted array: \n");
for (i = 0; i < n; i++)
printf("%d ",a[i]);
return 0;
}

Output:
Sorted array:
0 12 17 23 31 37 46 54 72 88 100

Time Complexity: The algorithm as a whole still has a running worst case running time of O(n2) because of the series of swaps required for each
insertion.

Insertion Sort for Singly Linked List

We have discussed Insertion Sort for arrays. In this article same for linked list is discussed.
Below is simple insertion sort algorithm for linked list.
1) Create an empty sorted (or result) list
2) Traverse the given list, do following for every node.
......a) Insert current node in sorted way in sorted or result list.
3) Change head of given linked list to head of sorted (or result) list.

The main step is (2.a) which has been covered in below post.
Sorted Insert for Singly Linked List
Below is C implementation of above algorithm
/* C program for insertion sort on a linked list */
#include<stdio.h>
#include<stdlib.h>
/* Link list node */
struct node
{
int data;
struct node* next;
};
// Function to insert a given node in a sorted linked list
void sortedInsert(struct node**, struct node*);
// function to sort a singly linked list using insertion sort
void insertionSort(struct node **head_ref)
{
// Initialize sorted linked list
struct node *sorted = NULL;
// Traverse the given linked list and insert every
// node to sorted
struct node *current = *head_ref;
while (current != NULL)
{
// Store next for next iteration
struct node *next = current->next;
// insert current in sorted linked list
sortedInsert(&sorted, current);
// Update current
current = next;
}
// Update head_ref to point to sorted linked list
*head_ref = sorted;
}
/* function to insert a new_node in a list. Note that this
function expects a pointer to head_ref as this can modify the
head of the input linked list (similar to push())*/
void sortedInsert(struct node** head_ref, struct node* new_node)
{
struct node* current;
/* Special case for the head end */
if (*head_ref == NULL || (*head_ref)->data >= new_node->data)
{
new_node->next = *head_ref;
*head_ref = new_node;
}
else
{
/* Locate the node before the point of insertion */
current = *head_ref;
while (current->next!=NULL &&
current->next->data < new_node->data)
{
current = current->next;
}
new_node->next = current->next;
current->next = new_node;

}
}
/* BELOW FUNCTIONS ARE JUST UTILITY TO TEST sortedInsert */
/* A utility function to create a new node */
struct node *newNode(int new_data)
{
/* allocate node */
struct node* new_node =
(struct node*) malloc(sizeof(struct node));
/* put in the data */
new_node->data = new_data;
new_node->next = NULL;
return new_node;
}
/* Function to print linked list */
void printList(struct node *head)
{
struct node *temp = head;
while(temp != NULL)
{
printf("%d ", temp->data);
temp = temp->next;
}
}
/* A utility function to insert a node at the beginning of linked list */
void push(struct node** head_ref, int new_data)
{
/* allocate node */
struct node* new_node = new node;
/* put in the data */
new_node->data = new_data;
/* link the old list off the new node */
new_node->next = (*head_ref);
/* move the head to point to the new node */
(*head_ref)
= new_node;
}
// Driver program to test above functions
int main()
{
struct node *a = NULL;
push(&a, 5);
push(&a, 20);
push(&a, 4);
push(&a, 3);
push(&a, 30);
printf("Linked List before sorting \n");
printList(a);
insertionSort(&a);
printf("\nLinked List after sorting \n");
printList(a);
return 0;
}
Linked List before sorting
30 3 4 20 5
Linked List after sorting
3 4 5 20 30

Why Quick Sort preferred for Arrays and Merge Sort for Linked Lists?
Why is Quick Sort preferred for arrays?
Below are recursive and iterative implementations of Quick Sort and Merge Sort for arrays.
Recursive Quick Sort for array.
Iterative Quick Sort for arrays.
Recursive Merge Sort for arrays
Iterative Merge Sort for arrays
Quick Sort in its general form is an in-place sort (i.e. it doesnt require any extra storage) whereas merge sort requires O(N) extra storage, N
denoting the array size which may be quite expensive. Allocating and de-allocating the extra space used for merge sort increases the running time
of the algorithm. Comparing average complexity we find that both type of sorts have O(NlogN) average complexity but the constants differ. For
arrays, merge sort loses due to the use of extra O(N) storage space.
Most practical implementations of Quick Sort use randomized version. The randomized version has expected time complexity of O(nLogn). The
worst case is possible in randomized version also, but worst case doesnt occur for a particular pattern (like sorted array) and randomized Quick
Sort works well in practice.
Quick Sort is also a cache friendly sorting algorithm as it has good locality of reference when used for arrays.
Quick Sort is also tail recursive, therefore tail call optimizations is done.
Why is Merge Sort preferred for Linked Lists?
Below are implementations of Quicksort and Mergesort for singly and doubly linked lists.
Quick Sort for Doubly Linked List
Quick Sort for Singly Linked List
Merge Sort for Singly Linked List
Merge Sort for Doubly Linked List
In case of linked lists the case is different mainly due to difference in memory allocation of arrays and linked lists. Unlike arrays, linked list nodes
may not be adjacent in memory. Unlike array, in linked list, we can insert items in the middle in O(1) extra space and O(1) time. Therefore merge
operation of merge sort can be implemented without extra space for linked lists.
In arrays, we can do random access as elements are continuous in memory. Let us say we have an integer (4-byte) array A and let the address of
A[0] be x then to access A[i], we can directly access the memory at (x + i*4). Unlike arrays, we can not do random access in linked list. Quick
Sort requires a lot of this kind of access. In linked list to access ith index, we have to travel each and every node from the head to ith node as we
dont have continuous block of memory. Therefore, the overhead increases for quick sort. Merge sort accesses data sequentially and the need of
random access is low.

Merge Sort for Doubly Linked List

Given a doubly linked list, write a function to sort the doubly linked list in increasing order using merge sort.
For example, the following doubly linked list should be changed to 2<->4<->8<->10

Below is C implementation of merge sort for doubly linked list.

// C program for merge sort on doubly linked list
#include<stdio.h>
#include<stdlib.h>
struct node
{
int data;
struct node *next, *prev;
};
struct node *split(struct node *head);
// Function to merge two linked lists
struct node *merge(struct node *first, struct node *second)
{
// If first linked list is empty
if (!first)
return second;
// If second linked list is empty
if (!second)
return first;
// Pick the smaller value
if (first->data < second->data)
{
first->next = merge(first->next,second);
first->next->prev = first;
first->prev = NULL;
return first;
}
else
{
second->next = merge(first,second->next);
second->next->prev = second;
second->prev = NULL;
return second;
}
}
// Function to do merge sort
struct node *mergeSort(struct node *head)
{
if (!head || !head->next)
return head;
struct node *second = split(head);
// Recur for left and right halves
head = mergeSort(head);
second = mergeSort(second);
// Merge the two sorted halves
return merge(head,second);
}
// A utility function to insert a new node at the
// beginning of doubly linked list
void insert(struct node **head, int data)
{
struct node *temp =
(struct node *)malloc(sizeof(struct node));
temp->data = data;
temp->next = temp->prev = NULL;
if (!(*head))
(*head) = temp;
else

{
temp->next = *head;
(*head)->prev = temp;
(*head) = temp;
}
}
// A utility function to print a doubly linked list in
// both forward and backward directions
void print(struct node *head)
{
struct node *temp = head;
printf("Forward Traversal using next poitner\n");
while (head)
{
printf("%d ",head->data);
temp = head;
head = head->next;
}
printf("\nBackword Traversal using prev pointer\n");
while (temp)
{
printf("%d ", temp->data);
temp = temp->prev;
}
}
// Utility function to swap two integers
void swap(int *A, int *B)
{
int temp = *A;
*A = *B;
*B = temp;
}
// Split a doubly linked list (DLL) into 2 DLLs of
// half sizes
struct node *split(struct node *head)
{
struct node *fast = head,*slow = head;
while (fast->next && fast->next->next)
{
fast = fast->next->next;
slow = slow->next;
}
struct node *temp = slow->next;
slow->next = NULL;
return temp;
}
// Driver program
int main(void)
{
struct node *head = NULL;
insert(&head,5);
insert(&head,20);
insert(&head,4);
insert(&head,3);
insert(&head,30);
insert(&head,10);
printf("Linked List before sorting\n");
print(head);
head = mergeSort(head);
printf("\n\nLinked List after sorting\n");
print(head);
return 0;
}

Output:
Linked List before sorting
Forward Traversal using next pointer
10 30 3 4 20 5
Backward Traversal using prev pointer
5 20 4 3 30 10
Linked List after sorting
Forward Traversal using next pointer
3 4 5 10 20 30
Backward Traversal using prev pointer
30 20 10 5 4 3

Thanks to Goku for providing above implementation in a comment here.

Time Complexity: Time complexity of the above implementation is same as time complexity of MergeSort for arrays. It takes ?(nLogn) time.
You may also like to see QuickSort for doubly linked list

Greedy Algorithms | Set 1 (Activity Selection Problem)

Greedy is an algorithmic paradigm that builds up a solution piece by piece, always choosing the next piece that offers the most obvious and
immediate benefit. Greedy algorithms are used for optimization problems. An optimization problem can be solved using Greedy if the problem has
the following property: At every step, we can make a choice that looks best at the moment, and we get the optimal solution of the complete
problem.
If a Greedy Algorithm can solve a problem, then it generally becomes the best method to solve that problem as the Greedy algorithms are in
general more efficient than other techniques like Dynamic Programming. But Greedy algorithms cannot always be applied. For example, Fractional
Knapsack problem (See this) can be solved using Greedy, but 0-1 Knapsack cannot be solved using Greedy.
Following are some standard algorithms that are Greedy algorithms.
1) Kruskals Minimum Spanning Tree (MST): In Kruskals algorithm, we create a MST by picking edges one by one. The Greedy Choice is to
pick the smallest weight edge that doesnt cause a cycle in the MST constructed so far.
2) Prims Minimum Spanning Tree: In Prims algorithm also, we create a MST by picking edges one by one. We maintain two sets: set of the
vertices already included in MST and the set of the vertices not yet included. The Greedy Choice is to pick the smallest weight edge that connects
the two sets.
3) Dijkstras Shortest Path: The Dijkstras algorithm is very similar to Prims algorithm. The shortest path tree is built up, edge by edge. We
maintain two sets: set of the vertices already included in the tree and the set of the vertices not yet included. The Greedy Choice is to pick the edge
that connects the two sets and is on the smallest weight path from source to the set that contains not yet included vertices.
4) Huffman Coding: Huffman Coding is a loss-less compression technique. It assigns variable length bit codes to different characters. The
Greedy Choice is to assign least bit length code to the most frequent character.
The greedy algorithms are sometimes also used to get an approximation for Hard optimization problems. For example, Traveling Salesman
Problem is a NP Hard problem. A Greedy choice for this problem is to pick the nearest unvisited city from the current city at every step. This
solutions doesnt always produce the best optimal solution, but can be used to get an approximate optimal solution.
Let us consider the Activity Selection problem as our first example of Greedy algorithms. Following is the problem statement.
You are given n activities with their start and finish times. Select the maximum number of activities that can be performed by a single
person, assuming that a person can only work on a single activity at a time.
Example:
Consider the following 6 activities.
start[] = {1, 3, 0, 5, 8, 5};
finish[] = {2, 4, 6, 7, 9, 9};
The maximum set of activities that can be executed
by a single person is {0, 1, 3, 4}

The greedy choice is to always pick the next activity whose finish time is least among the remaining activities and the start time is more than or
equal to the finish time of previously selected activity. We can sort the activities according to their finishing time so that we always consider the next
activity as minimum finishing time activity.
1) Sort the activities according to their finishing time
2) Select the first activity from the sorted array and print it.
3) Do following for remaining activities in the sorted array.
.a) If the start time of this activity is greater than the finish time of previously selected activity then select this activity and print it.
In the following C implementation, it is assumed that the activities are already sorted according to their finish time.

C++
#include<stdio.h>
// Prints a maximum set of activities that can be done by a single
// person, one at a time.
// n --> Total number of activities
// s[] --> An array that contains start time of all activities
// f[] --> An array that contains finish time of all activities
void printMaxActivities(int s[], int f[], int n)
{
int i, j;
printf ("Following activities are selected \n");
// The first activity always gets selected
i = 0;
printf("%d ", i);
// Consider rest of the activities
for (j = 1; j < n; j++)
{
// If this activity has start time greater than or

// equal to the finish time of previously selected

// activity, then select it
if (s[j] >= f[i])
{
printf ("%d ", j);
i = j;
}
}
}
// driver program to test above function
int main()
{
int s[] = {1, 3, 0, 5, 8, 5};
int f[] = {2, 4, 6, 7, 9, 9};
int n = sizeof(s)/sizeof(s[0]);
printMaxActivities(s, f, n);
getchar();
return 0;
}

Python
"""The following implemenatation assumes that the activities
are already sorted according to their finish time"""
"""Prints a maximum set of activities that can be done by a
single person, one at a time"""
# n --> Total number of activities
# s[]--> An array that contains start time of all activities
# f[] --> An array that conatins finish time of all activities
def printMaxActivities(s , f ):
n = len(f)
print "The following activities are selected"
# The first activity is always selected
i = 0
print i,
# Consider rest of the activities
for j in xrange(n):
# If this activity has start time greater than
# or equal to the finish time of previously
# selected activity, then select it
if s[j] >= f[i]:
print j,
i = j
# Driver program to test above function
s = [1 , 3 , 0 , 5 , 8 , 5]
f = [2 , 4 , 6 , 7 , 9 , 9]
printMaxActivities(s , f)
# This code is contributed by Nikhil Kumar Singh

Following activities are selected

0 1 3 4

How does Greedy Choice work for Activities sorted according to finish time?
Let the give set of activities be S = {1, 2, 3, ..n} and activities be sorted by finish time. The greedy choice is to always pick activity 1. How come
the activity 1 always provides one of the optimal solutions. We can prove it by showing that if there is another solution B with first activity other
than 1, then there is also a solution A of same size with activity 1 as first activity. Let the first activity selected by B be k, then there always exist A
= {B {k}} U {1}.(Note that the activities in B are independent and k has smallest finishing time among all. Since k is not 1, finish(k) >= finish(1)).
References:
Introduction to Algorithms by Thomas H. Cormen, Charles E. Leiserson, Ronald L. Rivest, Clifford Stein
Algorithms by S. Dasgupta, C.H. Papadimitriou, and U.V. Vazirani
http://en.wikipedia.org/wiki/Greedy_algorithm

Greedy Algorithms | Set 2 (Kruskals Minimum Spanning Tree Algorithm)

What is Minimum Spanning Tree?
Given a connected and undirected graph, a spanning tree of that graph is a subgraph that is a tree and connects all the vertices together. A single
graph can have many different spanning trees. A minimum spanning tree (MST) or minimum weight spanning tree for a weighted, connected and
undirected graph is a spanning tree with weight less than or equal to the weight of every other spanning tree. The weight of a spanning tree is the
sum of weights given to each edge of the spanning tree.
How many edges does a minimum spanning tree has?
A minimum spanning tree has (V 1) edges where V is the number of vertices in the given graph.
What are the applications of Minimum Spanning Tree?
See this for applications of MST.
Below are the steps for finding MST using Kruskals algorithm
1. Sort all the edges in non-decreasing order of their weight.
2. Pick the smallest edge. Check if it forms a cycle with the spanning tree
formed so far. If cycle is not formed, include this edge. Else, discard it.
3. Repeat step#2 until there are (V-1) edges in the spanning tree.

The step#2 uses Union-Find algorithm to detect cycle. So we recommend to read following post as a prerequisite.
Union-Find Algorithm | Set 1 (Detect Cycle in a Graph)
Union-Find Algorithm | Set 2 (Union By Rank and Path Compression)
The algorithm is a Greedy Algorithm. The Greedy Choice is to pick the smallest weight edge that does not cause a cycle in the MST constructed
so far. Let us understand it with an example: Consider the below input graph.

The graph contains 9 vertices and 14 edges. So, the minimum spanning tree formed will be having (9 1) = 8 edges.
After sorting:
Weight Src
Dest
1
7
6
2
8
2
2
6
5
4
0
1
4
2
5
6
8
6
7
2
3
7
7
8
8
0
7
8
1
2
9
3
4
10
5
4
11
1
7
14
3
5

Now pick all edges one by one from sorted list of edges
1. Pick edge 7-6: No cycle is formed, include it.

2. Pick edge 8-2: No cycle is formed, include it.

3. Pick edge 6-5: No cycle is formed, include it.

4. Pick edge 0-1: No cycle is formed, include it.

5. Pick edge 2-5: No cycle is formed, include it.

6. Pick edge 8-6: Since including this edge results in cycle, discard it.
7. Pick edge 2-3: No cycle is formed, include it.

8. Pick edge 7-8: Since including this edge results in cycle, discard it.
9. Pick edge 0-7: No cycle is formed, include it.

10. Pick edge 1-2: Since including this edge results in cycle, discard it.
11. Pick edge 3-4: No cycle is formed, include it.

Since the number of edges included equals (V 1), the algorithm stops here.

C/C++
// C++ program for Kruskal's algorithm to find Minimum Spanning Tree
// of a given connected, undirected and weighted graph
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
// a structure to represent a weighted edge in graph
struct Edge
{
int src, dest, weight;

};
// a structure to represent a connected, undirected and weighted graph
struct Graph
{
// V-> Number of vertices, E-> Number of edges
int V, E;
// graph is represented as an array of edges. Since the graph is
// undirected, the edge from src to dest is also edge from dest
// to src. Both are counted as 1 edge here.
struct Edge* edge;
};
// Creates a graph with V vertices and E edges
struct Graph* createGraph(int V, int E)
{
struct Graph* graph = (struct Graph*) malloc( sizeof(struct Graph) );
graph->V = V;
graph->E = E;
graph->edge = (struct Edge*) malloc( graph->E * sizeof( struct Edge ) );
return graph;
}
// A structure to represent a subset for union-find
struct subset
{
int parent;
int rank;
};
// A utility function to find set of an element i
// (uses path compression technique)
int find(struct subset subsets[], int i)
{
// find root and make root as parent of i (path compression)
if (subsets[i].parent != i)
subsets[i].parent = find(subsets, subsets[i].parent);
return subsets[i].parent;
}
// A function that does union of two sets of x and y
// (uses union by rank)
void Union(struct subset subsets[], int x, int y)
{
int xroot = find(subsets, x);
int yroot = find(subsets, y);
// Attach smaller rank tree under root of high rank tree
// (Union by Rank)
if (subsets[xroot].rank < subsets[yroot].rank)
subsets[xroot].parent = yroot;
else if (subsets[xroot].rank > subsets[yroot].rank)
subsets[yroot].parent = xroot;
// If ranks are same, then make one as root and increment
// its rank by one
else
{
subsets[yroot].parent = xroot;
subsets[xroot].rank++;
}
}
// Compare two edges according to their weights.
// Used in qsort() for sorting an array of edges
int myComp(const void* a, const void* b)
{
struct Edge* a1 = (struct Edge*)a;
struct Edge* b1 = (struct Edge*)b;
return a1->weight > b1->weight;
}
// The main function to construct MST using Kruskal's algorithm
void KruskalMST(struct Graph* graph)
{
int V = graph->V;
struct Edge result[V]; // Tnis will store the resultant MST

int e = 0; // An index variable, used for result[]

int i = 0; // An index variable, used for sorted edges
// Step 1: Sort all the edges in non-decreasing order of their weight
// If we are not allowed to change the given graph, we can create a copy of
// array of edges
qsort(graph->edge, graph->E, sizeof(graph->edge[0]), myComp);
// Allocate memory for creating V ssubsets
struct subset *subsets =
(struct subset*) malloc( V * sizeof(struct subset) );
// Create V subsets with single elements
for (int v = 0; v < V; ++v)
{
subsets[v].parent = v;
subsets[v].rank = 0;
}
// Number of edges to be taken is equal to V-1
while (e < V - 1)
{
// Step 2: Pick the smallest edge. And increment the index
// for next iteration
struct Edge next_edge = graph->edge[i++];
int x = find(subsets, next_edge.src);
int y = find(subsets, next_edge.dest);
// If including this edge does't cause cycle, include it
// in result and increment the index of result for next edge
if (x != y)
{
result[e++] = next_edge;
Union(subsets, x, y);
}
// Else discard the next_edge
}
// print the contents of result[] to display the built MST
printf("Following are the edges in the constructed MST\n");
for (i = 0; i < e; ++i)
printf("%d -- %d == %d\n", result[i].src, result[i].dest,
result[i].weight);
return;
}
// Driver program to test above functions
int main()
{
/* Let us create following weighted graph
10
0--------1
| \
|
6| 5\ |15
|
\ |
2--------3
4
*/
int V = 4; // Number of vertices in graph
int E = 5; // Number of edges in graph
struct Graph* graph = createGraph(V, E);
// add edge 0-1
graph->edge[0].src = 0;
graph->edge[0].dest = 1;
graph->edge[0].weight = 10;
// add edge 0-2
graph->edge[1].src = 0;
graph->edge[1].dest = 2;
graph->edge[1].weight = 6;
// add edge 0-3
graph->edge[2].src = 0;
graph->edge[2].dest = 3;
graph->edge[2].weight = 5;
// add edge 1-3
graph->edge[3].src = 1;
graph->edge[3].dest = 3;

graph->edge[3].weight = 15;
// add edge 2-3
graph->edge[4].src = 2;
graph->edge[4].dest = 3;
graph->edge[4].weight = 4;
KruskalMST(graph);
return 0;
}

Java
// Java program for Kruskal's algorithm to find Minimum Spanning Tree
// of a given connected, undirected and weighted graph
import java.util.*;
import java.lang.*;
import java.io.*;
class Graph
{
// A class to represent a graph edge
class Edge implements Comparable<Edge>
{
int src, dest, weight;
// Comparator function used for sorting edges based on
// their weight
public int compareTo(Edge compareEdge)
{
return this.weight-compareEdge.weight;
}
};
// A class to represent a subset for union-find
class subset
{
int parent, rank;
};
int V, E;
// V-> no. of vertices & E->no.of edges
Edge edge[]; // collection of all edges
// Creates a graph with V vertices and E edges
Graph(int v, int e)
{
V = v;
E = e;
edge = new Edge[E];
for (int i=0; i<e; ++i)
edge[i] = new Edge();
}
// A utility function to find set of an element i
// (uses path compression technique)
int find(subset subsets[], int i)
{
// find root and make root as parent of i (path compression)
if (subsets[i].parent != i)
subsets[i].parent = find(subsets, subsets[i].parent);
return subsets[i].parent;
}
// A function that does union of two sets of x and y
// (uses union by rank)
void Union(subset subsets[], int x, int y)
{
int xroot = find(subsets, x);
int yroot = find(subsets, y);
// Attach smaller rank tree under root of high rank tree
// (Union by Rank)
if (subsets[xroot].rank < subsets[yroot].rank)
subsets[xroot].parent = yroot;
else if (subsets[xroot].rank > subsets[yroot].rank)
subsets[yroot].parent = xroot;
// If ranks are same, then make one as root and increment

// its rank by one

else
{
subsets[yroot].parent = xroot;
subsets[xroot].rank++;
}
}
// The main function to construct MST using Kruskal's algorithm
void KruskalMST()
{
Edge result[] = new Edge[V]; // Tnis will store the resultant MST
int e = 0; // An index variable, used for result[]
int i = 0; // An index variable, used for sorted edges
for (i=0; i<V; ++i)
result[i] = new Edge();
// Step 1: Sort all the edges in non-decreasing order of their
// weight. If we are not allowed to change the given graph, we
// can create a copy of array of edges
Arrays.sort(edge);
// Allocate memory for creating V ssubsets
subset subsets[] = new subset[V];
for(i=0; i<V; ++i)
subsets[i]=new subset();
// Create V subsets with single elements
for (int v = 0; v < V; ++v)
{
subsets[v].parent = v;
subsets[v].rank = 0;
}
i = 0; // Index used to pick next edge
// Number of edges to be taken is equal to V-1
while (e < V - 1)
{
// Step 2: Pick the smallest edge. And increment the index
// for next iteration
Edge next_edge = new Edge();
next_edge = edge[i++];
int x = find(subsets, next_edge.src);
int y = find(subsets, next_edge.dest);
// If including this edge does't cause cycle, include it
// in result and increment the index of result for next edge
if (x != y)
{
result[e++] = next_edge;
Union(subsets, x, y);
}
// Else discard the next_edge
}
// print the contents of result[] to display the built MST
System.out.println("Following are the edges in the constructed MST");
for (i = 0; i < e; ++i)
System.out.println(result[i].src+" -- "+result[i].dest+" == "+
result[i].weight);
}
// Driver Program
public static void main (String[] args)
{
/* Let us create following weighted graph
10
0--------1
| \
|
6| 5\ |15
|
\ |
2--------3
4
*/
int V = 4; // Number of vertices in graph
int E = 5; // Number of edges in graph
Graph graph = new Graph(V, E);
// add edge 0-1

graph.edge[0].src = 0;
graph.edge[0].dest = 1;
graph.edge[0].weight = 10;
// add edge 0-2
graph.edge[1].src = 0;
graph.edge[1].dest = 2;
graph.edge[1].weight = 6;
// add edge 0-3
graph.edge[2].src = 0;
graph.edge[2].dest = 3;
graph.edge[2].weight = 5;
// add edge 1-3
graph.edge[3].src = 1;
graph.edge[3].dest = 3;
graph.edge[3].weight = 15;
// add edge 2-3
graph.edge[4].src = 2;
graph.edge[4].dest = 3;
graph.edge[4].weight = 4;
graph.KruskalMST();
}
}
//This code is contributed by Aakash Hasija
Following
2 -- 3 ==
0 -- 3 ==
0 -- 1 ==

are the edges in the constructed MST

4
5
10

Time Complexity: O(ElogE) or O(ElogV). Sorting of edges takes O(ELogE) time. After sorting, we iterate through all edges and apply findunion algorithm. The find and union operations can take atmost O(LogV) time. So overall complexity is O(ELogE + ELogV) time. The value of E
can be atmost V^2, so O(LogV) are O(LogE) same. Therefore, overall time complexity is O(ElogE) or O(ElogV)
References:
http://www.ics.uci.edu/~eppstein/161/960206.html
http://en.wikipedia.org/wiki/Minimum_spanning_tree

Greedy Algorithms | Set 3 (Huffman Coding)

Huffman coding is a lossless data compression algorithm. The idea is to assign variable-legth codes to input characters, lengths of the assigned
codes are based on the frequencies of corresponding characters. The most frequent character gets the smallest code and the least frequent
character gets the largest code.
The variable-length codes assigned to input characters are Prefix Codes, means the codes (bit sequences) are assigned in such a way that the code
assigned to one character is not prefix of code assigned to any other character. This is how Huffman Coding makes sure that there is no ambiguity
when decoding the generated bit stream.
Let us understand prefix codes with a counter example. Let there be four characters a, b, c and d, and their corresponding variable length codes
be 00, 01, 0 and 1. This coding leads to ambiguity because code assigned to c is prefix of codes assigned to a and b. If the compressed bit stream
is 0001, the de-compressed output may be cccd or ccb or acd or ab.
See this for applications of Huffman Coding.
There are mainly two major parts in Huffman Coding
1) Build a Huffman Tree from input characters.
2) Traverse the Huffman Tree and assign codes to characters.
Steps to build Huffman Tree
Input is array of unique characters along with their frequency of occurrences and output is Huffman Tree.
1. Create a leaf node for each unique character and build a min heap of all leaf nodes (Min Heap is used as a priority queue. The value of
frequency field is used to compare two nodes in min heap. Initially, the least frequent character is at root)
2. Extract two nodes with the minimum frequency from the min heap.
3. Create a new internal node with frequency equal to the sum of the two nodes frequencies. Make the first extracted node as its left child and the
other extracted node as its right child. Add this node to the min heap.
4. Repeat steps#2 and #3 until the heap contains only one node. The remaining node is the root node and the tree is complete.
Let us understand the algorithm with an example:
character
a
b
c
d
e
f

Frequency
5
9
12
13
16
45

Step 1. Build a min heap that contains 6 nodes where each node represents root of a tree with single node.
Step 2 Extract two minimum frequency nodes from min heap. Add a new internal node with frequency 5 + 9 = 14.

Now min heap contains 5 nodes where 4 nodes are roots of trees with single element each, and one heap node is root of tree with 3 elements
character
c
d
Internal Node
e
f

Frequency
12
13
14
16
45

Step 3: Extract two minimum frequency nodes from heap. Add a new internal node with frequency 12 + 13 = 25

Now min heap contains 4 nodes where 2 nodes are roots of trees with single element each, and two heap nodes are root of tree with more than
one nodes.
character
Internal Node
e
Internal Node
f

Frequency
14
16
25
45

Step 4: Extract two minimum frequency nodes. Add a new internal node with frequency 14 + 16 = 30

Now min heap contains 3 nodes.

character
Internal Node
Internal Node
f

Frequency
25
30
45

Step 5: Extract two minimum frequency nodes. Add a new internal node with frequency 25 + 30 = 55

Now min heap contains 2 nodes.

character
Frequency
f
45
Internal Node
55

Step 6: Extract two minimum frequency nodes. Add a new internal node with frequency 45 + 55 = 100

Now min heap contains only one node.

character
Frequency
Internal Node
100

Since the heap contains only one node, the algorithm stops here.
Steps to print codes from Huffman Tree:
Traverse the tree formed starting from the root. Maintain an auxiliary array. While moving to the left child, write 0 to the array. While moving to the
right child, write 1 to the array. Print the array when a leaf node is encountered.

The codes are as follows:

character
f
c
d
a
b
e

code-word
0
100
101
1100
1101
111

// C program for Huffman Coding

#include <stdio.h>
#include <stdlib.h>

// This constant can be avoided by explicitly calculating height of Huffman Tree

#define MAX_TREE_HT 100
// A Huffman tree node
struct MinHeapNode
{
char data; // One of the input characters
unsigned freq; // Frequency of the character
struct MinHeapNode *left, *right; // Left and right child of this node
};
// A Min Heap: Collection of min heap (or Hufmman tree) nodes
struct MinHeap
{
unsigned size;
// Current size of min heap
unsigned capacity; // capacity of min heap
struct MinHeapNode **array; // Attay of minheap node pointers
};
// A utility function allocate a new min heap node with given character
// and frequency of the character
struct MinHeapNode* newNode(char data, unsigned freq)
{
struct MinHeapNode* temp =
(struct MinHeapNode*) malloc(sizeof(struct MinHeapNode));
temp->left = temp->right = NULL;
temp->data = data;
temp->freq = freq;
return temp;
}
// A utility function to create a min heap of given capacity
struct MinHeap* createMinHeap(unsigned capacity)
{
struct MinHeap* minHeap =
(struct MinHeap*) malloc(sizeof(struct MinHeap));
minHeap->size = 0; // current size is 0
minHeap->capacity = capacity;
minHeap->array =
(struct MinHeapNode**)malloc(minHeap->capacity * sizeof(struct MinHeapNode*));
return minHeap;
}
// A utility function to swap two min heap nodes
void swapMinHeapNode(struct MinHeapNode** a, struct MinHeapNode** b)
{
struct MinHeapNode* t = *a;
*a = *b;
*b = t;
}
// The standard minHeapify function.
void minHeapify(struct MinHeap* minHeap, int idx)
{
int smallest = idx;
int left = 2 * idx + 1;
int right = 2 * idx + 2;
if (left < minHeap->size &&
minHeap->array[left]->freq < minHeap->array[smallest]->freq)
smallest = left;
if (right < minHeap->size &&
minHeap->array[right]->freq < minHeap->array[smallest]->freq)
smallest = right;
if (smallest != idx)
{
swapMinHeapNode(&minHeap->array[smallest], &minHeap->array[idx]);
minHeapify(minHeap, smallest);
}
}
// A utility function to check if size of heap is 1 or not
int isSizeOne(struct MinHeap* minHeap)
{
return (minHeap->size == 1);
}
// A standard function to extract minimum value node from heap
struct MinHeapNode* extractMin(struct MinHeap* minHeap)

{
struct MinHeapNode* temp = minHeap->array[0];
minHeap->array[0] = minHeap->array[minHeap->size - 1];
--minHeap->size;
minHeapify(minHeap, 0);
return temp;
}
// A utility function to insert a new node to Min Heap
void insertMinHeap(struct MinHeap* minHeap, struct MinHeapNode* minHeapNode)
{
++minHeap->size;
int i = minHeap->size - 1;
while (i && minHeapNode->freq < minHeap->array[(i - 1)/2]->freq)
{
minHeap->array[i] = minHeap->array[(i - 1)/2];
i = (i - 1)/2;
}
minHeap->array[i] = minHeapNode;
}
// A standard funvtion to build min heap
void buildMinHeap(struct MinHeap* minHeap)
{
int n = minHeap->size - 1;
int i;
for (i = (n - 1) / 2; i >= 0; --i)
minHeapify(minHeap, i);
}
// A utility function to print an array of size n
void printArr(int arr[], int n)
{
int i;
for (i = 0; i < n; ++i)
printf("%d", arr[i]);
printf("\n");
}
// Utility function to check if this node is leaf
int isLeaf(struct MinHeapNode* root)
{
return !(root->left) && !(root->right) ;
}
// Creates a min heap of capacity equal to size and inserts all character of
// data[] in min heap. Initially size of min heap is equal to capacity
struct MinHeap* createAndBuildMinHeap(char data[], int freq[], int size)
{
struct MinHeap* minHeap = createMinHeap(size);
for (int i = 0; i < size; ++i)
minHeap->array[i] = newNode(data[i], freq[i]);
minHeap->size = size;
buildMinHeap(minHeap);
return minHeap;
}
// The main function that builds Huffman tree
struct MinHeapNode* buildHuffmanTree(char data[], int freq[], int size)
{
struct MinHeapNode *left, *right, *top;
// Step 1: Create a min heap of capacity equal to size. Initially, there are
// modes equal to size.
struct MinHeap* minHeap = createAndBuildMinHeap(data, freq, size);
// Iterate while size of heap doesn't become 1
while (!isSizeOne(minHeap))
{
// Step 2: Extract the two minimum freq items from min heap
left = extractMin(minHeap);
right = extractMin(minHeap);
// Step 3: Create a new internal node with frequency equal to the
// sum of the two nodes frequencies. Make the two extracted node as
// left and right children of this new node. Add this node to the min heap
// '$' is a special value for internal nodes, not used
top = newNode('$', left->freq + right->freq);
top->left = left;
top->right = right;
insertMinHeap(minHeap, top);

}
// Step 4: The remaining node is the root node and the tree is complete.
return extractMin(minHeap);
}
// Prints huffman codes from the root of Huffman Tree. It uses arr[] to
// store codes
void printCodes(struct MinHeapNode* root, int arr[], int top)
{
// Assign 0 to left edge and recur
if (root->left)
{
arr[top] = 0;
printCodes(root->left, arr, top + 1);
}
// Assign 1 to right edge and recur
if (root->right)
{
arr[top] = 1;
printCodes(root->right, arr, top + 1);
}
// If this is a leaf node, then it contains one of the input
// characters, print the character and its code from arr[]
if (isLeaf(root))
{
printf("%c: ", root->data);
printArr(arr, top);
}
}
// The main function that builds a Huffman Tree and print codes by traversing
// the built Huffman Tree
void HuffmanCodes(char data[], int freq[], int size)
{
// Construct Huffman Tree
struct MinHeapNode* root = buildHuffmanTree(data, freq, size);
// Print Huffman codes using the Huffman tree built above
int arr[MAX_TREE_HT], top = 0;
printCodes(root, arr, top);
}
// Driver program to test above functions
int main()
{
char arr[] = {'a', 'b', 'c', 'd', 'e', 'f'};
int freq[] = {5, 9, 12, 13, 16, 45};
int size = sizeof(arr)/sizeof(arr[0]);
HuffmanCodes(arr, freq, size);
return 0;
}
f:
c:
d:
a:
b:
e:

0
100
101
1100
1101
111

Time complexity: O(nlogn) where n is the number of unique characters. If there are n nodes, extractMin() is called 2*(n 1) times. extractMin()
takes O(logn) time as it calles minHeapify(). So, overall complexity is O(nlogn).
If the input array is sorted, there exists a linear time algorithm. We will soon be discussing in our next post.
Reference:
http://en.wikipedia.org/wiki/Huffman_coding

Greedy Algorithms | Set 4 (Efficient Huffman Coding for Sorted Input)

We recommend to read following post as a prerequisite for this.
Greedy Algorithms | Set 3 (Huffman Coding)
Time complexity of the algorithm discussed in above post is O(nLogn). If we know that the given array is sorted (by non-decreasing order of
frequency), we can generate Huffman codes in O(n) time. Following is a O(n) algorithm for sorted input.
1. Create two empty queues.
2. Create a leaf node for each unique character and Enqueue it to the first queue in non-decreasing order of frequency. Initially second queue is
empty.
3. Dequeue two nodes with the minimum frequency by examining the front of both queues. Repeat following steps two times
..a) If second queue is empty, dequeue from first queue.
..b) If first queue is empty, dequeue from second queue.
..c) Else, compare the front of two queues and dequeue the minimum.
4. Create a new internal node with frequency equal to the sum of the two nodes frequencies. Make the first Dequeued node as its left child and the
second Dequeued node as right child. Enqueue this node to second queue.
5. Repeat steps#3 and #4 until there is more than one node in the queues. The remaining node is the root node and the tree is complete.
// C Program for Efficient Huffman Coding for Sorted input
#include <stdio.h>
#include <stdlib.h>
// This constant can be avoided by explicitly calculating height of Huffman Tree
#define MAX_TREE_HT 100
// A node of huffman tree
struct QueueNode
{
char data;
unsigned freq;
struct QueueNode *left, *right;
};
// Structure for Queue: collection of Huffman Tree nodes (or QueueNodes)
struct Queue
{
int front, rear;
int capacity;
struct QueueNode **array;
};
// A utility function to create a new Queuenode
struct QueueNode* newNode(char data, unsigned freq)
{
struct QueueNode* temp =
(struct QueueNode*) malloc(sizeof(struct QueueNode));
temp->left = temp->right = NULL;
temp->data = data;
temp->freq = freq;
return temp;
}
// A utility function to create a Queue of given capacity
struct Queue* createQueue(int capacity)
{
struct Queue* queue = (struct Queue*) malloc(sizeof(struct Queue));
queue->front = queue->rear = -1;
queue->capacity = capacity;
queue->array =
(struct QueueNode**) malloc(queue->capacity * sizeof(struct QueueNode*));
return queue;
}
// A utility function to check if size of given queue is 1
int isSizeOne(struct Queue* queue)
{
return queue->front == queue->rear && queue->front != -1;
}
// A utility function to check if given queue is empty
int isEmpty(struct Queue* queue)

{
return queue->front == -1;
}
// A utility function to check if given queue is full
int isFull(struct Queue* queue)
{
return queue->rear == queue->capacity - 1;
}
// A utility function to add an item to queue
void enQueue(struct Queue* queue, struct QueueNode* item)
{
if (isFull(queue))
return;
queue->array[++queue->rear] = item;
if (queue->front == -1)
++queue->front;
}
// A utility function to remove an item from queue
struct QueueNode* deQueue(struct Queue* queue)
{
if (isEmpty(queue))
return NULL;
struct QueueNode* temp = queue->array[queue->front];
if (queue->front == queue->rear) // If there is only one item in queue
queue->front = queue->rear = -1;
else
++queue->front;
return temp;
}
// A utility function to get from of queue
struct QueueNode* getFront(struct Queue* queue)
{
if (isEmpty(queue))
return NULL;
return queue->array[queue->front];
}
/* A function to get minimum item from two queues */
struct QueueNode* findMin(struct Queue* firstQueue, struct Queue* secondQueue)
{
// Step 3.a: If second queue is empty, dequeue from first queue
if (isEmpty(firstQueue))
return deQueue(secondQueue);
// Step 3.b: If first queue is empty, dequeue from second queue
if (isEmpty(secondQueue))
return deQueue(firstQueue);
// Step 3.c: Else, compare the front of two queues and dequeue minimum
if (getFront(firstQueue)->freq < getFront(secondQueue)->freq)
return deQueue(firstQueue);
return deQueue(secondQueue);
}
// Utility function to check if this node is leaf
int isLeaf(struct QueueNode* root)
{
return !(root->left) && !(root->right) ;
}
// A utility function to print an array of size n
void printArr(int arr[], int n)
{
int i;
for (i = 0; i < n; ++i)
printf("%d", arr[i]);
printf("\n");
}
// The main function that builds Huffman tree
struct QueueNode* buildHuffmanTree(char data[], int freq[], int size)
{
struct QueueNode *left, *right, *top;
// Step 1: Create two empty queues
struct Queue* firstQueue = createQueue(size);

struct Queue* secondQueue = createQueue(size);

// Step 2:Create a leaf node for each unique character and Enqueue it to
// the first queue in non-decreasing order of frequency. Initially second
// queue is empty
for (int i = 0; i < size; ++i)
enQueue(firstQueue, newNode(data[i], freq[i]));
// Run while Queues contain more than one node. Finally, first queue will
// be empty and second queue will contain only one node
while (!(isEmpty(firstQueue) && isSizeOne(secondQueue)))
{
// Step 3: Dequeue two nodes with the minimum frequency by examining
// the front of both queues
left = findMin(firstQueue, secondQueue);
right = findMin(firstQueue, secondQueue);
// Step 4: Create a new internal node with frequency equal to the sum
// of the two nodes frequencies. Enqueue this node to second queue.
top = newNode('$' , left->freq + right->freq);
top->left = left;
top->right = right;
enQueue(secondQueue, top);
}
return deQueue(secondQueue);
}
// Prints huffman codes from the root of Huffman Tree. It uses arr[] to
// store codes
void printCodes(struct QueueNode* root, int arr[], int top)
{
// Assign 0 to left edge and recur
if (root->left)
{
arr[top] = 0;
printCodes(root->left, arr, top + 1);
}
// Assign 1 to right edge and recur
if (root->right)
{
arr[top] = 1;
printCodes(root->right, arr, top + 1);
}
// If this is a leaf node, then it contains one of the input
// characters, print the character and its code from arr[]
if (isLeaf(root))
{
printf("%c: ", root->data);
printArr(arr, top);
}
}
// The main function that builds a Huffman Tree and print codes by traversing
// the built Huffman Tree
void HuffmanCodes(char data[], int freq[], int size)
{
// Construct Huffman Tree
struct QueueNode* root = buildHuffmanTree(data, freq, size);
// Print Huffman codes using the Huffman tree built above
int arr[MAX_TREE_HT], top = 0;
printCodes(root, arr, top);
}
// Driver program to test above functions
int main()
{
char arr[] = {'a', 'b', 'c', 'd', 'e', 'f'};
int freq[] = {5, 9, 12, 13, 16, 45};
int size = sizeof(arr)/sizeof(arr[0]);
HuffmanCodes(arr, freq, size);
return 0;
}

Output:
f: 0
c: 100

d:
a:
b:
e:

101
1100
1101
111

Time complexity: O(n)

If the input is not sorted, it need to be sorted first before it can be processed by the above algorithm. Sorting can be done using heap-sort or
merge-sort both of which run in Theta(nlogn). So, the overall time complexity becomes O(nlogn) for unsorted input.
Reference:
http://en.wikipedia.org/wiki/Huffman_coding

Greedy Algorithms | Set 5 (Prims Minimum Spanning Tree (MST))

We have discussedKruskals algorithm for Minimum Spanning Tree. Like Kruskals algorithm, Prims algorithm is also aGreedy algorithm. It starts
with an empty spanning tree. The idea is to maintain two sets of vertices. The first set contains the vertices already included in the MST, the other
set contains the vertices not yet included. At every step, it considers all the edges that connect the two sets, and picks the minimum weight edge
from these edges. After picking the edge, it moves the other endpoint of the edge to the set containing MST.
A group of edges that connects two set of vertices in a graph is calledcut in graph theory.So, at every step of Prims algorithm, we find a cut (of
two sets, one contains the vertices already included in MST and other contains rest of the verices), pick the minimum weight edge from
the cut and include this vertex to MST Set (the set that contains already included vertices).
How does Prims Algorithm Work? The idea behind Prims algorithm is simple, a spanning tree means all vertices must be connected. So the two
disjoint subsets (discussed above) of vertices must be connected to make aSpanningTree. And they must be connected with the minimum weight
edge to make it aMinimumSpanning Tree.
Algorithm
1) Create a set mstSet that keeps track of vertices already included in MST.
2) Assign a key value to all vertices in the input graph. Initialize all key values as INFINITE. Assign key value as 0 for the first vertex so that it is
picked first.
3) While mstSet doesnt include all vertices
.a) Pick a vertex u which is not there in mstSet and has minimum key value.
.b) Include u to mstSet.
.c) Update key value of all adjacent vertices of u. To update the key values, iterate through all adjacent vertices. For every adjacent vertex v, if
weight of edge u-v is less than the previous key value of v, update the key value as weight of u-v
The idea of using key values is to pick the minimum weight edge from cut. The key values are used only for vertices which are not yet included in
MST, the key value for these vertices indicate the minimum weight edges connecting them to the set of vertices included in MST.
Let us understand with the following example:

The set mstSet is initially empty and keys assigned to vertices are {0, INF, INF, INF, INF, INF, INF, INF} where INF indicates infinite. Now
pick the vertex with minimum key value. The vertex 0 is picked, include it in mstSet. So mstSet becomes {0}. After including to mstSet, update
key values of adjacent vertices. Adjacent vertices of 0 are 1 and 7. The key values of 1 and 7 are updated as 4 and 8. Following subgraph shows
vertices and their key values, only the vertices with finite key values are shown. The vertices included in MST are shown in green color.

Pick the vertex with minimum key value and not already included in MST (not in mstSET). The vertex 1 is picked and added to mstSet. So mstSet
now becomes {0, 1}. Update the key values of adjacent vertices of 1. The key value of vertex 2 becomes 8.

Pick the vertex with minimum key value and not already included in MST (not in mstSET). We can either pick vertex 7 or vertex 2, let vertex 7 is
picked. So mstSet now becomes {0, 1, 7}. Update the key values of adjacent vertices of 7. The key value of vertex 6 and 8 becomes finite (7
and 1 respectively).

Pick the vertex with minimum key value and not already included in MST (not in mstSET). Vertex 6 is picked. So mstSet now becomes {0, 1, 7,
6}. Update the key values of adjacent vertices of 6. The key value of vertex 5 and 8 are updated.

We repeat the above steps until mstSet includes all vertices of given graph. Finally, we get the following graph.

How to implement the above algorithm?

We use a boolean array mstSet[] to represent the set of vertices included in MST. If a value mstSet[v] is true, then vertex v is included in MST,
otherwise not. Array key[] is used to store key values of all vertices. Another array parent[] to store indexes of parent nodes in MST. The parent
array is the output array which is used to show the constructed MST.

C/C++
// A C / C++ program for Prim's Minimum Spanning Tree (MST) algorithm.
// The program is for adjacency matrix representation of the graph
#include <stdio.h>
#include <limits.h>
// Number of vertices in the graph
#define V 5
// A utility function to find the vertex with minimum key value, from
// the set of vertices not yet included in MST
int minKey(int key[], bool mstSet[])
{
// Initialize min value
int min = INT_MAX, min_index;
for (int v = 0; v < V; v++)
if (mstSet[v] == false && key[v] < min)
min = key[v], min_index = v;
return min_index;
}
// A utility function to print the constructed MST stored in parent[]
int printMST(int parent[], int n, int graph[V][V])
{
printf("Edge Weight\n");
for (int i = 1; i < V; i++)
printf("%d - %d
%d \n", parent[i], i, graph[i][parent[i]]);
}
// Function to construct and print MST for a graph represented using adjacency
// matrix representation
void primMST(int graph[V][V])
{
int parent[V]; // Array to store constructed MST
int key[V]; // Key values used to pick minimum weight edge in cut
bool mstSet[V]; // To represent set of vertices not yet included in MST

// Initialize all keys as INFINITE

for (int i = 0; i < V; i++)
key[i] = INT_MAX, mstSet[i] = false;
// Always include first 1st vertex in MST.
key[0] = 0;
// Make key 0 so that this vertex is picked as first vertex
parent[0] = -1; // First node is always root of MST
// The MST will have V vertices
for (int count = 0; count < V-1; count++)
{
// Pick thd minimum key vertex from the set of vertices
// not yet included in MST
int u = minKey(key, mstSet);
// Add the picked vertex to the MST Set
mstSet[u] = true;
// Update key value and parent index of the adjacent vertices of
// the picked vertex. Consider only those vertices which are not yet
// included in MST
for (int v = 0; v < V; v++)
// graph[u][v] is non zero only for adjacent vertices of m
// mstSet[v] is false for vertices not yet included in MST
// Update the key only if graph[u][v] is smaller than key[v]
if (graph[u][v] && mstSet[v] == false && graph[u][v] < key[v])
parent[v] = u, key[v] = graph[u][v];
}
// print the constructed MST
printMST(parent, V, graph);
}
// driver program to test above function
int main()
{
/* Let us create the following graph
2
3
(0)--(1)--(2)
| / \ |
6| 8/ \5 |7
| /
\ |
(3)-------(4)
9
*/
int graph[V][V] = {{0, 2, 0, 6, 0},
{2, 0, 3, 8, 5},
{0, 3, 0, 0, 7},
{6, 8, 0, 0, 9},
{0, 5, 7, 9, 0},
};
// Print the solution
primMST(graph);
return 0;
}

Java
// A Java program for Prim's Minimum Spanning Tree (MST) algorithm.
// The program is for adjacency matrix representation of the graph
import java.util.*;
import java.lang.*;
import java.io.*;
class MST
{
// Number of vertices in the graph
private static final int V=5;
// A utility function to find the vertex with minimum key
// value, from the set of vertices not yet included in MST
int minKey(int key[], Boolean mstSet[])
{
// Initialize min value
int min = Integer.MAX_VALUE, min_index=-1;

for (int v = 0; v < V; v++)

if (mstSet[v] == false && key[v] < min)
{
min = key[v];
min_index = v;
}
return min_index;
}
// A utility function to print the constructed MST stored in
// parent[]
void printMST(int parent[], int n, int graph[][])
{
System.out.println("Edge Weight");
for (int i = 1; i < V; i++)
System.out.println(parent[i]+" - "+ i+"
"+
graph[i][parent[i]]);
}
// Function to construct and print MST for a graph represented
// using adjacency matrix representation
void primMST(int graph[][])
{
// Array to store constructed MST
int parent[] = new int[V];
// Key values used to pick minimum weight edge in cut
int key[] = new int [V];
// To represent set of vertices not yet included in MST
Boolean mstSet[] = new Boolean[V];
// Initialize all keys as INFINITE
for (int i = 0; i < V; i++)
{
key[i] = Integer.MAX_VALUE;
mstSet[i] = false;
}
// Always include first 1st vertex in MST.
key[0] = 0;
// Make key 0 so that this vertex is
// picked as first vertex
parent[0] = -1; // First node is always root of MST
// The MST will have V vertices
for (int count = 0; count < V-1; count++)
{
// Pick thd minimum key vertex from the set of vertices
// not yet included in MST
int u = minKey(key, mstSet);
// Add the picked vertex to the MST Set
mstSet[u] = true;
// Update key value and parent index of the adjacent
// vertices of the picked vertex. Consider only those
// vertices which are not yet included in MST
for (int v = 0; v < V; v++)
//
//
//
if

graph[u][v] is non zero only for adjacent vertices of m

mstSet[v] is false for vertices not yet included in MST
Update the key only if graph[u][v] is smaller than key[v]
(graph[u][v]!=0 && mstSet[v] == false &&
graph[u][v] < key[v])

{
parent[v] = u;
key[v] = graph[u][v];
}
}
// print the constructed MST
printMST(parent, V, graph);
}
public static void main (String[] args)
{
/* Let us create the following graph
2
3
(0)--(1)--(2)

|
/ \ |
6| 8/ \5 |7
| /
\ |
(3)-------(4)
9
*/
MST t = new MST();
int graph[][] = new int[][] {{0, 2, 0, 6, 0},
{2, 0, 3, 8, 5},
{0, 3, 0, 0, 7},
{6, 8, 0, 0, 9},
{0, 5, 7, 9, 0},
};
// Print the solution
t.primMST(graph);
}
}
// This code is contributed by Aakash Hasija

Edge Weight
0 - 1
2
1 - 2
3
0 - 3
6
1 - 4
5

Time Complexity of the above program is O(V^2). If the input graph is represented using adjacency list, then the time complexity of Prims
algorithm can be reduced to O(E log V) with the help of binary heap. Please see Prims MST for Adjacency List Representation for more details.

Greedy Algorithms | Set 6 (Prims MST for Adjacency List Representation)

We recommend to read following two posts as a prerequisite of this post.
1. Greedy Algorithms | Set 5 (Prims Minimum Spanning Tree (MST))
2. Graph and its representations
We have discussed Prims algorithm and its implementation for adjacency matrix representation of graphs. The time complexity for the matrix
representation is O(V^2). In this post, O(ELogV) algorithm for adjacency list representation is discussed.
As discussed in the previous post, in Prims algorithm, two sets are maintained, one set contains list of vertices already included in MST, other set
contains vertices not yet included. With adjacency list representation, all vertices of a graph can be traversed in O(V+E) time using BFS. The idea
is to traverse all vertices of graph using BFS and use a Min Heap to store the vertices not yet included in MST. Min Heap is used as a priority
queue to get the minimum weight edge from the cut. Min Heap is used as time complexity of operations like extracting minimum element and
decreasing key value is O(LogV) in Min Heap.
Following are the detailed steps.
1) Create a Min Heap of size V where V is the number of vertices in the given graph. Every node of min heap contains vertex number and key
value of the vertex.
2) Initialize Min Heap with first vertex as root (the key value assigned to first vertex is 0). The key value assigned to all other vertices is INF
(infinite).
3) While Min Heap is not empty, do following
..a) Extract the min value node from Min Heap. Let the extracted vertex be u.
..b) For every adjacent vertex v of u, check if v is in Min Heap (not yet included in MST). If v is in Min Heap and its key value is more than weight
of u-v, then update the key value of v as weight of u-v.
Let us understand the above algorithm with the following example:

Initially, key value of first vertex is 0 and INF (infinite) for all other vertices. So vertex 0 is extracted from Min Heap and key values of vertices
adjacent to 0 (1 and 7) are updated. Min Heap contains all vertices except vertex 0.
The vertices in green color are the vertices included in MST.

Since key value of vertex 1 is minimum among all nodes in Min Heap, it is extracted from Min Heap and key values of vertices adjacent to 1 are
updated (Key is updated if the a vertex is not in Min Heap and previous key value is greater than the weight of edge from 1 to the adjacent). Min
Heap contains all vertices except vertex 0 and 1.

Since key value of vertex 7 is minimum among all nodes in Min Heap, it is extracted from Min Heap and key values of vertices adjacent to 7 are
updated (Key is updated if the a vertex is not in Min Heap and previous key value is greater than the weight of edge from 7 to the adjacent). Min
Heap contains all vertices except vertex 0, 1 and 7.

Since key value of vertex 6 is minimum among all nodes in Min Heap, it is extracted from Min Heap and key values of vertices adjacent to 6 are
updated (Key is updated if the a vertex is not in Min Heap and previous key value is greater than the weight of edge from 6 to the adjacent). Min
Heap contains all vertices except vertex 0, 1, 7 and 6.

The above steps are repeated for rest of the nodes in Min Heap till Min Heap becomes empty

// C / C++ program for Prim's MST for adjacency list representation of graph
#include <stdio.h>
#include <stdlib.h>
#include <limits.h>
// A structure to represent a node in adjacency list
struct AdjListNode
{
int dest;
int weight;
struct AdjListNode* next;
};
// A structure to represent an adjacency liat
struct AdjList
{
struct AdjListNode *head; // pointer to head node of list
};
// A structure to represent a graph. A graph is an array of adjacency lists.
// Size of array will be V (number of vertices in graph)
struct Graph
{
int V;
struct AdjList* array;
};
// A utility function to create a new adjacency list node
struct AdjListNode* newAdjListNode(int dest, int weight)
{
struct AdjListNode* newNode =
(struct AdjListNode*) malloc(sizeof(struct AdjListNode));
newNode->dest = dest;
newNode->weight = weight;
newNode->next = NULL;
return newNode;
}
// A utility function that creates a graph of V vertices
struct Graph* createGraph(int V)
{
struct Graph* graph = (struct Graph*) malloc(sizeof(struct Graph));
graph->V = V;
// Create an array of adjacency lists. Size of array will be V
graph->array = (struct AdjList*) malloc(V * sizeof(struct AdjList));
// Initialize each adjacency list as empty by making head as NULL
for (int i = 0; i < V; ++i)
graph->array[i].head = NULL;
return graph;
}
// Adds an edge to an undirected graph
void addEdge(struct Graph* graph, int src, int dest, int weight)
{
// Add an edge from src to dest. A new node is added to the adjacency
// list of src. The node is added at the begining
struct AdjListNode* newNode = newAdjListNode(dest, weight);
newNode->next = graph->array[src].head;

graph->array[src].head = newNode;
// Since graph is undirected, add an edge from dest to src also
newNode = newAdjListNode(src, weight);
newNode->next = graph->array[dest].head;
graph->array[dest].head = newNode;
}
// Structure to represent a min heap node
struct MinHeapNode
{
int v;
int key;
};
// Structure to represent a min heap
struct MinHeap
{
int size;
// Number of heap nodes present currently
int capacity; // Capacity of min heap
int *pos;
// This is needed for decreaseKey()
struct MinHeapNode **array;
};
// A utility function to create a new Min Heap Node
struct MinHeapNode* newMinHeapNode(int v, int key)
{
struct MinHeapNode* minHeapNode =
(struct MinHeapNode*) malloc(sizeof(struct MinHeapNode));
minHeapNode->v = v;
minHeapNode->key = key;
return minHeapNode;
}
// A utilit function to create a Min Heap
struct MinHeap* createMinHeap(int capacity)
{
struct MinHeap* minHeap =
(struct MinHeap*) malloc(sizeof(struct MinHeap));
minHeap->pos = (int *)malloc(capacity * sizeof(int));
minHeap->size = 0;
minHeap->capacity = capacity;
minHeap->array =
(struct MinHeapNode**) malloc(capacity * sizeof(struct MinHeapNode*));
return minHeap;
}
// A utility function to swap two nodes of min heap. Needed for min heapify
void swapMinHeapNode(struct MinHeapNode** a, struct MinHeapNode** b)
{
struct MinHeapNode* t = *a;
*a = *b;
*b = t;
}
// A standard function to heapify at given idx
// This function also updates position of nodes when they are swapped.
// Position is needed for decreaseKey()
void minHeapify(struct MinHeap* minHeap, int idx)
{
int smallest, left, right;
smallest = idx;
left = 2 * idx + 1;
right = 2 * idx + 2;
if (left < minHeap->size &&
minHeap->array[left]->key < minHeap->array[smallest]->key )
smallest = left;
if (right < minHeap->size &&
minHeap->array[right]->key < minHeap->array[smallest]->key )
smallest = right;
if (smallest != idx)
{
// The nodes to be swapped in min heap
MinHeapNode *smallestNode = minHeap->array[smallest];
MinHeapNode *idxNode = minHeap->array[idx];
// Swap positions
minHeap->pos[smallestNode->v] = idx;

minHeap->pos[idxNode->v] = smallest;
// Swap nodes
swapMinHeapNode(&minHeap->array[smallest], &minHeap->array[idx]);
minHeapify(minHeap, smallest);
}
}
// A utility function to check if the given minHeap is ampty or not
int isEmpty(struct MinHeap* minHeap)
{
return minHeap->size == 0;
}
// Standard function to extract minimum node from heap
struct MinHeapNode* extractMin(struct MinHeap* minHeap)
{
if (isEmpty(minHeap))
return NULL;
// Store the root node
struct MinHeapNode* root = minHeap->array[0];
// Replace root node with last node
struct MinHeapNode* lastNode = minHeap->array[minHeap->size - 1];
minHeap->array[0] = lastNode;
// Update position of last node
minHeap->pos[root->v] = minHeap->size-1;
minHeap->pos[lastNode->v] = 0;
// Reduce heap size and heapify root
--minHeap->size;
minHeapify(minHeap, 0);
return root;
}
// Function to decreasy key value of a given vertex v. This function
// uses pos[] of min heap to get the current index of node in min heap
void decreaseKey(struct MinHeap* minHeap, int v, int key)
{
// Get the index of v in heap array
int i = minHeap->pos[v];
// Get the node and update its key value
minHeap->array[i]->key = key;
// Travel up while the complete tree is not hepified.
// This is a O(Logn) loop
while (i && minHeap->array[i]->key < minHeap->array[(i - 1) / 2]->key)
{
// Swap this node with its parent
minHeap->pos[minHeap->array[i]->v] = (i-1)/2;
minHeap->pos[minHeap->array[(i-1)/2]->v] = i;
swapMinHeapNode(&minHeap->array[i], &minHeap->array[(i - 1) / 2]);
// move to parent index
i = (i - 1) / 2;
}
}
// A utility function to check if a given vertex
// 'v' is in min heap or not
bool isInMinHeap(struct MinHeap *minHeap, int v)
{
if (minHeap->pos[v] < minHeap->size)
return true;
return false;
}
// A utility function used to print the constructed MST
void printArr(int arr[], int n)
{
for (int i = 1; i < n; ++i)
printf("%d - %d\n", arr[i], i);
}
// The main function that constructs Minimum Spanning Tree (MST)
// using Prim's algorithm

void PrimMST(struct Graph* graph)

{
int V = graph->V;// Get the number of vertices in graph
int parent[V]; // Array to store constructed MST
int key[V];
// Key values used to pick minimum weight edge in cut
// minHeap represents set E
struct MinHeap* minHeap = createMinHeap(V);
// Initialize min heap with all vertices. Key value of
// all vertices (except 0th vertex) is initially infinite
for (int v = 1; v < V; ++v)
{
parent[v] = -1;
key[v] = INT_MAX;
minHeap->array[v] = newMinHeapNode(v, key[v]);
minHeap->pos[v] = v;
}
// Make key value of 0th vertex as 0 so that it
// is extracted first
key[0] = 0;
minHeap->array[0] = newMinHeapNode(0, key[0]);
minHeap->pos[0] = 0;
// Initially size of min heap is equal to V
minHeap->size = V;
// In the followin loop, min heap contains all nodes
// not yet added to MST.
while (!isEmpty(minHeap))
{
// Extract the vertex with minimum key value
struct MinHeapNode* minHeapNode = extractMin(minHeap);
int u = minHeapNode->v; // Store the extracted vertex number
// Traverse through all adjacent vertices of u (the extracted
// vertex) and update their key values
struct AdjListNode* pCrawl = graph->array[u].head;
while (pCrawl != NULL)
{
int v = pCrawl->dest;
//
//
//
if
{

If v is not yet included in MST and weight of u-v is

less than key value of v, then update key value and
parent of v
(isInMinHeap(minHeap, v) && pCrawl->weight < key[v])
key[v] = pCrawl->weight;
parent[v] = u;
decreaseKey(minHeap, v, key[v]);

}
pCrawl = pCrawl->next;
}
}
// print edges of MST
printArr(parent, V);
}
// Driver program to test above functions
int main()
{
// Let us create the graph given in above fugure
int V = 9;
struct Graph* graph = createGraph(V);
addEdge(graph, 0, 1, 4);
addEdge(graph, 0, 7, 8);
addEdge(graph, 1, 2, 8);
addEdge(graph, 1, 7, 11);
addEdge(graph, 2, 3, 7);
addEdge(graph, 2, 8, 2);
addEdge(graph, 2, 5, 4);
addEdge(graph, 3, 4, 9);
addEdge(graph, 3, 5, 14);
addEdge(graph, 4, 5, 10);
addEdge(graph, 5, 6, 2);
addEdge(graph, 6, 7, 1);
addEdge(graph, 6, 8, 6);
addEdge(graph, 7, 8, 7);

PrimMST(graph);
return 0;
}

Output:
0
5
2
3
6
7
0
2

1
2
3
4
5
6
7
8

Time Complexity: The time complexity of the above code/algorithm looks O(V^2) as there are two nested while loops. If we take a closer look,
we can observe that the statements in inner loop are executed O(V+E) times (similar to BFS). The inner loop has decreaseKey() operation which
takes O(LogV) time. So overall time complexity is O(E+V)*O(LogV) which is O((E+V)*LogV) = O(ELogV) (For a connected graph, V =
O(E))
References:
Introduction to Algorithms by Clifford Stein, Thomas H. Cormen, Charles E. Leiserson, Ronald L.
http://en.wikipedia.org/wiki/Prims_algorithm

Greedy Algorithms | Set 7 (Dijkstras shortest path algorithm)

Given a graph and a source vertex in graph, find shortest paths from source to all vertices in the given graph.
Dijkstras algorithm is very similar to Prims algorithm for minimum spanning tree. Like Prims MST, we generate a SPT (shortest path tree) with
given source as root. We maintain two sets, one set contains vertices included in shortest path tree, other set includes vertices not yet included in
shortest path tree. At every step of the algorithm, we find a vertex which is in the other set (set of not yet included) and has minimum distance from
source.
Below are the detailed steps used in Dijkstras algorithm to find the shortest path from a single source vertex to all other vertices in the given graph.
Algorithm
1) Create a set sptSet (shortest path tree set) that keeps track of vertices included in shortest path tree, i.e., whose minimum distance from source
is calculated and finalized. Initially, this set is empty.
2) Assign a distance value to all vertices in the input graph. Initialize all distance values as INFINITE. Assign distance value as 0 for the source
vertex so that it is picked first.
3) While sptSet doesnt include all vertices
.a) Pick a vertex u which is not there in sptSetand has minimum distance value.
.b) Include u to sptSet.
.c) Update distance value of all adjacent vertices of u. To update the distance values, iterate through all adjacent vertices. For every adjacent
vertex v, if sum of distance value of u (from source) and weight of edge u-v, is less than the distance value of v, then update the distance value of v.
Let us understand with the following example:

The set sptSetis initially empty and distances assigned to vertices are {0, INF, INF, INF, INF, INF, INF, INF} where INF indicates infinite.
Now pick the vertex with minimum distance value. The vertex 0 is picked, include it in sptSet. So sptSet becomes {0}. After including 0 to
sptSet, update distance values of its adjacent vertices. Adjacent vertices of 0 are 1 and 7. The distance values of 1 and 7 are updated as 4 and 8.
Following subgraph shows vertices and their distance values, only the vertices with finite distance values are shown. The vertices included in SPT
are shown in green color.

Pick the vertex with minimum distance value and not already included in SPT (not in sptSET). The vertex 1 is picked and added to sptSet. So
sptSet now becomes {0, 1}. Update the distance values of adjacent vertices of 1. The distance value of vertex 2 becomes 12.

Pick the vertex with minimum distance value and not already included in SPT (not in sptSET). Vertex 7 is picked. So sptSet now becomes {0, 1,
7}. Update the distance values of adjacent vertices of 7. The distance value of vertex 6 and 8 becomes finite (15 and 9 respectively).

Pick the vertex with minimum distance value and not already included in SPT (not in sptSET). Vertex 6 is picked. So sptSet now becomes {0, 1,
7, 6}. Update the distance values of adjacent vertices of 6. The distance value of vertex 5 and 8 are updated.

We repeat the above steps until sptSet doesnt include all vertices of given graph. Finally, we get the following Shortest Path Tree (SPT).

How to implement the above algorithm?

We use a boolean array sptSet[] to represent the set of vertices included in SPT. If a value sptSet[v] is true, then vertex v is included in SPT,
otherwise not. Array dist[] is used to store shortest distance values of all vertices.

C/C++
// A C / C++ program for Dijkstra's single source shortest path algorithm.
// The program is for adjacency matrix representation of the graph
#include <stdio.h>
#include <limits.h>
// Number of vertices in the graph
#define V 9
// A utility function to find the vertex with minimum distance value, from
// the set of vertices not yet included in shortest path tree
int minDistance(int dist[], bool sptSet[])
{
// Initialize min value
int min = INT_MAX, min_index;
for (int v = 0; v < V; v++)
if (sptSet[v] == false && dist[v] <= min)
min = dist[v], min_index = v;
return min_index;
}
// A utility function to print the constructed distance array
int printSolution(int dist[], int n)
{
printf("Vertex Distance from Source\n");
for (int i = 0; i < V; i++)
printf("%d \t\t %d\n", i, dist[i]);
}
// Funtion that implements Dijkstra's single source shortest path algorithm
// for a graph represented using adjacency matrix representation
void dijkstra(int graph[V][V], int src)
{
int dist[V];
// The output array. dist[i] will hold the shortest
// distance from src to i
bool sptSet[V]; // sptSet[i] will true if vertex i is included in shortest
// path tree or shortest distance from src to i is finalized
// Initialize all distances as INFINITE and stpSet[] as false
for (int i = 0; i < V; i++)
dist[i] = INT_MAX, sptSet[i] = false;
// Distance of source vertex from itself is always 0
dist[src] = 0;
// Find shortest path for all vertices
for (int count = 0; count < V-1; count++)
{
// Pick the minimum distance vertex from the set of vertices not

// yet processed. u is always equal to src in first iteration.

int u = minDistance(dist, sptSet);
// Mark the picked vertex as processed
sptSet[u] = true;
// Update dist value of the adjacent vertices of the picked vertex.
for (int v = 0; v < V; v++)
//
//
//
if

Update dist[v] only if is not in sptSet, there is an edge from

u to v, and total weight of path from src to v through u is
smaller than current value of dist[v]
(!sptSet[v] && graph[u][v] && dist[u] != INT_MAX
&& dist[u]+graph[u][v] < dist[v])
dist[v] = dist[u] + graph[u][v];

}
// print the constructed distance array
printSolution(dist, V);
}
// driver program to test above function
int main()
{
/* Let us create the example graph discussed above */
int graph[V][V] = {{0, 4, 0, 0, 0, 0, 0, 8, 0},
{4, 0, 8, 0, 0, 0, 0, 11, 0},
{0, 8, 0, 7, 0, 4, 0, 0, 2},
{0, 0, 7, 0, 9, 14, 0, 0, 0},
{0, 0, 0, 9, 0, 10, 0, 0, 0},
{0, 0, 4, 0, 10, 0, 2, 0, 0},
{0, 0, 0, 14, 0, 2, 0, 1, 6},
{8, 11, 0, 0, 0, 0, 1, 0, 7},
{0, 0, 2, 0, 0, 0, 6, 7, 0}
};
dijkstra(graph, 0);
return 0;
}

Java
// A Java program for Dijkstra's single source shortest path algorithm.
// The program is for adjacency matrix representation of the graph
import java.util.*;
import java.lang.*;
import java.io.*;
class ShortestPath
{
// A utility function to find the vertex with minimum distance value,
// from the set of vertices not yet included in shortest path tree
static final int V=9;
int minDistance(int dist[], Boolean sptSet[])
{
// Initialize min value
int min = Integer.MAX_VALUE, min_index=-1;
for (int v = 0; v < V; v++)
if (sptSet[v] == false && dist[v] <= min)
{
min = dist[v];
min_index = v;
}
return min_index;
}
// A utility function to print the constructed distance array
void printSolution(int dist[], int n)
{
System.out.println("Vertex Distance from Source");
for (int i = 0; i < V; i++)
System.out.println(i+" \t\t "+dist[i]);
}
// Funtion that implements Dijkstra's single source shortest path
// algorithm for a graph represented using adjacency matrix
// representation

void dijkstra(int graph[][], int src)

{
int dist[] = new int[V]; // The output array. dist[i] will hold
// the shortest distance from src to i
// sptSet[i] will true if vertex i is included in shortest
// path tree or shortest distance from src to i is finalized
Boolean sptSet[] = new Boolean[V];
// Initialize all distances as INFINITE and stpSet[] as false
for (int i = 0; i < V; i++)
{
dist[i] = Integer.MAX_VALUE;
sptSet[i] = false;
}
// Distance of source vertex from itself is always 0
dist[src] = 0;
// Find shortest path for all vertices
for (int count = 0; count < V-1; count++)
{
// Pick the minimum distance vertex from the set of vertices
// not yet processed. u is always equal to src in first
// iteration.
int u = minDistance(dist, sptSet);
// Mark the picked vertex as processed
sptSet[u] = true;
// Update dist value of the adjacent vertices of the
// picked vertex.
for (int v = 0; v < V; v++)
//
//
//
if

Update dist[v] only if is not in sptSet, there is an

edge from u to v, and total weight of path from src to
v through u is smaller than current value of dist[v]
(!sptSet[v] && graph[u][v]!=0 &&
dist[u] != Integer.MAX_VALUE &&
dist[u]+graph[u][v] < dist[v])
dist[v] = dist[u] + graph[u][v];

}
// print the constructed distance array
printSolution(dist, V);
}
// Driver method
public static void main (String[] args)
{
/* Let us create the example graph discussed above */
int graph[][] = new int[][]{{0, 4, 0, 0, 0, 0, 0, 8, 0},
{4, 0, 8, 0, 0, 0, 0, 11, 0},
{0, 8, 0, 7, 0, 4, 0, 0, 2},
{0, 0, 7, 0, 9, 14, 0, 0, 0},
{0, 0, 0, 9, 0, 10, 0, 0, 0},
{0, 0, 4, 0, 10, 0, 2, 0, 0},
{0, 0, 0, 14, 0, 2, 0, 1, 6},
{8, 11, 0, 0, 0, 0, 1, 0, 7},
{0, 0, 2, 0, 0, 0, 6, 7, 0}
};
ShortestPath t = new ShortestPath();
t.dijkstra(graph, 0);
}
}
//This code is contributed by Aakash Hasija

Vertex
0
1
2
3
4
5
6
7
8

Notes:

Distance from Source

0
4
12
19
21
11
9
8
14

1) The code calculates shortest distance, but doesnt calculate the path information. We can create a parent array, update the parent array when
distance is updated (like prims implementation) and use it show the shortest path from source to different vertices.
2) The code is for undirected graph, same dijekstra function can be used for directed graphs also.
3) The code finds shortest distances from source to all vertices. If we are interested only in shortest distance from source to a single target, we can
break the for loop when the picked minimum distance vertex is equal to target (Step 3.a of algorithm).
4) Time Complexity of the implementation is O(V^2). If the input graph is represented using adjacency list, it can be reduced to O(E log V) with
the help of binary heap. Please see
Dijkstras Algorithm for Adjacency List Representation for more details.
5) Dijkstras algorithm doesnt work for graphs with negative weight edges. For graphs with negative weight edges, BellmanFord algorithm can be
used, we will soon be discussing it as a separate post.
Dijkstras Algorithm for Adjacency List Representation

Greedy Algorithms | Set 8 (Dijkstras Algorithm for Adjacency List Representation)

We recommend to read following two posts as a prerequisite of this post.
1. Greedy Algorithms | Set 7 (Dijkstras shortest path algorithm)
2.Graph and its representations
We have discussed Dijkstras algorithm and its implementation for adjacency matrix representation of graphs. The time complexity for the matrix
representation is O(V^2). In this post, O(ELogV) algorithm for adjacency list representation is discussed.
As discussed in the previous post, in Dijkstras algorithm, two sets are maintained, one set contains list of vertices already included in SPT (Shortest
Path Tree), other set contains vertices not yet included. With adjacency list representation, all vertices of a graph can be traversed in O(V+E) time
usingBFS. The idea is to traverse all vertices of graph usingBFSand use a Min Heap to store the vertices not yet included in SPT (or the vertices
for which shortest distance is not finalized yet). Min Heap is used as a priority queue to get the minimum distance vertex from set of not
yetincludedvertices. Time complexity of operations like extract-min and decrease-key value is O(LogV) for Min Heap.
Following are the detailed steps.
1)Create a Min Heap of size V where V is the number of vertices in the given graph. Every node of min heap contains vertex number and distance
value of the vertex.
2)Initialize Min Heap with source vertex as root (the distance value assigned to source vertex is 0). The distance value assigned to all other vertices
is INF (infinite).
3)While Min Heap is not empty, do following
..a)Extract the vertex with minimum distance value node from Min Heap. Let the extracted vertex be u.
..b)For every adjacent vertex v of u, check if v is in Min Heap. If v is in Min Heap and distance value is more than weight of u-v plus distance
value of u, then update the distance value of v.
Let us understand with the following example. Let the given source vertex be 0

Initially, distance value of source vertex is 0 and INF (infinite) for all other vertices. So source vertex is extracted from Min Heap and distance
values of vertices adjacent to 0 (1 and 7) are updated. Min Heap contains all vertices except vertex 0.
The vertices in green color are the vertices for which minimum distances are finalized and are not in Min Heap

Since distance value of vertex 1 is minimum among all nodes in Min Heap, it is extracted from Min Heap and distance values of vertices adjacent
to 1 are updated (distance is updated if the a vertex is not in Min Heap and distance through 1 is shorter than the previous distance). Min Heap
contains all vertices except vertex 0 and 1.

Pick the vertex with minimum distance value from min heap. Vertex 7 is picked. So min heap now contains all vertices except 0, 1 and 7. Update
the distance values of adjacent vertices of 7. The distance value of vertex 6 and 8 becomes finite (15 and 9 respectively).

Pick the vertex with minimum distance from min heap. Vertex 6 is picked. So min heap now contains all vertices except 0, 1, 7 and 6. Update the
distance values of adjacent vertices of 6. The distance value of vertex 5 and 8 are updated.

Above steps are repeated till min heap doesnt become empty. Finally, we get the following shortest path tree.

// C / C++ program for Dijkstra's shortest path algorithm for adjacency

// list representation of graph
#include <stdio.h>
#include <stdlib.h>
#include <limits.h>
// A structure to represent a node in adjacency list
struct AdjListNode
{
int dest;
int weight;
struct AdjListNode* next;
};
// A structure to represent an adjacency liat
struct AdjList
{
struct AdjListNode *head; // pointer to head node of list
};
// A structure to represent a graph. A graph is an array of adjacency lists.
// Size of array will be V (number of vertices in graph)
struct Graph
{
int V;
struct AdjList* array;
};
// A utility function to create a new adjacency list node
struct AdjListNode* newAdjListNode(int dest, int weight)
{
struct AdjListNode* newNode =
(struct AdjListNode*) malloc(sizeof(struct AdjListNode));
newNode->dest = dest;
newNode->weight = weight;
newNode->next = NULL;
return newNode;
}
// A utility function that creates a graph of V vertices
struct Graph* createGraph(int V)
{
struct Graph* graph = (struct Graph*) malloc(sizeof(struct Graph));
graph->V = V;
// Create an array of adjacency lists. Size of array will be V
graph->array = (struct AdjList*) malloc(V * sizeof(struct AdjList));
// Initialize each adjacency list as empty by making head as NULL
for (int i = 0; i < V; ++i)
graph->array[i].head = NULL;
return graph;
}

// Adds an edge to an undirected graph

void addEdge(struct Graph* graph, int src, int dest, int weight)
{
// Add an edge from src to dest. A new node is added to the adjacency
// list of src. The node is added at the begining
struct AdjListNode* newNode = newAdjListNode(dest, weight);
newNode->next = graph->array[src].head;
graph->array[src].head = newNode;
// Since graph is undirected, add an edge from dest to src also
newNode = newAdjListNode(src, weight);
newNode->next = graph->array[dest].head;
graph->array[dest].head = newNode;
}
// Structure to represent a min heap node
struct MinHeapNode
{
int v;
int dist;
};
// Structure to represent a min heap
struct MinHeap
{
int size;
// Number of heap nodes present currently
int capacity; // Capacity of min heap
int *pos;
// This is needed for decreaseKey()
struct MinHeapNode **array;
};
// A utility function to create a new Min Heap Node
struct MinHeapNode* newMinHeapNode(int v, int dist)
{
struct MinHeapNode* minHeapNode =
(struct MinHeapNode*) malloc(sizeof(struct MinHeapNode));
minHeapNode->v = v;
minHeapNode->dist = dist;
return minHeapNode;
}
// A utility function to create a Min Heap
struct MinHeap* createMinHeap(int capacity)
{
struct MinHeap* minHeap =
(struct MinHeap*) malloc(sizeof(struct MinHeap));
minHeap->pos = (int *)malloc(capacity * sizeof(int));
minHeap->size = 0;
minHeap->capacity = capacity;
minHeap->array =
(struct MinHeapNode**) malloc(capacity * sizeof(struct MinHeapNode*));
return minHeap;
}
// A utility function to swap two nodes of min heap. Needed for min heapify
void swapMinHeapNode(struct MinHeapNode** a, struct MinHeapNode** b)
{
struct MinHeapNode* t = *a;
*a = *b;
*b = t;
}
// A standard function to heapify at given idx
// This function also updates position of nodes when they are swapped.
// Position is needed for decreaseKey()
void minHeapify(struct MinHeap* minHeap, int idx)
{
int smallest, left, right;
smallest = idx;
left = 2 * idx + 1;
right = 2 * idx + 2;
if (left < minHeap->size &&
minHeap->array[left]->dist < minHeap->array[smallest]->dist )
smallest = left;
if (right < minHeap->size &&
minHeap->array[right]->dist < minHeap->array[smallest]->dist )
smallest = right;
if (smallest != idx)

{
// The nodes to be swapped in min heap
MinHeapNode *smallestNode = minHeap->array[smallest];
MinHeapNode *idxNode = minHeap->array[idx];
// Swap positions
minHeap->pos[smallestNode->v] = idx;
minHeap->pos[idxNode->v] = smallest;
// Swap nodes
swapMinHeapNode(&minHeap->array[smallest], &minHeap->array[idx]);
minHeapify(minHeap, smallest);
}
}
// A utility function to check if the given minHeap is ampty or not
int isEmpty(struct MinHeap* minHeap)
{
return minHeap->size == 0;
}
// Standard function to extract minimum node from heap
struct MinHeapNode* extractMin(struct MinHeap* minHeap)
{
if (isEmpty(minHeap))
return NULL;
// Store the root node
struct MinHeapNode* root = minHeap->array[0];
// Replace root node with last node
struct MinHeapNode* lastNode = minHeap->array[minHeap->size - 1];
minHeap->array[0] = lastNode;
// Update position of last node
minHeap->pos[root->v] = minHeap->size-1;
minHeap->pos[lastNode->v] = 0;
// Reduce heap size and heapify root
--minHeap->size;
minHeapify(minHeap, 0);
return root;
}
// Function to decreasy dist value of a given vertex v. This function
// uses pos[] of min heap to get the current index of node in min heap
void decreaseKey(struct MinHeap* minHeap, int v, int dist)
{
// Get the index of v in heap array
int i = minHeap->pos[v];
// Get the node and update its dist value
minHeap->array[i]->dist = dist;
// Travel up while the complete tree is not hepified.
// This is a O(Logn) loop
while (i && minHeap->array[i]->dist < minHeap->array[(i - 1) / 2]->dist)
{
// Swap this node with its parent
minHeap->pos[minHeap->array[i]->v] = (i-1)/2;
minHeap->pos[minHeap->array[(i-1)/2]->v] = i;
swapMinHeapNode(&minHeap->array[i], &minHeap->array[(i - 1) / 2]);
// move to parent index
i = (i - 1) / 2;
}
}
// A utility function to check if a given vertex
// 'v' is in min heap or not
bool isInMinHeap(struct MinHeap *minHeap, int v)
{
if (minHeap->pos[v] < minHeap->size)
return true;
return false;
}
// A utility function used to print the solution
void printArr(int dist[], int n)

{
printf("Vertex Distance from Source\n");
for (int i = 0; i < n; ++i)
printf("%d \t\t %d\n", i, dist[i]);
}
// The main function that calulates distances of shortest paths from src to all
// vertices. It is a O(ELogV) function
void dijkstra(struct Graph* graph, int src)
{
int V = graph->V;// Get the number of vertices in graph
int dist[V];
// dist values used to pick minimum weight edge in cut
// minHeap represents set E
struct MinHeap* minHeap = createMinHeap(V);
// Initialize min heap with all vertices. dist value of all vertices
for (int v = 0; v < V; ++v)
{
dist[v] = INT_MAX;
minHeap->array[v] = newMinHeapNode(v, dist[v]);
minHeap->pos[v] = v;
}
// Make dist value of src vertex as 0 so that it is extracted first
minHeap->array[src] = newMinHeapNode(src, dist[src]);
minHeap->pos[src] = src;
dist[src] = 0;
decreaseKey(minHeap, src, dist[src]);
// Initially size of min heap is equal to V
minHeap->size = V;
// In the followin loop, min heap contains all nodes
// whose shortest distance is not yet finalized.
while (!isEmpty(minHeap))
{
// Extract the vertex with minimum distance value
struct MinHeapNode* minHeapNode = extractMin(minHeap);
int u = minHeapNode->v; // Store the extracted vertex number
// Traverse through all adjacent vertices of u (the extracted
// vertex) and update their distance values
struct AdjListNode* pCrawl = graph->array[u].head;
while (pCrawl != NULL)
{
int v = pCrawl->dest;
// If shortest distance to v is not finalized yet, and distance to v
// through u is less than its previously calculated distance
if (isInMinHeap(minHeap, v) && dist[u] != INT_MAX &&
pCrawl->weight + dist[u] < dist[v])
{
dist[v] = dist[u] + pCrawl->weight;
// update distance value in min heap also
decreaseKey(minHeap, v, dist[v]);
}
pCrawl = pCrawl->next;
}
}
// print the calculated shortest distances
printArr(dist, V);
}
// Driver program to test above functions
int main()
{
// create the graph given in above fugure
int V = 9;
struct Graph* graph = createGraph(V);
addEdge(graph, 0, 1, 4);
addEdge(graph, 0, 7, 8);
addEdge(graph, 1, 2, 8);
addEdge(graph, 1, 7, 11);
addEdge(graph, 2, 3, 7);
addEdge(graph, 2, 8, 2);
addEdge(graph, 2, 5, 4);
addEdge(graph, 3, 4, 9);

addEdge(graph,
addEdge(graph,
addEdge(graph,
addEdge(graph,
addEdge(graph,
addEdge(graph,

3,
4,
5,
6,
6,
7,

5,
5,
6,
7,
8,
8,

14);
10);
2);
1);
6);
7);

dijkstra(graph, 0);
return 0;
}

Output:
Vertex
0
1
2
3
4
5
6
7
8

Distance from Source

0
4
12
19
21
11
9
8
14

Time Complexity:The time complexity of the above code/algorithm looks O(V^2) as there are two nested while loops. If we take a closer look,
we can observe that the statements in inner loop are executed O(V+E) times (similar to BFS). The inner loop has decreaseKey() operation which
takes O(LogV) time. So overall time complexity is O(E+V)*O(LogV) which is O((E+V)*LogV) = O(ELogV)
Note that the above code uses Binary Heap for Priority Queue implementation. Time complexity can be reduced to O(E + VLogV) using
Fibonacci Heap. The reason is, Fibonacci Heap takes O(1) time for decrease-key operation while Binary Heap takes O(Logn) time.
Notes:
1)The code calculates shortest distance, but doesnt calculate the path information. We can create a parent array, update the parent array when
distance is updated (likeprims implementation) and use it show the shortest path from source to different vertices.
2)The code is for undirected graph, same dijekstra function can be used for directed graphs also.
3)The code finds shortest distances from source to all vertices. If we are interested only in shortest distance from source to a single target, we can
break the for loop when the picked minimum distance vertex is equal to target (Step 3.a of algorithm).
4)Dijkstras algorithm doesnt work for graphs with negative weight edges. For graphs with negative weight edges,BellmanFord algorithmcan be
used, we will soon be discussing it as a separate post.
References:
Introduction to Algorithms by Clifford Stein, Thomas H. Cormen, Charles E. Leiserson, Ronald L.
Algorithms by Sanjoy Dasgupta, Christos Papadimitriou, Umesh Vazirani

Job Sequencing Problem | Set 1 (Greedy Algorithm)

Given an array of jobs where every job has a deadline and associated profit if the job is finished before the deadline. It is also given that every job
takes single unit of time, so the minimum possible deadline for any job is 1. How to maximize total profit if only one job can be scheduled at a time.
Examples:
Input: Four Jobs with following deadlines and profits
JobID
Deadline
Profit
a
4
20
b
1
10
c
1
40
d
1
30
Output: Following is maximum profit sequence of jobs
c, a
Input: Five Jobs with following deadlines and profits
JobID
Deadline
Profit
a
2
100
b
1
19
c
2
27
d
1
25
e
3
15
Output: Following is maximum profit sequence of jobs
c, a, e

A Simple Solution is to generate all subsets of given set of jobs and check individual subset for feasibility of jobs in that subset. Keep track of
maximum profit among all feasible subsets. The time complexity of this solution is exponential.
This is a standard Greedy Algorithm problem. Following is algorithm.
1) Sort all jobs in decreasing order of profit.
2) Initialize the result sequence as first job in sorted jobs.
3) Do following for remaining n-1 jobs
.......a) If the current job can fit in the current result sequence
without missing the deadline, add current job to the result.
Else ignore the current job.

The Following is C++ implementation of above algorithm.

// Program to find the maximum profit job sequence from a given array
// of jobs with deadlines and profits
#include<iostream>
#include<algorithm>
using namespace std;
// A structure to represent a job
struct Job
{
char id;
// Job Id
int dead;
// Deadline of job
int profit; // Profit if job is over before or on deadline
};
// This function is used for sorting all jobs according to profit
bool comparison(Job a, Job b)
{
return (a.profit > b.profit);
}
// Returns minimum number of platforms reqquired
void printJobScheduling(Job arr[], int n)
{
// Sort all jobs according to decreasing order of prfit
sort(arr, arr+n, comparison);
int result[n]; // To store result (Sequence of jobs)
bool slot[n]; // To keep track of free time slots
// Initialize all slots to be free
for (int i=0; i<n; i++)
slot[i] = false;
// Iterate through all given jobs
for (int i=0; i<n; i++)
{
// Find a free slot for this job (Note that we start
// from the last possible slot)

for (int j=min(n, arr[i].dead)-1; j>=0; j--)

{
// Free slot found
if (slot[j]==false)
{
result[j] = i; // Add this job to result
slot[j] = true; // Make this slot occupied
break;
}
}
}
// Print the result
for (int i=0; i<n; i++)
if (slot[i])
cout << arr[result[i]].id << " ";
}
// Driver program to test methods
int main()
{
Job arr[5] = { {'a', 2, 100}, {'b', 1, 19}, {'c', 2, 27},
{'d', 1, 25}, {'e', 3, 15}};
int n = sizeof(arr)/sizeof(arr[0]);
cout << "Following is maximum profit sequence of jobs\n";
printJobScheduling(arr, n);
return 0;
}

Output:
Following is maximum profit sequence of jobs
c a e

Time Complexity of the above solution is O(n2). It can be optimized to almost O(n) by using union-find data structure. We will son be discussing
the optimized solution.
Sources:
http://ocw.mit.edu/courses/civil-and-environmental-engineering/1-204-computer-algorithms-in-systems-engineering-spring-2010/lecturenotes/MIT1_204S10_lec10.pdf

Greedy Algorithm to find Minimum number of Coins

Given a value V, if we want to make change for V Rs, and we have infinite supply of each of the denominations in Indian currency, i.e., we have
infinite supply of { 1, 2, 5, 10, 20, 50, 100, 500, 1000} valued coins/notes, what is the minimum number of coins and/or notes needed to make
the change?
Examples:
Input: V = 70
Output: 2
We need a 50 Rs note and a 20 Rs note.
Input: V = 121
Output: 3
We need a 100 Rs note, a 20 Rs note and a
1 Rs coin.
1) Initialize result as empty.
2) find the largest denomination that is
smaller than V.
3) Add found denomination to result. Subtract
value of found denomination from V.
4) If V becomes 0, then print result.
Else repeat steps 2 and 3 for new value of V

Below is C++ implementation of above algorithm.

// C++ program to find minimum number of denominations
#include <bits/stdc++.h>
using namespace std;
// All denominations of Indian Currency
int deno[] = {1, 2, 5, 10, 20, 50, 100, 500, 1000};
int n = sizeof(deno)/sizeof(deno[0]);
// Driver program
void findMin(int V)
{
// Initialize result
vector<int> ans;
// Traverse through all denomination
for (int i=n-1; i>=0; i--)
{
// Find denominations
while (V >= deno[i])
{
V -= deno[i];
ans.push_back(deno[i]);
}
}
// Print result
for (int i = 0; i < ans.size(); i++)
cout << ans[i] << " ";
}
// Driver program
int main()
{
int n = 93;
cout << "Following is minimal number of change for " << n << " is ";
findMin(n);
return 0;
}

Output:
Following is minimal number of change for 93 is 50 20 20 2 1

Note that above approach may not work for all denominations. For example, it doesnt work for denominations {9, 6, 5, 1} and V = 11. The
above approach would print 9, 1 and 1. But we can use 2 denominations 5 and 6.
For general input, we use below dynamic programming approach.
Find minimum number of coins that make a given value
Thanks to Utkarsh for providing above solution here.

K Centers Problem | Set 1 (Greedy Approximate Algorithm)

Given n cities and distances between every pair of cities, select k cities to place warehouses (or ATMs) such that the maximum distance of a city
to a warehouse (or ATM) is minimized.
For example consider the following four cities, 0, 1, 2 and 3 and distances between them, how do place 2 ATMs among these 4 cities so that the
maximum distance of a city to an ATM is minimized.

There is no polynomial time solution available for this problem as the problem is a known NP-Hard problem. There is a polynomial time Greedy
approximate algorithm, the greedy algorithm provides a solution which is never worse that twice the optimal solution. The greedy solution works
only if the distances between cities follow Triangular Inequality (Distance between two points is always smaller than sum of distances through a
third point).
The 2-Approximate Greedy Algorithm:
1) Choose the first center arbitrarily.
2) Choose remaining k-1 centers using the following criteria.
Let c1, c2, c3, ci be the already chosen centers. Choose
(i+1)th center by picking the city which is farthest from already
selected centers, i.e, the point p which has following value as maximum
Min[dist(p, c1), dist(p, c2), dist(p, c3), . dist(p, ci)]
The following diagram taken from here illustrates above algorithm.

Example (k = 3 in the above shown Graph)

a) Let the first arbitrarily picked vertex be 0.
b) The next vertex is 1 because 1 is the farthest vertex from 0.
c) Remaining cities are 2 and 3. Calculate their distances from already selected centers (0 and 1). The greedy algorithm basically calculates
following values.
Minimum of all distanced from 2 to already considered centers
Min[dist(2, 0), dist(2, 1)] = Min[7, 8] = 7
Minimum of all distanced from 3 to already considered centers
Min[dist(3, 0), dist(3, 1)] = Min[6, 5] = 5
After computing the above values, the city 2 is picked as the value corresponding to 2 is maximum.
Note that the greedy algorithm doesnt give best solution for k = 2 as this is just an approximate algorithm with bound as twice of optimal.
Proof that the above greedy algorithm is 2 approximate.
Let OPT be the maximum distance of a city from a center in the Optimal solution. We need to show that the maximum distance obtained from
Greedy algorithm is 2*OPT.
The proof can be done using contradiction.
a) Assume that the distance from the furthest point to all centers is > 2OPT.

b) This means that distances between all centers are also > 2OPT.
c) We have k + 1 points with distances > 2OPT between every pair.
d) Each point has a center of the optimal solution with distance ? OPT to it.
e) There exists a pair of points with the same center X in the optimal solution (pigeonhole principle: k optimal centers, k+1 points)
f) The distance between them is at most 2OPT (triangle inequality) which is a contradiction.
Source:
http://algo2.iti.kit.edu/vanstee/courses/kcenter.pdf

Dynamic Programming | Set 1 (Overlapping Subproblems Property)

Dynamic Programming is an algorithmic paradigm that solves a given complex problem by breaking it into subproblems and stores the results of
subproblems to avoid computing the same results again. Following are the two main properties of a problem that suggest that the given problem
can be solved using Dynamic programming.
1) Overlapping Subproblems
2) Optimal Substructure
1) Overlapping Subproblems:
Like Divide and Conquer, Dynamic Programming combines solutions to sub-problems. Dynamic Programming is mainly used when solutions of
same subproblems are needed again and again. In dynamic programming, computed solutions to subproblems are stored in a table so that these
dont have to recomputed. So Dynamic Programming is not useful when there are no common (overlapping) subproblems because there is no point
storing the solutions if they are not needed again. For example, Binary Search doesnt have common subproblems. If we take example of following
recursive program for Fibonacci Numbers, there are many subproblems which are solved again and again.
/* simple recursive program for Fibonacci numbers */
int fib(int n)
{
if ( n <= 1 )
return n;
return fib(n-1) + fib(n-2);
}

Recursion tree for execution of fib(5)

fib(5)
/
\
fib(4)
fib(3)
/
\
/
\
fib(3)
fib(2)
fib(2)
fib(1)
/
\
/
\
/
\
fib(2) fib(1) fib(1) fib(0) fib(1) fib(0)
/
\
fib(1) fib(0)

We can see that the function f(3) is being called 2 times. If we would have stored the value of f(3), then instead of computing it again, we would
have reused the old stored value. There are following two different ways to store the values so that these values can be reused.
a) Memoization (Top Down):
b) Tabulation (Bottom Up):
a) Memoization (Top Down): The memoized program for a problem is similar to the recursive version with a small modification that it looks into a
lookup table before computing solutions. We initialize a lookup array with all initial values as NIL. Whenever we need solution to a subproblem,
we first look into the lookup table. If the precomputed value is there then we return that value, otherwise we calculate the value and put the result in
lookup table so that it can be reused later.
Following is the memoized version for nth Fibonacci Number.
/* Memoized version for nth Fibonacci number */
#include<stdio.h>
#define NIL -1
#define MAX 100
int lookup[MAX];
/* Function to initialize NIL values in lookup table */
void _initialize()
{
int i;
for (i = 0; i < MAX; i++)
lookup[i] = NIL;
}
/* function for
int fib(int n)
{
if(lookup[n]
{
if ( n <= 1
lookup[n]
else
lookup[n]
}

nth Fibonacci number */

== NIL)
)
= n;
= fib(n-1) + fib(n-2);

return lookup[n];
}
int main ()
{
int n = 40;
_initialize();
printf("Fibonacci number is %d ", fib(n));
getchar();
return 0;
}

b) Tabulation (Bottom Up): The tabulated program for a given problem builds a table in bottom up fashion and returns the last entry from table.
/* tabulated version */
#include<stdio.h>
int fib(int n)
{
int f[n+1];
int i;
f[0] = 0; f[1] = 1;
for (i = 2; i <= n; i++)
f[i] = f[i-1] + f[i-2];
return f[n];
}
int main ()
{
int n = 9;
printf("Fibonacci number is %d ", fib(n));
getchar();
return 0;
}

Both tabulated and Memoized store the solutions of subproblems. In Memoized version, table is filled on demand while in tabulated version,
starting from the first entry, all entries are filled one by one. Unlike the tabulated version, all entries of the lookup table are not necessarily filled in
memoized version. For example, memoized solution of LCS problem doesnt necessarily fill all entries.
To see the optimization achieved by memoized and tabulated versions over the basic recursive version, see the time taken by following runs for
40th Fibonacci number.
Simple recursive program
Memoized version
tabulated version
Also see method 2 of Ugly Number post for one more simple example where we have overlapping subproblems and we store the results of
subproblems.
We will be covering Optimal Substructure Property and some more example problems in future posts on Dynamic Programming.
Try following questions as an exercise of this post.
1) Write a memoized version for LCS problem. Note that the tabular version is given in the CLRS book.
2) How would you choose between Memoization and Tabulation?
References:
http://www.youtube.com/watch?v=V5hZoJ6uK-s

Dynamic Programming | Set 2 (Optimal Substructure Property)

As we discussed in Set 1, following are the two main properties of a problem that suggest that the given problem can be solved using Dynamic
programming.
1) Overlapping Subproblems
2) Optimal Substructure
We have already discussed Overlapping Subproblem property in the Set 1. Let us discuss Optimal Substructure property here.
2) Optimal Substructure: A given problems has Optimal Substructure Property if optimal solution of the given problem can be obtained by using
optimal solutions of its subproblems.
For example the shortest path problem has following optimal substructure property: If a node x lies in the shortest path from a source node u to
destination node v then the shortest path from u to v is combination of shortest path from u to x and shortest path from x to v. The standard All
Pair Shortest Path algorithms like FloydWarshall and BellmanFord are typical examples of Dynamic Programming.
On the other hand the Longest path problem doesnt have the Optimal Substructure property. Here by Longest Path we mean longest simple path
(path without cycle) between two nodes. Consider the following unweighted graph given in the CLRS book. There are two longest paths from q to
t: q -> r ->t and q ->s->t. Unlike shortest paths, these longest paths do not have the optimal substructure property. For example, the longest path
q->r->t is not a combination of longest path from q to r and longest path from r to t, because the longest path from q to r is q->s->t->r.

We will be covering some example problems in future posts on Dynamic Programming.

References:
http://en.wikipedia.org/wiki/Optimal_substructure
CLRS book

Dynamic Programming | Set 3 (Longest Increasing Subsequence)

We have discussed Overlapping Subproblems and Optimal Substructure properties in Set 1 and Set 2 respectively.
Let us discuss Longest Increasing Subsequence (LIS) problem as an example problem that can be solved using Dynamic Programming.
The longest Increasing Subsequence (LIS) problem is to find the length of the longest subsequence of a given sequence such that all elements of the
subsequence are sorted in increasing order. For example, length of LIS for { 10, 22, 9, 33, 21, 50, 41, 60, 80 } is 6 and LIS is {10, 22, 33, 50,
60, 80}.
Optimal Substructure:
Let arr[0..n-1] be the input array and L(i) be the length of the LIS till index i such that arr[i] is part of LIS and arr[i] is the last element in LIS, then
L(i) can be recursively written as.
L(i) = { 1 + Max ( L(j) ) } where j < i and arr[j] < arr[i] and if there is no such j then L(i) = 1
To get LIS of a given array, we need to return max(L(i)) where 0 < i < n So the LIS problem has optimal substructure property as the main
problem can be solved using solutions to subproblems. Overlapping Subproblems:
Following is simple recursive implementation of the LIS problem. The implementation simply follows the recursive structure mentioned above. The
value of lis ending with every element is returned using max_ending_here. The overall lis is returned using pointer to a variable max.

C/C++
/* A Naive C/C++ recursive implementation of LIS problem */
#include<stdio.h>
#include<stdlib.h>
/* To make use of recursive calls, this function must return
two things:
1) Length of LIS ending with element arr[n-1]. We use
max_ending_here for this purpose
2) Overall maximum as the LIS may end with an element
before arr[n-1] max_ref is used this purpose.
The value of LIS of full array of size n is stored in
*max_ref which is our final result */
int _lis( int arr[], int n, int *max_ref)
{
/* Base case */
if (n == 1)
return 1;
// 'max_ending_here' is length of LIS ending with arr[n-1]
int res, max_ending_here = 1;
/* Recursively get all LIS ending with arr[0], arr[1] ...
arr[n-2]. If arr[i-1] is smaller than arr[n-1], and
max ending with arr[n-1] needs to be updated, then
update it */
for (int i = 1; i < n; i++)
{
res = _lis(arr, i, max_ref);
if (arr[i-1] < arr[n-1] && res + 1 > max_ending_here)
max_ending_here = res + 1;
}
// Compare max_ending_here with the overall max. And
// update the overall max if needed
if (*max_ref < max_ending_here)
*max_ref = max_ending_here;
// Return length of LIS ending with arr[n-1]
return max_ending_here;
}
// The wrapper function for _lis()
int lis(int arr[], int n)
{
// The max variable holds the result
int max = 1;
// The function _lis() stores its result in max
_lis( arr, n, &max );
// returns max
return max;
}
/* Driver program to test above function */
int main()

{
int arr[] = { 10, 22, 9, 33, 21, 50, 41, 60 };
int n = sizeof(arr)/sizeof(arr[0]);
printf("Length of LIS is %d\n", lis( arr, n ));
return 0;
}

Python
# A naive Python implementation of LIS problem
""" To make use of recursive calls, this function must return
two things:
1) Length of LIS ending with element arr[n-1]. We use
max_ending_here for this purpose
2) Overall maximum as the LIS may end with an element
before arr[n-1] max_ref is used this purpose.
The value of LIS of full array of size n is stored in
*max_ref which is our final result """
# global variable to store the maximum
global maximum
def _lis(arr , n ):
# to allow the access of global variable
global maximum
# Base Case
if n == 1 :
return 1
# maxEndingHere is the length of LIS ending with arr[n-1]
maxEndingHere = 1
"""Recursively get all LIS ending with arr[0], arr[1]..arr[n-2]
IF arr[n-1] is maller than arr[n-1], and max ending with
arr[n-1] needs to be updated, then update it"""
for i in xrange(1, n):
res = _lis(arr , i)
if arr[i-1] < arr[n-1] and res+1 > maxEndingHere:
maxEndingHere = res +1
# Compare maxEndingHere with overall maximum.And update
# the overall maximum if needed
maximum = max(maximum , maxEndingHere)
return maxEndingHere
def lis(arr):
# to allow the access of global variable
global maximum
# lenght of arr
n = len(arr)
# maximum variable holds the result
maximum = 1
# The function _lis() stores its result in maximum
_lis(arr , n)
return maximum
# Driver program to test the above function
arr = [10 , 22 , 9 , 33 , 21 , 41 , 60]
n = len(arr)
print "Length of LIS is ", lis(arr)
# This code is contributed by NIKHIL KUMAR SINGH

/
lis(3)
/
\
lis(2) lis(1)

lis(4)
|
lis(2)
/
lis(1)

\
lis(1)

/
lis(1)

We can see that there are many subproblems which are solved again and again. So this problem has Overlapping Substructure property and
recomputation of same subproblems can be avoided by either using Memoization or Tabulation. Following is a tabluated implementation for the
LIS problem.

C/C++
/* Dynamic Programming C/C++ implementation of LIS problem */
#include<stdio.h>
#include<stdlib.h>
/* lis() returns the length of the longest increasing
subsequence in arr[] of size n */
int lis( int arr[], int n )
{
int *lis, i, j, max = 0;
lis = (int*) malloc ( sizeof( int ) * n );
/* Initialize LIS values for all indexes */
for ( i = 0; i < n; i++ )
lis[i] = 1;
/* Compute optimized LIS values in bottom up manner */
for ( i = 1; i < n; i++ )
for ( j = 0; j < i; j++ )
if ( arr[i] > arr[j] && lis[i] < lis[j] + 1)
lis[i] = lis[j] + 1;
/* Pick maximum of all LIS values */
for ( i = 0; i < n; i++ )
if ( max < lis[i] )
max = lis[i];
/* Free memory to avoid memory leak */
free( lis );
return max;
}
/* Driver program to test above function */
int main()
{
int arr[] = { 10, 22, 9, 33, 21, 50, 41, 60 };
int n = sizeof(arr)/sizeof(arr[0]);
printf("Length of LIS is %d\n", lis( arr, n ) );
return 0;
}

Python
# Dynamic programming Python implementation of LIS problem
# lis returns length of the longest increasing subsequence
# in arr of size n
def lis(arr):
n = len(arr)
# Declare the list (array) for LIS and initialize LIS
# values for all indexes
lis = [1]*n
# Compute optimized LIS values in bottom up manner
for i in range (1 , n):
for j in range(0 , i):
if arr[i] > arr[j] and lis[i]< lis[j] + 1 :
lis[i] = lis[j]+1
# Initialize maximum to 0 to get the maximum of all
# LIS
maximum = 0
# Pick maximum of all LIS values
for i in range(n):
maximum = max(maximum , lis[i])
return maximum

# end of lis function

# Driver program to test above function
arr = [10, 22, 9, 33, 21, 50, 41, 60]
print "Length of LIS is", lis(arr)
# This code is contributed by Nikhil Kumar Singh

Length of LIS is 5

Note that the time complexity of the above Dynamic Programming (DP) solution is O(n^2) and there is a O(nLogn) solution for the LIS problem.
We have not discussed the O(n Log n) solution here as the purpose of this post is to explain Dynamic Programming with a simple example. See
below post for O(n Log n) solution.
Longest Increasing Subsequence Size (N log N)

Dynamic Programming | Set 4 (Longest Common Subsequence)

We have discussed Overlapping Subproblems and Optimal Substructure properties in Set 1 and Set 2 respectively. We also discussed one
example problem in Set 3. Let us discuss Longest Common Subsequence (LCS) problem as one more example problem that can be solved using
Dynamic Programming.
LCS Problem Statement: Given two sequences, find the length of longest subsequence present in both of them. A subsequence is a sequence that
appears in the same relative order, but not necessarily contiguous. For example, abc, abg, bdf, aeg, acefg, .. etc are subsequences of abcdefg. So
a string of length n has 2^n different possible subsequences.
It is a classic computer science problem, the basis of diff (a file comparison program that outputs the differences between two files), and has
applications in bioinformatics.
Examples:
LCS for input Sequences ABCDGH and AEDFHR is ADH of length 3.
LCS for input Sequences AGGTAB and GXTXAYB is GTAB of length 4.
The naive solution for this problem is to generate all subsequences of both given sequences and find the longest matching subsequence. This
solution is exponential in term of time complexity. Let us see how this problem possesses both important properties of a Dynamic Programming
(DP) Problem.
1) Optimal Substructure:
Let the input sequences be X[0..m-1] and Y[0..n-1] of lengths m and n respectively. And let L(X[0..m-1], Y[0..n-1]) be the length of LCS of the
two sequences X and Y. Following is the recursive definition of L(X[0..m-1], Y[0..n-1]).
If last characters of both sequences match (or X[m-1] == Y[n-1]) then
L(X[0..m-1], Y[0..n-1]) = 1 + L(X[0..m-2], Y[0..n-2])
If last characters of both sequences do not match (or X[m-1] != Y[n-1]) then
L(X[0..m-1], Y[0..n-1]) = MAX ( L(X[0..m-2], Y[0..n-1]), L(X[0..m-1], Y[0..n-2])
Examples:
1) Consider the input strings AGGTAB and GXTXAYB. Last characters match for the strings. So length of LCS can be written as:
L(AGGTAB, GXTXAYB) = 1 + L(AGGTA, GXTXAY)
2) Consider the input strings ABCDGH and AEDFHR. Last characters do not match for the strings. So length of LCS can be written as:
L(ABCDGH, AEDFHR) = MAX ( L(ABCDG, AEDFHR), L(ABCDGH, AEDFH) )
So the LCS problem has optimal substructure property as the main problem can be solved using solutions to subproblems.
2) Overlapping Subproblems:
Following is simple recursive implementation of the LCS problem. The implementation simply follows the recursive structure mentioned above.

C/C++
/* A Naive recursive implementation of LCS problem */
#include<stdio.h>
#include<stdlib.h>
#include<string.h>
int max(int a, int b);
/* Returns length of LCS for X[0..m-1], Y[0..n-1] */
int lcs( char *X, char *Y, int m, int n )
{
if (m == 0 || n == 0)
return 0;
if (X[m-1] == Y[n-1])
return 1 + lcs(X, Y, m-1, n-1);
else
return max(lcs(X, Y, m, n-1), lcs(X, Y, m-1, n));
}
/* Utility function to get max of 2 integers */
int max(int a, int b)
{
return (a > b)? a : b;
}
/* Driver program to test above function */
int main()
{

char X[] = "AGGTAB";

char Y[] = "GXTXAYB";
int m = strlen(X);
int n = strlen(Y);
printf("Length of LCS is %d\n", lcs( X, Y, m, n ) );
return 0;
}

Python
# A Naive recursive Python implementation of LCS problem
def lcs(X, Y, m, n):
if m == 0 or n == 0:
return 0;
elif X[m-1] == Y[n-1]:
return 1 + lcs(X, Y, m-1, n-1);
else:
return max(lcs(X, Y, m, n-1), lcs(X, Y, m-1, n));
# Driver program to test the above function
X = "AGGTAB"
Y = "GXTXAYB"
print "Length of LCS is ", lcs(X , Y, len(X), len(Y))

Length of LCS is 4

Time complexity of the above naive recursive approach is O(2^n) in worst case and worst case happens when all characters of X and Y mismatch
i.e., length of LCS is 0.
Considering the above implementation, following is a partial recursion tree for input strings AXYT and AYZX
lcs("AXYT", "AYZX")
/
\
lcs("AXY", "AYZX")
lcs("AXYT", "AYZ")
/
\
/
\
lcs("AX", "AYZX") lcs("AXY", "AYZ") lcs("AXY", "AYZ") lcs("AXYT", "AY")

In the above partial recursion tree, lcs(AXY, AYZ) is being solved twice. If we draw the complete recursion tree, then we can see that there are
many subproblems which are solved again and again. So this problem has Overlapping Substructure property and recomputation of same
subproblems can be avoided by either using Memoization or Tabulation. Following is a tabulated implementation for the LCS problem.

C/C++
/* Dynamic Programming C/C++ implementation of LCS problem */
#include<stdio.h>
#include<stdlib.h>
int max(int a, int b);
/* Returns length of LCS for X[0..m-1], Y[0..n-1] */
int lcs( char *X, char *Y, int m, int n )
{
int L[m+1][n+1];
int i, j;
/* Following steps build L[m+1][n+1] in bottom up fashion. Note
that L[i][j] contains length of LCS of X[0..i-1] and Y[0..j-1] */
for (i=0; i<=m; i++)
{
for (j=0; j<=n; j++)
{
if (i == 0 || j == 0)
L[i][j] = 0;
else if (X[i-1] == Y[j-1])
L[i][j] = L[i-1][j-1] + 1;
else
L[i][j] = max(L[i-1][j], L[i][j-1]);
}
}

/* L[m][n] contains length of LCS for X[0..n-1] and Y[0..m-1] */

return L[m][n];
}
/* Utility function to get max of 2 integers */
int max(int a, int b)
{
return (a > b)? a : b;
}
/* Driver program to test above function */
int main()
{
char X[] = "AGGTAB";
char Y[] = "GXTXAYB";
int m = strlen(X);
int n = strlen(Y);
printf("Length of LCS is %d\n", lcs( X, Y, m, n ) );
return 0;
}

Python
# Dynamic Programming implementation of LCS problem
def lcs(X , Y):
# find the length of the strings
m = len(X)
n = len(Y)
# declaring the array for storing the dp values
L = [[None]*(n+1) for i in xrange(m+1)]
"""Following steps build L[m+1][n+1] in bottom up fashion
Note: L[i][j] contains length of LCS of X[0..i-1]
and Y[0..j-1]"""
for i in range(m+1):
for j in range(n+1):
if i == 0 or j == 0 :
L[i][j] = 0
elif X[i-1] == Y[j-1]:
L[i][j] = L[i-1][j-1]+1
else:
L[i][j] = max(L[i-1][j] , L[i][j-1])
# L[m][n] contains the length of LCS of X[0..n-1] & Y[0..m-1]
return L[m][n]
#end of function lcs
# Driver program to test the above function
X = "AGGTAB"
Y = "GXTXAYB"
print "Length of LCS is ", lcs(X, Y)
# This code is contributed by Nikhil Kumar Singh(nickzuck_007)

The above algorithm/code returns only length of LCS. Please see the following post for printing the LCS.
Printing Longest Common Subsequence
References:
http://www.youtube.com/watch?v=V5hZoJ6uK-s
http://www.algorithmist.com/index.php/Longest_Common_Subsequence
http://www.ics.uci.edu/~eppstein/161/960229.html
http://en.wikipedia.org/wiki/Longest_common_subsequence_problem

Dynamic Programming | Set 5 (Edit Distance)

Given two strings str1 and str2 and below operations that can performed on str1. Find minimum number of edits (operations) required to convert
str1? into str2?.
a. Insert
b. Remove
c. Replace
All of the above operations are of equal cost.
Examples:
Input: str1 = "geek", str2 = "gesek"
Output: 1
We can convert str1 into str2 by inserting a 's'.
Input: str1 = "cat", str2 = "cut"
Output: 1
We can convert str1 into str2 by replacing 'a' with 'u'.
Input: str1 = "sunday", str2 = "saturday"
Output: 3
Last three and first characters are same. We basically
need to convert "un" to "atur". This can be done using
below three operations.
Replace 'n' with 'r', insert t, insert a

What are the subproblems in this case?

The idea is process all characters one by one staring from either from left or right sides of both strings.
Let we traverse from right corner, there are two possibilities for every pair of character being traversed.
m: Length of str1 (first string)
n: Length of str2 (second string)

1. If last characters of two strings are same, nothing much to do. Ignore last characters and get count for remaining strings. So we recur for
lengths m-1 and n-1.
2. Else (If last characters are not same), we consider all operations on str1?, consider all three operations on last character of first string,
recursively compute minimum cost for all three operations and take minimum of three values.
a. Insert: Recur for m and n-1
b. Remove: Recur for m-1 and n
c. Replace: Recur for m-1 and n-1
Below is C++ implementation of above Naive recursive solution.

C++
// A Naive recursive C++ program to find minimum number
// operations to convert str1 to str2
#include<bits/stdc++.h>
using namespace std;
// Utility function to find minimum of three numbers
int min(int x, int y, int z)
{
return min(min(x, y), z);
}
int editDist(string str1 , string str2 , int m ,int n)
{
// If first string is empty, the only option is to
// insert all characters of second string into first
if (m == 0) return n;
// If second string is empty, the only option is to
// remove all characters of first string
if (n == 0) return m;
//
//
//
if

If last characters of two strings are same, nothing

much to do. Ignore last characters and get count for
remaining strings.
(str1[m-1] == str2[n-1])
return editDist(str1, str2, m-1, n-1);

// If last characters are not same, consider all three

// operations on last character

// compute minimum cost for all
// minimum of three values.
return 1 + min ( editDist(str1,
editDist(str1,
editDist(str1,
);

of first string, recursively

three operations and take
str2, m, n-1),
// Insert
str2, m-1, n), // Remove
str2, m-1, n-1) // Replace

}
// Driver program
int main()
{
// your code goes here
string str1 = "sunday";
string str2 = "saturday";
cout << editDist( str1 , str2 , str1.length(), str2.length());
return 0;
}

Python
# A Naive recursive Python program to fin minimum number
# operations to convert str1 to str2
def editDistance(str1, str2, m , n):
# If first string is empty, the only option is to
# insert all characters of second string into first
if m==0:
return n
# If second string is empty, the only option is to
# remove all characters of first string
if n==0:
return m
# If last characters of two strings are same, nothing
# much to do. Ignore last characters and get count for
# remaining strings.
if str1[m-1]==str2[n-1]:
return editDistance(str1,str2,m-1,n-1)
# If last characters are not same, consider all three
# operations on last character of first string, recursively
# compute minimum cost for all three operations and take
# minimum of three values.
return 1 + min(editDistance(str1, str2, m, n-1),
# Insert
editDistance(str1, str2, m-1, n),
# Remove
editDistance(str1, str2, m-1, n-1)
# Replace
)
# Driver program to test the above function
str1 = "sunday"
str2 = "saturday"
print editDistance(str1, str2, len(str1), len(str2))
# This code is contributed by Bhavya Jain

The time complexity of above solution is exponential. In worst case, we may end up doing O(3m) operations. The worst case happens when none
of characters of two strings match. Below is a recursive call diagram for worst case.

We can see that many subproblems are solved again and again, for example eD(2,2) is called three times. Since same suproblems are called again,
this problem has Overlapping Subprolems property. So Edit Distance problem has both properties (see this and this) of a dynamic programming
problem. Like other typical Dynamic Programming(DP) problems, recomputations of same subproblems can be avoided by constructing a
temporary array that stores results of subpriblems.

C++
// A Dynamic Programming based C++ program to find minimum
// number operations to convert str1 to str2
#include<bits/stdc++.h>
using namespace std;
// Utility function to find minimum of three numbers
int min(int x, int y, int z)
{
return min(min(x, y), z);
}
int editDistDP(string str1, string str2, int m, int n)
{
// Create a table to store results of subproblems
int dp[m+1][n+1];
// Fill d[][] in bottom up manner
for (int i=0; i<=m; i++)
{
for (int j=0; j<=n; j++)
{
// If first string is empty, only option is to
// isnert all characters of second string
if (i==0)
dp[i][j] = j; // Min. operations = j
// If second string is empty, only option is to
// remove all characters of second string
else if (j==0)
dp[i][j] = i; // Min. operations = i
// If last characters are same, ignore last char
// and recur for remaining string
else if (str1[i-1] == str2[j-1])
dp[i][j] = dp[i-1][j-1];
// If last character are different, consider all
// possibilities and find minimum
else
dp[i][j] = 1 + min(dp[i][j-1], // Insert
dp[i-1][j], // Remove
dp[i-1][j-1]); // Replace
}
}
return dp[m][n];
}
// Driver program
int main()
{

// your code goes here

string str1 = "sunday";
string str2 = "saturday";
cout << editDistDP(str1, str2, str1.length(), str2.length());
return 0;
}

Python
# A Dynamic Programming based Python program for edit
# distance problem
def editDistDP(str1, str2, m, n):
# Create a table to store results of subproblems
dp = [[0 for x in range(n+1)] for x in range(m+1)]
# Fill d[][] in bottom up manner
for i in range(m+1):
for j in range(n+1):
# If first string is empty, only option is to
# isnert all characters of second string
if i == 0:
dp[i][j] = j
# Min. operations = j
# If second string is empty, only option is to
# remove all characters of second string
elif j == 0:
dp[i][j] = i
# Min. operations = i
# If last characters are same, ignore last char
# and recur for remaining string
elif str1[i-1] == str2[j-1]:
dp[i][j] = dp[i-1][j-1]
# If last character are different, consider all
# possibilities and find minimum
else:
dp[i][j] = 1 + min(dp[i][j-1],
# Insert
dp[i-1][j],
# Remove
dp[i-1][j-1])
# Replace
return dp[m][n]
# Driver program
str1 = "sunday"
str2 = "saturday"
print(editDistDP(str1, str2, len(str1), len(str2)))
# This code is contributed by Bhavya Jain

Output:
3

Time Complexity: O(m x n)

Auxiliary Space: O(m x n)
Applications: There are many practical applications of edit distance algorithm, refer Lucene API for sample. Another example, display all the
words in a dictionary that are near proximity to a given word\incorrectly spelled word.
Thanks to Vivek Kumar for suggesting above updates.

Dynamic Programming | Set 6 (Min Cost Path)

Given a cost matrix cost[][] and a position (m, n) in cost[][], write a function that returns cost of minimum cost path to reach (m, n) from (0, 0).
Each cell of the matrix represents a cost to traverse through that cell. Total cost of a path to reach (m, n) is sum of all the costs on that path
(including both source and destination). You can only traverse down, right and diagonally lower cells from a given cell, i.e., from a given cell (i, j),
cells (i+1, j), (i, j+1) and (i+1, j+1) can be traversed. You may assume that all costs are positive integers.
For example, in the following figure, what is the minimum cost path to (2, 2)?

The path with minimum cost is highlighted in the following figure. The path is (0, 0) > (0, 1) > (1, 2) > (2, 2). The cost of the path is 8 (1 + 2 + 2 +
3).

1) Optimal Substructure
The path to reach (m, n) must be through one of the 3 cells: (m-1, n-1) or (m-1, n) or (m, n-1). So minimum cost to reach (m, n) can be written as
minimum of the 3 cells plus cost[m][n].
minCost(m, n) = min (minCost(m-1, n-1), minCost(m-1, n), minCost(m, n-1)) + cost[m][n]
2) Overlapping Subproblems
Following is simple recursive implementation of the MCP (Minimum Cost Path) problem. The implementation simply follows the recursive structure
mentioned above.
/* A Naive recursive implementation of MCP(Minimum Cost Path) problem */
#include<stdio.h>
#include<limits.h>
#define R 3
#define C 3
int min(int x, int y, int z);
/* Returns cost of minimum cost path from (0,0) to (m, n) in mat[R][C]*/
int minCost(int cost[R][C], int m, int n)
{
if (n < 0 || m < 0)
return INT_MAX;
else if (m == 0 && n == 0)
return cost[m][n];
else
return cost[m][n] + min( minCost(cost, m-1, n-1),
minCost(cost, m-1, n),
minCost(cost, m, n-1) );
}
/* A utility function
int min(int x, int y,
{
if (x < y)
return (x < z)?
else
return (y < z)?
}

that returns minimum of 3 integers */

int z)
x : z;
y : z;

/* Driver program to test above functions */

int main()
{
int cost[R][C] = { {1, 2, 3},
{4, 8, 2},

{1, 5, 3} };
printf(" %d ", minCost(cost, 2, 2));
return 0;
}

It should be noted that the above function computes the same subproblems again and again. See the following recursion tree, there are many nodes
which apear more than once. Time complexity of this naive recursive solution is exponential and it is terribly slow.
mC refers to minCost()

mC(1,
/
|
/
|
mC(0,0) mC(0,1)

mC(2, 2)
/
|
\
/
|
\
1)
mC(1, 2)
mC(2,
\
/
|
\
/
\
/
|
\
/
mC(1,0) mC(0,1) mC(0,2) mC(1,1) mC(1,0)

1)
|
\
|
\
mC(1,1) mC(2,0)

So the MCP problem has both properties (see this and this) of a dynamic programming problem. Like other typical Dynamic Programming(DP)
problems, recomputations of same subproblems can be avoided by constructing a temporary array tc[][] in bottom up manner.

C++
/* Dynamic Programming implementation of MCP problem */
#include<stdio.h>
#include<limits.h>
#define R 3
#define C 3
int min(int x, int y, int z);
int minCost(int cost[R][C], int m, int n)
{
int i, j;
// Instead of following line, we can use int tc[m+1][n+1] or
// dynamically allocate memoery to save space. The following line is
// used to keep te program simple and make it working on all compilers.
int tc[R][C];
tc[0][0] = cost[0][0];
/* Initialize first column of total cost(tc) array */
for (i = 1; i <= m; i++)
tc[i][0] = tc[i-1][0] + cost[i][0];
/* Initialize first row of tc array */
for (j = 1; j <= n; j++)
tc[0][j] = tc[0][j-1] + cost[0][j];
/* Construct rest of the tc array */
for (i = 1; i <= m; i++)
for (j = 1; j <= n; j++)
tc[i][j] = min(tc[i-1][j-1], tc[i-1][j], tc[i][j-1]) + cost[i][j];
return tc[m][n];
}
/* A utility function
int min(int x, int y,
{
if (x < y)
return (x < z)?
else
return (y < z)?
}

that returns minimum of 3 integers */

int z)
x : z;
y : z;

/* Driver program to test above functions */

int main()
{
int cost[R][C] = { {1, 2, 3},
{4, 8, 2},
{1, 5, 3} };
printf(" %d ", minCost(cost, 2, 2));
return 0;
}

Python

#
#
R
C

Dynamic Programming Python implementation of Min Cost Path

problem
= 3
= 3

def minCost(cost, m, n):

# Instead of following line, we can use int tc[m+1][n+1] or
# dynamically allocate memoery to save space. The following
# line is used to keep te program simple and make it working
# on all compilers.
tc = [[0 for x in range(C)] for x in range(R)]
tc[0][0] = cost[0][0]
# Initialize first column of total cost(tc) array
for i in range(1, m+1):
tc[i][0] = tc[i-1][0] + cost[i][0]
# Initialize first row of tc array
for j in range(1, n+1):
tc[0][j] = tc[0][j-1] + cost[0][j]
# Construct rest of the tc array
for i in range(1, m+1):
for j in range(1, n+1):
tc[i][j] = min(tc[i-1][j-1], tc[i-1][j], tc[i][j-1]) + cost[i][j]
return tc[m][n]
# Driver program to test above functions
cost = [[1, 2, 3],
[4, 8, 2],
[1, 5, 3]]
print(minCost(cost, 2, 2))
# This code is contributed by Bhavya Jain

Time Complexity of the DP implementation is O(mn) which is much better than Naive Recursive implementation.

Dynamic Programming | Set 7 (Coin Change)

Given a value N, if we want to make change for N cents, and we have infinite supply of each of S = { S1, S2, .. , Sm} valued coins, how many
ways can we make the change? The order of coins doesnt matter.
For example, for N = 4 and S = {1,2,3}, there are four solutions: {1,1,1,1},{1,1,2},{2,2},{1,3}. So output should be 4. For N = 10 and S = {2,
5, 3, 6}, there are five solutions: {2,2,2,2,2}, {2,2,3,3}, {2,2,6}, {2,3,5} and {5,5}. So the output should be 5.
1) Optimal Substructure
To count total number solutions, we can divide all set solutions in two sets.
1) Solutions that do not contain mth coin (or Sm).
2) Solutions that contain at least one Sm.
Let count(S[], m, n) be the function to count the number of solutions, then it can be written as sum of count(S[], m-1, n) and count(S[], m, n-Sm).
Therefore, the problem has optimal substructure property as the problem can be solved using solutions to subproblems.
2) Overlapping Subproblems
Following is a simple recursive implementation of the Coin Change problem. The implementation simply follows the recursive structure mentioned
above.
#include<stdio.h>
// Returns the count of ways we can sum S[0...m-1] coins to get sum n
int count( int S[], int m, int n )
{
// If n is 0 then there is 1 solution (do not include any coin)
if (n == 0)
return 1;
// If n is less than 0 then no solution exists
if (n < 0)
return 0;
// If there are no coins and n is greater than 0, then no solution exist
if (m <=0 && n >= 1)
return 0;
// count is sum of solutions (i) including S[m-1] (ii) excluding S[m-1]
return count( S, m - 1, n ) + count( S, m, n-S[m-1] );
}
// Driver program to test above function
int main()
{
int i, j;
int arr[] = {1, 2, 3};
int m = sizeof(arr)/sizeof(arr[0]);
printf("%d ", count(arr, m, 4));
getchar();
return 0;
}

It should be noted that the above function computes the same subproblems again and again. See the following recursion tree for S = {1, 2, 3} and
n = 5.
The function C({1}, 3) is called two times. If we draw the complete tree, then we can see that there are many subproblems being called more than
once.
C() --> count()
C({1,2,3}, 5)
/

\
/
\
C({1,2,3}, 2)
C({1,2}, 5)
/
\
/
\
/
\
/
\
C({1,2,3}, -1) C({1,2}, 2)
C({1,2}, 3)
C({1}, 5)
/
\
/
\
/
\
/
\
/
\
/
\
C({1,2},0) C({1},2) C({1,2},1) C({1},3)
C({1}, 4) C({}, 5)
/ \
/ \
/ \
/
\
/ \
/ \
/ \
/
\
.
. .
. .
. C({1}, 3) C({}, 4)
/ \
/
\
.
.

Since same suproblems are called again, this problem has Overlapping Subprolems property. So the Coin Change problem has both properties

(see this and this) of a dynamic programming problem. Like other typical Dynamic Programming(DP) problems, recomputations of same
subproblems can be avoided by constructing a temporary array table[][] in bottom up manner.
Dynamic Programming Solution

C
#include<stdio.h>
int count( int S[], int m, int n )
{
int i, j, x, y;
// We need n+1 rows as the table is consturcted in bottom up manner using
// the base case 0 value case (n = 0)
int table[n+1][m];
// Fill the enteries for 0 value case (n = 0)
for (i=0; i<m; i++)
table[0][i] = 1;
// Fill rest of the table enteries in bottom up manner
for (i = 1; i < n+1; i++)
{
for (j = 0; j < m; j++)
{
// Count of solutions including S[j]
x = (i-S[j] >= 0)? table[i - S[j]][j]: 0;
// Count of solutions excluding S[j]
y = (j >= 1)? table[i][j-1]: 0;
// total count
table[i][j] = x + y;
}
}
return table[n][m-1];
}
// Driver program to test above function
int main()
{
int arr[] = {1, 2, 3};
int m = sizeof(arr)/sizeof(arr[0]);
int n = 4;
printf(" %d ", count(arr, m, n));
return 0;
}

Python
# Dynamic Programming Python implementation of Coin Change problem
def count(S, m, n):
# We need n+1 rows as the table is consturcted in bottom up
# manner using the base case 0 value case (n = 0)
table = [[0 for x in range(m)] for x in range(n+1)]
# Fill the enteries for 0 value case (n = 0)
for i in range(m):
table[0][i] = 1
# Fill rest of the table enteries in bottom up manner
for i in range(1, n+1):
for j in range(m):
# Count of solutions including S[j]
x = table[i - S[j]][j] if i-S[j] >= 0 else 0
# Count of solutions excluding S[j]
y = table[i][j-1] if j >= 1 else 0
# total count
table[i][j] = x + y
return table[n][m-1]
# Driver program to test above function
arr = [1, 2, 3]
m = len(arr)

n = 4
print(count(arr, m, n))
# This code is contributed by Bhavya Jain

Time Complexity: O(mn)

Following is a simplified version of method 2. The auxiliary space required here is O(n) only.
int count( int S[], int m, int n )
{
// table[i] will be storing the number of solutions for
// value i. We need n+1 rows as the table is consturcted
// in bottom up manner using the base case (n = 0)
int table[n+1];
// Initialize all table values as 0
memset(table, 0, sizeof(table));
// Base case (If given value is 0)
table[0] = 1;
// Pick all coins one by one and update the table[] values
// after the index greater than or equal to the value of the
// picked coin
for(int i=0; i<m; i++)
for(int j=S[i]; j<=n; j++)
table[j] += table[j-S[i]];
return table[n];
}

Thanks to Rohan Laishram for suggesting this space optimized version.

References:
http://www.algorithmist.com/index.php/Coin_Change

Dynamic Programming | Set 8 (Matrix Chain Multiplication)

Given a sequence of matrices, find the most efficient way to multiply these matrices together. The problem is not actually to perform the
multiplications, but merely to decide in which order to perform the multiplications.
We have many options to multiply a chain of matrices because matrix multiplication is associative. In other words, no matter how we parenthesize
the product, the result will be the same. For example, if we had four matrices A, B, C, and D, we would have:
(ABC)D = (AB)(CD) = A(BCD) = ....

However, the order in which we parenthesize the product affects the number of simple arithmetic operations needed to compute the product, or
the efficiency. For example, suppose A is a 10 30 matrix, B is a 30 5 matrix, and C is a 5 60 matrix. Then,
(AB)C = (10305) + (10560) = 1500 + 3000 = 4500 operations
A(BC) = (30560) + (103060) = 9000 + 18000 = 27000 operations.

Clearly the first parenthesization requires less number of operations.

Given an array p[] which represents the chain of matrices such that the ith matrix Ai is of dimension p[i-1] x p[i]. We need to write a
function MatrixChainOrder() that should return the minimum number of multiplications needed to multiply the chain.
Input: p[] = {40, 20, 30, 10, 30}
Output: 26000
There are 4 matrices of dimensions 40x20, 20x30, 30x10 and 10x30.
Let the input 4 matrices be A, B, C and D. The minimum number of
multiplications are obtained by putting parenthesis in following way
(A(BC))D --> 20*30*10 + 40*20*10 + 40*10*30
Input: p[] = {10, 20, 30, 40, 30}
Output: 30000
There are 4 matrices of dimensions 10x20, 20x30, 30x40 and 40x30.
Let the input 4 matrices be A, B, C and D. The minimum number of
multiplications are obtained by putting parenthesis in following way
((AB)C)D --> 10*20*30 + 10*30*40 + 10*40*30
Input: p[] = {10, 20, 30}
Output: 6000
There are only two matrices of dimensions 10x20 and 20x30. So there
is only one way to multiply the matrices, cost of which is 10*20*30

1) Optimal Substructure:
A simple solution is to place parenthesis at all possible places, calculate the cost for each placement and return the minimum value. In a chain of
matrices of size n, we can place the first set of parenthesis in n-1 ways. For example, if the given chain is of 4 matrices. let the chain be ABCD,
then there are 3 way to place first set of parenthesis: A(BCD), (AB)CD and (ABC)D. So when we place a set of parenthesis, we divide the
problem into subproblems of smaller size. Therefore, the problem has optimal substructure property and can be easily solved using recursion.
Minimum number of multiplication needed to multiply a chain of size n = Minimum of all n-1 placements (these placements create subproblems of
smaller size)
2) Overlapping Subproblems
Following is a recursive implementation that simply follows the above optimal substructure property.
/* A naive recursive implementation that simply follows the above optimal
substructure property */
#include<stdio.h>
#include<limits.h>
// Matrix Ai has dimension p[i-1] x p[i] for i = 1..n
int MatrixChainOrder(int p[], int i, int j)
{
if(i == j)
return 0;
int k;
int min = INT_MAX;
int count;
// place parenthesis at different places between first and last matrix,
// recursively calculate count of multiplcations for each parenthesis
// placement and return the minimum count
for (k = i; k <j; k++)
{
count = MatrixChainOrder(p, i, k) +
MatrixChainOrder(p, k+1, j) +
p[i-1]*p[k]*p[j];
if (count < min)

min = count;
}
// Return minimum count
return min;
}
// Driver program to test above function
int main()
{
int arr[] = {1, 2, 3, 4, 3};
int n = sizeof(arr)/sizeof(arr[0]);
printf("Minimum number of multiplications is %d ",
MatrixChainOrder(arr, 1, n-1));
getchar();
return 0;
}

Time complexity of the above naive recursive approach is exponential. It should be noted that the above function computes the same subproblems
again and again. See the following recursion tree for a matrix chain of size 4. The function MatrixChainOrder(p, 3, 4) is called two times. We can
see that there are many subproblems being called more than once.

Since same suproblems are called again, this problem has Overlapping Subprolems property. So Matrix Chain Multiplication problem has both
properties (see this and this) of a dynamic programming problem. Like other typical Dynamic Programming(DP) problems, recomputations of
same subproblems can be avoided by constructing a temporary array m[][] in bottom up manner.
Dynamic Programming Solution
Following is C/C++ implementation for Matrix Chain Multiplication problem using Dynamic Programming.

C
// See the Cormen book for details of the following algorithm
#include<stdio.h>
#include<limits.h>
// Matrix Ai has dimension p[i-1] x p[i] for i = 1..n
int MatrixChainOrder(int p[], int n)
{
/* For simplicity of the program, one extra row and one extra column are
allocated in m[][]. 0th row and 0th column of m[][] are not used */
int m[n][n];
int i, j, k, L, q;
/* m[i,j] = Minimum number of scalar multiplications needed to compute
the matrix A[i]A[i+1]...A[j] = A[i..j] where dimention of A[i] is
p[i-1] x p[i] */
// cost is zero when multiplying one matrix.
for (i = 1; i < n; i++)
m[i][i] = 0;
// L is chain length.
for (L=2; L<n; L++)
{
for (i=1; i<=n-L+1; i++)
{
j = i+L-1;
m[i][j] = INT_MAX;
for (k=i; k<=j-1; k++)
{
// q = cost/scalar multiplications
q = m[i][k] + m[k+1][j] + p[i-1]*p[k]*p[j];

if (q < m[i][j])
m[i][j] = q;
}
}
}
return m[1][n-1];
}
int main()
{
int arr[] = {1, 2, 3, 4};
int size = sizeof(arr)/sizeof(arr[0]);
printf("Minimum number of multiplications is %d ",
MatrixChainOrder(arr, size));
getchar();
return 0;
}

Python
# Dynamic Programming Python implementation of Matrix Chain Multiplication
# See the Cormen book for details of the following algorithm
import sys
# Matrix Ai has dimension p[i-1] x p[i] for i = 1..n
def MatrixChainOrder(p, n):
# For simplicity of the program, one extra row and one extra column are
# allocated in m[][]. 0th row and 0th column of m[][] are not used
m = [[0 for x in range(n)] for x in range(n)]
# m[i,j] = Minimum number of scalar multiplications needed to compute
# the matrix A[i]A[i+1]...A[j] = A[i..j] where dimention of A[i] is
# p[i-1] x p[i]
# cost is zero when multiplying one matrix.
for i in range(1, n):
m[i][i] = 0
# L is chain length.
for L in range(2, n):
for i in range(1, n-L+1):
j = i+L-1
m[i][j] = sys.maxint
for k in range(i, j):
# q = cost/scalar multiplications
q = m[i][k] + m[k+1][j] + p[i-1]*p[k]*p[j]
if q < m[i][j]:
m[i][j] = q
return m[1][n-1]
# Driver program to test above function
arr = [1, 2, 3 ,4]
size = len(arr)
print("Minimum number of multiplications is " + str(MatrixChainOrder(arr, size)))
# This Code is contributed by Bhavya Jain

Minimum number of multiplications is 18

Time Complexity: O(n^3)

Auxiliary Space: O(n^2)
References:
http://en.wikipedia.org/wiki/Matrix_chain_multiplication
http://www.personal.kent.edu/~rmuhamma/Algorithms/MyAlgorithms/Dynamic/chainMatrixMult.htm

Dynamic Programming | Set 9 (Binomial Coefficient)

Following are common definition of Binomial Coefficients.
1) A binomial coefficient C(n, k) can be defined as the coefficient of X^k in the expansion of (1 + X)^n.
2) A binomial coefficient C(n, k) also gives the number of ways, disregarding order, that k objects can be chosen from among n objects; more
formally, the number of k-element subsets (or k-combinations) of an n-element set.
The Problem
Write a function that takes two parameters n and k and returns the value of Binomial Coefficient C(n, k). For example, your function
should return 6 for n = 4 and k = 2, and it should return 10 for n = 5 and k = 2.
1) Optimal Substructure
The value of C(n, k) can recursively calculated using following standard formula for Binomial Cofficients.
C(n, k) = C(n-1, k-1) + C(n-1, k)
C(n, 0) = C(n, n) = 1

2) Overlapping Subproblems
Following is simple recursive implementation that simply follows the recursive structure mentioned above.
// A Naive Recursive Implementation
#include<stdio.h>
// Returns value of Binomial Coefficient C(n, k)
int binomialCoeff(int n, int k)
{
// Base Cases
if (k==0 || k==n)
return 1;
// Recur
return binomialCoeff(n-1, k-1) + binomialCoeff(n-1, k);
}
/* Drier program to test above function*/
int main()
{
int n = 5, k = 2;
printf("Value of C(%d, %d) is %d ", n, k, binomialCoeff(n, k));
return 0;
}

It should be noted that the above function computes the same subproblems again and again. See the following recursion tree for n = 5 an k = 2.
The function C(3, 1) is called two times. For large values of n, there will be many common subproblems.
C(5, 2)
/

\
C(4, 1)
C(4, 2)
/ \
/
\
C(3, 0) C(3, 1)
C(3, 1)
C(3, 2)
/
\
/
\
/
\
C(2, 0)
C(2, 1)
C(2, 0) C(2, 1)
C(2, 1) C(2, 2)
/
\
/ \
/
\
C(1, 0) C(1, 1)
C(1, 0) C(1, 1) C(1, 0) C(1, 1)

Since same suproblems are called again, this problem has Overlapping Subprolems property. So the Binomial Coefficient problem has both
properties (see this and this) of a dynamic programming problem. Like other typical Dynamic Programming(DP) problems, recomputations of
same subproblems can be avoided by constructing a temporary array C[][] in bottom up manner. Following is Dynamic Programming based
implementation.

C
// A Dynamic Programming based solution that uses table C[][] to
// calculate the Binomial Coefficient
#include<stdio.h>
// Prototype of a utility function that returns minimum of two integers
int min(int a, int b);
// Returns value of Binomial Coefficient C(n, k)
int binomialCoeff(int n, int k)
{
int C[n+1][k+1];
int i, j;

// Caculate value of Binomial Coefficient in bottom up manner

for (i = 0; i <= n; i++)
{
for (j = 0; j <= min(i, k); j++)
{
// Base Cases
if (j == 0 || j == i)
C[i][j] = 1;
// Calculate value using previosly stored values
else
C[i][j] = C[i-1][j-1] + C[i-1][j];
}
}
return C[n][k];
}
// A utility function to return minimum of two integers
int min(int a, int b)
{
return (a<b)? a: b;
}
/* Drier program to test above function*/
int main()
{
int n = 5, k = 2;
printf ("Value of C(%d, %d) is %d ", n, k, binomialCoeff(n, k) );
return 0;
}

Python
# A Dynamic Programming based Python Program that uses table C[][]
# to calculate the Binomial Coefficient
# Returns value of Binomial Coefficient C(n, k)
def binomialCoef(n, k):
C = [[0 for x in range(k+1)] for x in range(n+1)]
# Calculate value of Binomial Coefficient in bottom up manner
for i in range(n+1):
for j in range(min(i, k)+1):
# Base Cases
if j == 0 or j == i:
C[i][j] = 1
# Calculate value using previosly stored values
else:
C[i][j] = C[i-1][j-1] + C[i-1][j]
return C[n][k]
# Driver program to test above function
n = 5
k = 2
print("Value of C[" + str(n) + "][" + str(k) + "] is "
+ str(binomialCoef(n,k)))
# This code is contributed by Bhavya Jain

Value of C[5][2] is 10

Time Complexity: O(n*k)

Auxiliary Space: O(n*k)
Following is a space optimized version of the above code. The following code only uses O(k). Thanks to AK for suggesting this method.
// A space optimized Dynamic Programming Solution
int binomialCoeff(int n, int k)
{
int* C = (int*)calloc(k+1, sizeof(int));
int i, j, res;
C[0] = 1;

for(i = 1; i <= n; i++)

{
for(j = min(i, k); j > 0; j--)
C[j] = C[j] + C[j-1];
}
res = C[k]; // Store the result before freeing memory
free(C); // free dynamically allocated memory to avoid memory leak
return res;
}

Time Complexity: O(n*k)

Auxiliary Space: O(k)
References:
http://www.csl.mtu.edu/cs4321/www/Lectures/Lecture%2015%20-%20Dynamic%20Programming%20Binomial%20Coefficients.htm

Dynamic Programming | Set 10 ( 0-1 Knapsack Problem)

Given weights and values of n items, put these items in a knapsack of capacity W to get the maximum total value in the knapsack. In other words,
given two integer arrays val[0..n-1] and wt[0..n-1] which represent values and weights associated with n items respectively. Also given an integer
W which represents knapsack capacity, find out the maximum value subset of val[] such that sum of the weights of this subset is smaller than or
equal to W. You cannot break an item, either pick the complete item, or dont pick it (0-1 property).
A simple solution is to consider all subsets of items and calculate the total weight and value of all subsets. Consider the only subsets whose total
weight is smaller than W. From all such subsets, pick the maximum value subset.
1) Optimal Substructure:
To consider all subsets of items, there can be two cases for every item: (1) the item is included in the optimal subset, (2) not included in the optimal
set.
Therefore, the maximum value that can be obtained from n items is max of following two values.
1) Maximum value obtained by n-1 items and W weight (excluding nth item).
2) Value of nth item plus maximum value obtained by n-1 items and W minus weight of the nth item (including nth item).
If weight of nth item is greater than W, then the nth item cannot be included and case 1 is the only possibility.
2) Overlapping Subproblems
Following is recursive implementation that simply follows the recursive structure mentioned above.

C/C++
/* A Naive recursive implementation of 0-1 Knapsack problem */
#include<stdio.h>
// A utility function that returns maximum of two integers
int max(int a, int b) { return (a > b)? a : b; }
// Returns the maximum value that can be put in a knapsack of capacity W
int knapSack(int W, int wt[], int val[], int n)
{
// Base Case
if (n == 0 || W == 0)
return 0;
// If weight of the nth item is more than Knapsack capacity W, then
// this item cannot be included in the optimal solution
if (wt[n-1] > W)
return knapSack(W, wt, val, n-1);
// Return the maximum of two cases:
// (1) nth item included
// (2) not included
else return max( val[n-1] + knapSack(W-wt[n-1], wt, val, n-1),
knapSack(W, wt, val, n-1)
);
}
// Driver program to test above function
int main()
{
int val[] = {60, 100, 120};
int wt[] = {10, 20, 30};
int W = 50;
int n = sizeof(val)/sizeof(val[0]);
printf("%d", knapSack(W, wt, val, n));
return 0;
}

Python
#A naive recursive implementation of 0-1 Knapsack Problem
# Returns the maximum value that can be put in a knapsack of
# capacity W
def knapSack(W , wt , val , n):
# Base Case
if n == 0 or W == 0 :
return 0
# If weight of the nth item is more than Knapsack of capacity
# W, then this item cannot be included in the optimal solution
if (wt[n-1] > W):

return knapSack(W , wt , val , n-1)

# return the maximum of two cases:
# (1) nth item included
# (2) not included
else:
return max(val[n-1] + knapSack(W-wt[n-1] , wt , val , n-1),
knapSack(W , wt , val , n-1))
# end of function knapSack
# To test above function
val = [60, 100, 120]
wt = [10, 20, 30]
W = 50
n = len(val)
print knapSack(W , wt , val , n)
# This code is contributed by Nikhil Kumar Singh

220

It should be noted that the above function computes the same subproblems again and again. See the following recursion tree, K(1, 1) is being
evaluated twice. Time complexity of this naive recursive solution is exponential (2^n).
In the following recursion tree, K() refers to knapSack(). The two
parameters indicated in the following recursion tree are n and W.
The recursion tree is for following sample inputs.
wt[] = {1, 1, 1}, W = 2, val[] = {10, 20, 30}
K(3, 2)
/
/
K(2,2)
/
\
/
K(1,2)
/ \
/
\
K(0,2) K(0,1)
Recursion tree

---------> K(n, W)
\
\

K(2,1)
/
\
\
/
\
K(1,1)
K(1,1)
K(1,0)
/ \
/ \
/
\
/
\
K(0,1) K(0,0) K(0,1) K(0,0)
for Knapsack capacity 2 units and 3 items of 1 unit weight.

Since suproblems are evaluated again, this problem has Overlapping Subprolems property. So the 0-1 Knapsack problem has both properties
(see this and this) of a dynamic programming problem. Like other typical Dynamic Programming(DP) problems, recomputations of same
subproblems can be avoided by constructing a temporary array K[][] in bottom up manner. Following is Dynamic Programming based
implementation.

C++
// A Dynamic Programming based solution for 0-1 Knapsack problem
#include<stdio.h>
// A utility function that returns maximum of two integers
int max(int a, int b) { return (a > b)? a : b; }
// Returns the maximum value that can be put in a knapsack of capacity W
int knapSack(int W, int wt[], int val[], int n)
{
int i, w;
int K[n+1][W+1];
// Build table K[][] in bottom up manner
for (i = 0; i <= n; i++)
{
for (w = 0; w <= W; w++)
{
if (i==0 || w==0)
K[i][w] = 0;
else if (wt[i-1] <= w)
K[i][w] = max(val[i-1] + K[i-1][w-wt[i-1]], K[i-1][w]);
else
K[i][w] = K[i-1][w];
}
}
return K[n][W];
}
int main()

{
int val[] = {60, 100, 120};
int wt[] = {10, 20, 30};
int W = 50;
int n = sizeof(val)/sizeof(val[0]);
printf("%d", knapSack(W, wt, val, n));
return 0;
}

Pyhton
# A Dynamic Programming based Python Program for 0-1 Knapsack problem
# Returns the maximum value that can be put in a knapsack of capacity W
def knapSack(W, wt, val, n):
K = [[0 for x in range(W+1)] for x in range(n+1)]
# Build table K[][] in bottom up manner
for i in range(n+1):
for w in range(W+1):
if i==0 or w==0:
K[i][w] = 0
elif wt[i-1] <= w:
K[i][w] = max(val[i-1] + K[i-1][w-wt[i-1]], K[i-1][w])
else:
K[i][w] = K[i-1][w]
return K[n][W]
# Driver program to test above function
val = [60, 100, 120]
wt = [10, 20, 30]
W = 50
n = len(val)
print(knapSack(W, wt, val, n))
# This code is contributed by Bhavya Jain

220

Time Complexity: O(nW) where n is the number of items and W is the capacity of knapsack.
References:
http://www.es.ele.tue.nl/education/5MC10/Solutions/knapsack.pdf
http://www.cse.unl.edu/~goddard/Courses/CSCE310J/Lectures/Lecture8-DynamicProgramming.pdf

Dynamic Programming | Set 11 (Egg Dropping Puzzle)

The following is a description of the instance of this famous puzzle involving n=2 eggs and a building with k=36 floors.
Suppose that we wish to know which stories in a 36-story building are safe to drop eggs from, and which will cause the eggs to break on landing.
We make a few assumptions:
..An egg that survives a fall can be used again.
..A broken egg must be discarded.
..The effect of a fall is the same for all eggs.
..If an egg breaks when dropped, then it would break if dropped from a higher floor.
..If an egg survives a fall then it would survive a shorter fall.
..It is not ruled out that the first-floor windows break eggs, nor is it ruled out that the 36th-floor do not cause an egg to break.
If only one egg is available and we wish to be sure of obtaining the right result, the experiment can be carried out in only one way. Drop the egg
from the first-floor window; if it survives, drop it from the second floor window. Continue upward until it breaks. In the worst case, this method
may require 36 droppings. Suppose 2 eggs are available. What is the least number of egg-droppings that is guaranteed to work in all cases?
The problem is not actually to find the critical floor, but merely to decide floors from which eggs should be dropped so that total number of trials
are minimized.
Source: Wiki for Dynamic Programming
In this post, we will discuss solution to a general problem with n eggs and k floors. The solution is to try dropping an egg from every floor (from 1
to k) and recursively calculate the minimum number of droppings needed in worst case. The floor which gives the minimum value in worst case is
going to be part of the solution.
In the following solutions, we return the minimum number of trails in worst case; these solutions can be easily modified to print floor numbers of
every trials also.
1) Optimal Substructure:
When we drop an egg from a floor x, there can be two cases (1) The egg breaks (2) The egg doesnt break.
1) If the egg breaks after dropping from xth floor, then we only need to check for floors lower than x with remaining eggs; so the problem reduces
to x-1 floors and n-1 eggs
2) If the egg doesnt break after dropping from the xth floor, then we only need to check for floors higher than x; so the problem reduces to k-x
floors and n eggs.
Since we need to minimize the number of trials in worst case, we take the maximum of two cases. We consider the max of above two cases for
every floor and choose the floor which yields minimum number of trials.
k ==> Number of floors
n ==> Number of Eggs
eggDrop(n, k) ==> Minimum number of trails needed to find the critical
floor in worst case.
eggDrop(n, k) = 1 + min{max(eggDrop(n - 1, x - 1), eggDrop(n, k - x)):
x in {1, 2, ..., k}}

2) Overlapping Subproblems
Following is recursive implementation that simply follows the recursive structure mentioned above.
# include <stdio.h>
# include <limits.h>
// A utility function to get maximum of two integers
int max(int a, int b) { return (a > b)? a: b; }
/* Function to get minimum number of trails needed in worst
case with n eggs and k floors */
int eggDrop(int n, int k)
{
// If there are no floors, then no trials needed. OR if there is
// one floor, one trial needed.
if (k == 1 || k == 0)
return k;
// We need k trials for one egg and k floors
if (n == 1)
return k;
int min = INT_MAX, x, res;
// Consider all droppings from 1st floor to kth floor and
// return the minimum of these values plus 1.
for (x = 1; x <= k; x++)

{
res = max(eggDrop(n-1, x-1), eggDrop(n, k-x));
if (res < min)
min = res;
}
return min + 1;
}
/* Driver program to test to pront printDups*/
int main()
{
int n = 2, k = 10;
printf ("\nMinimum number of trials in worst case with %d eggs and "
"%d floors is %d \n", n, k, eggDrop(n, k));
return 0;
}
Output:
Minimum number of trials in worst case with 2 eggs and 10 floors is 4

It should be noted that the above function computes the same subproblems again and again. See the following partial recursion tree, E(2, 2) is
being evaluated twice. There will many repeated subproblems when you draw the complete recursion tree even for small values of n and k.
E(2,4)
|
------------------------------------|
|
|
|
|
|
|
|
x=1/\
x=2/\
x=3/ \
x=4/ \
/ \
/ \
....
....
/
\
/
\
E(1,0) E(2,3)
E(1,1) E(2,2)
/\ /\...
/ \
x=1/ \
.....
/
\
E(1,0) E(2,2)
/ \
......
Partial recursion tree for 2 eggs and 4 floors.

Since same suproblems are called again, this problem has Overlapping Subprolems property. So Egg Dropping Puzzle has both properties (see
this and this) of a dynamic programming problem. Like other typical Dynamic Programming(DP) problems, recomputations of same subproblems
can be avoided by constructing a temporary array eggFloor[][] in bottom up manner.
Dynamic Programming Solution
Following are C++ and Python implementations for Egg Dropping problem using Dynamic Programming.

C++
# A Dynamic Programming based C++ Program for the Egg Dropping Puzzle
# include <stdio.h>
# include <limits.h>
// A utility function to get maximum of two integers
int max(int a, int b) { return (a > b)? a: b; }
/* Function to get minimum number of trails needed in worst
case with n eggs and k floors */
int eggDrop(int n, int k)
{
/* A 2D table where entery eggFloor[i][j] will represent minimum
number of trials needed for i eggs and j floors. */
int eggFloor[n+1][k+1];
int res;
int i, j, x;
// We need one trial for one floor and0 trials for 0 floors
for (i = 1; i <= n; i++)
{
eggFloor[i][1] = 1;
eggFloor[i][0] = 0;
}
// We always need j trials for one egg and j floors.
for (j = 1; j <= k; j++)
eggFloor[1][j] = j;

// Fill rest of the entries in table using optimal substructure

// property
for (i = 2; i <= n; i++)
{
for (j = 2; j <= k; j++)
{
eggFloor[i][j] = INT_MAX;
for (x = 1; x <= j; x++)
{
res = 1 + max(eggFloor[i-1][x-1], eggFloor[i][j-x]);
if (res < eggFloor[i][j])
eggFloor[i][j] = res;
}
}
}
// eggFloor[n][k] holds the result
return eggFloor[n][k];
}
/* Driver program to test to pront printDups*/
int main()
{
int n = 2, k = 36;
printf ("\nMinimum number of trials in worst case with %d eggs and "
"%d floors is %d \n", n, k, eggDrop(n, k));
return 0;
}

Python
# A Dynamic Programming based Python Program for the Egg Dropping Puzzle
INT_MAX = 32767
# Function to get minimum number of trails needed in worst
# case with n eggs and k floors
def eggDrop(n, k):
# A 2D table where entery eggFloor[i][j] will represent minimum
# number of trials needed for i eggs and j floors.
eggFloor = [[0 for x in range(k+1)] for x in range(n+1)]
# We need one trial for one floor and0 trials for 0 floors
for i in range(1, n+1):
eggFloor[i][1] = 1
eggFloor[i][0] = 0
# We always need j trials for one egg and j floors.
for j in range(1, k+1):
eggFloor[1][j] = j
# Fill rest of the entries in table using optimal substructure
# property
for i in range(2, n+1):
for j in range(2, k+1):
eggFloor[i][j] = INT_MAX
for x in range(1, j+1):
res = 1 + max(eggFloor[i-1][x-1], eggFloor[i][j-x])
if res < eggFloor[i][j]:
eggFloor[i][j] = res
# eggFloor[n][k] holds the result
return eggFloor[n][k]
# Driver program to test to pront printDups
n = 2
k = 36
print("Minimum number of trials in worst case with" + str(n) + "eggs and " \
+ str(k) + " floors is " + str(eggDrop(n, k)))
# This code is contributed by Bhavya Jain
Output:
Minimum number of trials in worst case with 2 eggs and 36 floors is 8

Time Complexity: O(nk^2)

Auxiliary Space: O(nk)
As an exercise, you may try modifying the above DP solution to print all intermediate floors (The floors used for minimum trail solution).

References:
http://archive.ite.journal.informs.org/Vol4No1/Sniedovich/index.php

Dynamic Programming | Set 12 (Longest Palindromic Subsequence)

Given a sequence, find the length of the longest palindromic subsequence in it. For example, if the given sequence is BBABCBCAB, then the
output should be 7 as BABCBAB is the longest palindromic subseuqnce in it. BBBBB and BBCBB are also palindromic subsequences of the
given sequence, but not the longest ones.
The naive solution for this problem is to generate all subsequences of the given sequence and find the longest palindromic subsequence. This
solution is exponential in term of time complexity. Let us see how this problem possesses both important properties of a Dynamic Programming
(DP) Problem and can efficiently solved using Dynamic Programming.
1) Optimal Substructure:
Let X[0..n-1] be the input sequence of length n and L(0, n-1) be the length of the longest palindromic subsequence of X[0..n-1].
If last and first characters of X are same, then L(0, n-1) = L(1, n-2) + 2.
Else L(0, n-1) = MAX (L(1, n-1), L(0, n-2)).
Following is a general recursive solution with all cases handled.
// Everay single character is a palindrom of length 1
L(i, i) = 1 for all indexes i in given sequence
// IF first and last characters are not same
If (X[i] != X[j]) L(i, j) = max{L(i + 1, j),L(i, j - 1)}
// If there are only 2 characters and both are same
Else if (j == i + 1) L(i, j) = 2
// If there are more than two characters, and first and last
// characters are same
Else L(i, j) = L(i + 1, j - 1) + 2

2) Overlapping Subproblems
Following is simple recursive implementation of the LPS problem. The implementation simply follows the recursive structure mentioned above.
#include<stdio.h>
#include<string.h>
// A utility function to get max of two integers
int max (int x, int y) { return (x > y)? x : y; }
// Returns the length of the longest palindromic subsequence in seq
int lps(char *seq, int i, int j)
{
// Base Case 1: If there is only 1 character
if (i == j)
return 1;
// Base Case 2: If there are only 2 characters and both are same
if (seq[i] == seq[j] && i + 1 == j)
return 2;
// If the first and last characters match
if (seq[i] == seq[j])
return lps (seq, i+1, j-1) + 2;
// If the first and last characters do not match
return max( lps(seq, i, j-1), lps(seq, i+1, j) );
}
/* Driver program to test above functions */
int main()
{
char seq[] = "GEEKSFORGEEKS";
int n = strlen(seq);
printf ("The lnegth of the LPS is %d", lps(seq, 0, n-1));
getchar();
return 0;
}

Output:
The lnegth of the LPS is 5

Considering the above implementation, following is a partial recursion tree for a sequence of length 6 with all different characters.
/

L(0, 5)
\

/
\
L(1,5)
L(0,4)
/
\
/
\
/
\
/
\
L(2,5)
L(1,4) L(1,4) L(0,3)

In the above partial recursion tree, L(1, 4) is being solved twice. If we draw the complete recursion tree, then we can see that there are many
subproblems which are solved again and again. Since same suproblems are called again, this problem has Overlapping Subprolems property. So
LPS problem has both properties (see this and this) of a dynamic programming problem. Like other typical Dynamic Programming(DP) problems,
recomputations of same subproblems can be avoided by constructing a temporary array L[][] in bottom up manner.
Dynamic Programming Solution

C++
# A Dynamic Programming based Python program for LPS problem
# Returns the length of the longest palindromic subsequence in seq
#include<stdio.h>
#include<string.h>
// A utility function to get max of two integers
int max (int x, int y) { return (x > y)? x : y; }
// Returns the length of the longest palindromic subsequence in seq
int lps(char *str)
{
int n = strlen(str);
int i, j, cl;
int L[n][n]; // Create a table to store results of subproblems
// Strings of length 1 are palindrome of lentgh 1
for (i = 0; i < n; i++)
L[i][i] = 1;
// Build the table. Note that the lower diagonal values of table are
// useless and not filled in the process. The values are filled in a
// manner similar to Matrix Chain Multiplication DP solution (See
// http://www.geeksforgeeks.org/archives/15553). cl is length of
// substring
for (cl=2; cl<=n; cl++)
{
for (i=0; i<n-cl+1; i++)
{
j = i+cl-1;
if (str[i] == str[j] && cl == 2)
L[i][j] = 2;
else if (str[i] == str[j])
L[i][j] = L[i+1][j-1] + 2;
else
L[i][j] = max(L[i][j-1], L[i+1][j]);
}
}
return L[0][n-1];
}
/* Driver program to test above functions */
int main()
{
char seq[] = "GEEKS FOR GEEKS";
int n = strlen(seq);
printf ("The lnegth of the LPS is %d", lps(seq));
getchar();
return 0;
}

Python
# A Dynamic Programming based Python program for LPS problem
# Returns the length of the longest palindromic subsequence in seq
def lps(str):
n = len(str)
# Create a table to store results of subproblems
L = [[0 for x in range(n)] for x in range(n)]
# Strings of length 1 are palindrome of length 1

for i in range(n):
L[i][i] = 1
# Build the table. Note that the lower diagonal values of table are
# useless and not filled in the process. The values are filled in a
# manner similar to Matrix Chain Multiplication DP solution (See
# http://www.geeksforgeeks.org/dynamic-programming-set-8-matrix-chain-multiplication/
# cl is length of substring
for cl in range(2, n+1):
for i in range(n-cl+1):
j = i+cl-1
if str[i] == str[j] and cl == 2:
L[i][j] = 2
elif str[i] == str[j]:
L[i][j] = L[i+1][j-1] + 2
else:
L[i][j] = max(L[i][j-1], L[i+1][j]);
return L[0][n-1]
# Driver program to test above functions
seq = "GEEKS FOR GEEKS"
n = len(seq)
print("The length of the LPS is " + str(lps(seq)))
# This code is contributed by Bhavya Jain

The lnegth of the LPS is 7

Time Complexity of the above implementation is O(n^2) which is much better than the worst case time complexity of Naive Recursive
implementation.
This problem is close to the Longest Common Subsequence (LCS) problem. In fact, we can use LCS as a subroutine to solve this problem.
Following is the two step solution that uses LCS.
1) Reverse the given sequence and store the reverse in another array say rev[0..n-1]
2) LCS of the given sequence and rev[] will be the longest palindromic sequence.
This solution is also a O(n^2) solution.
References:
http://users.eecs.northwestern.edu/~dda902/336/hw6-sol.pdf

Dynamic Programming | Set 13 (Cutting a Rod)

Given a rod of length n inches and an array of prices that contains prices of all pieces of size smaller than n. Determine the maximum value
obtainable by cutting up the rod and selling the pieces. For example, if length of the rod is 8 and the values of different pieces are given as
following, then the maximum obtainable value is 22 (by cutting in two pieces of lengths 2 and 6)
length | 1 2 3 4 5 6 7 8
-------------------------------------------price
| 1 5 8 9 10 17 17 20

And if the prices are as following, then the maximum obtainable value is 24 (by cutting in eight pieces of length 1)
length | 1 2 3 4 5 6 7 8
-------------------------------------------price
| 3 5 8 9 10 17 17 20

The naive solution for this problem is to generate all configurations of different pieces and find the highest priced configuration. This solution is
exponential in term of time complexity. Let us see how this problem possesses both important properties of a Dynamic Programming (DP)
Problem and can efficiently solved using Dynamic Programming.
1) Optimal Substructure:
We can get the best price by making a cut at different positions and comparing the values obtained after a cut. We can recursively call the same
function for a piece obtained after a cut.
Let cutRoad(n) be the required (best possible price) value for a rod of lenght n. cutRod(n) can be written as following.
cutRod(n) = max(price[i] + cutRod(n-i-1)) for all i in {0, 1 .. n-1}
2) Overlapping Subproblems
Following is simple recursive implementation of the Rod Cutting problem. The implementation simply follows the recursive structure mentioned
above.
// A Naive recursive solution for Rod cutting problem
#include<stdio.h>
#include<limits.h>
// A utility function to get the maximum of two integers
int max(int a, int b) { return (a > b)? a : b;}
/* Returns the best obtainable price for a rod of length n and
price[] as prices of different pieces */
int cutRod(int price[], int n)
{
if (n <= 0)
return 0;
int max_val = INT_MIN;
// Recursively cut the rod in different pieces and compare different
// configurations
for (int i = 0; i<n; i++)
max_val = max(max_val, price[i] + cutRod(price, n-i-1));
return max_val;
}
/* Driver program to test above functions */
int main()
{
int arr[] = {1, 5, 8, 9, 10, 17, 17, 20};
int size = sizeof(arr)/sizeof(arr[0]);
printf("Maximum Obtainable Value is %d\n", cutRod(arr, size));
getchar();
return 0;
}

Output:
Maximum Obtainable Value is 22

Considering the above implementation, following is recursion tree for a Rod of length 4.
cR() ---> cutRod()
/
/

/
/

cR(4)
\
\

\
\

cR(3)
/ | \
/ | \
cR(2) cR(1) cR(0)
/ \
|
/ \
|
cR(1) cR(0) cR(0)

cR(2)
cR(1) cR(0)
/ \
|
/
\
|
cR(1) cR(0) cR(0)
|
|
cR(0)

In the above partial recursion tree, cR(2) is being solved twice. We can see that there are many subproblems which are solved again and again.
Since same suproblems are called again, this problem has Overlapping Subprolems property. So the Rod Cutting problem has both properties
(see this and this) of a dynamic programming problem. Like other typical Dynamic Programming(DP) problems, recomputations of same
subproblems can be avoided by constructing a temporary array val[] in bottom up manner.

C++
// A Dynamic Programming solution for Rod cutting problem
#include<stdio.h>
#include<limits.h>
// A utility function to get the maximum of two integers
int max(int a, int b) { return (a > b)? a : b;}
/* Returns the best obtainable price for a rod of length n and
price[] as prices of different pieces */
int cutRod(int price[], int n)
{
int val[n+1];
val[0] = 0;
int i, j;
// Build the table val[] in bottom up manner and return the last entry
// from the table
for (i = 1; i<=n; i++)
{
int max_val = INT_MIN;
for (j = 0; j < i; j++)
max_val = max(max_val, price[j] + val[i-j-1]);
val[i] = max_val;
}
return val[n];
}
/* Driver program to test above functions */
int main()
{
int arr[] = {1, 5, 8, 9, 10, 17, 17, 20};
int size = sizeof(arr)/sizeof(arr[0]);
printf("Maximum Obtainable Value is %d\n", cutRod(arr, size));
getchar();
return 0;
}

Python
# A Dynamic Programming solution for Rod cutting problem
INT_MIN = -32767
# Returns the best obtainable price for a rod of length n and
# price[] as prices of different pieces
def cutRod(price, n):
val = [0 for x in range(n+1)]
val[0] = 0
# Build the table val[] in bottom up manner and return
# the last entry from the table
for i in range(1, n+1):
max_val = INT_MIN
for j in range(i):
max_val = max(max_val, price[j] + val[i-j-1])
val[i] = max_val
return val[n]
# Driver program to test above functions
arr = [1, 5, 8, 9, 10, 17, 17, 20]
size = len(arr)
print("Maximum Obtainable Value is " + str(cutRod(arr, size)))

# This code is contributed by Bhavya Jain

Maximum Obtainable Value is 22

Time Complexity of the above implementation is O(n^2) which is much better than the worst case time complexity of Naive Recursive
implementation.

Dynamic Programming | Set 14 (Maximum Sum Increasing Subsequence)

Given an array of n positive integers. Write a program to find the sum of maximum sum subsequence of the given array such that the intgers in the
subsequence are sorted in increasing order. For example, if input is {1, 101, 2, 3, 100, 4, 5}, then output should be 106 (1 + 2 + 3 + 100), if the
input array is {3, 4, 5, 10}, then output should be 22 (3 + 4 + 5 + 10) and if the input array is {10, 5, 4, 3}, then output should be 10
Solution
This problem is a variation of standard Longest Increasing Subsequence (LIS) problem. We need a slight change in the Dynamic Programming
solution of LIS problem. All we need to change is to use sum as a criteria instead of length of increasing subsequence.
Following are C/C++ and Python implementations for Dynamic Programming solution of the problem.

C/C++
/* Dynamic Programming implementation of Maximum Sum Increasing
Subsequence (MSIS) problem */
#include<stdio.h>
/* maxSumIS() returns the maximum sum of increasing subsequence
in arr[] of size n */
int maxSumIS( int arr[], int n )
{
int i, j, max = 0;
int msis[n];
/* Initialize msis values for all indexes */
for ( i = 0; i < n; i++ )
msis[i] = arr[i];
/* Compute maximum sum values in bottom up manner */
for ( i = 1; i < n; i++ )
for ( j = 0; j < i; j++ )
if ( arr[i] > arr[j] && msis[i] < msis[j] + arr[i])
msis[i] = msis[j] + arr[i];
/* Pick maximum of all msis values */
for ( i = 0; i < n; i++ )
if ( max < msis[i] )
max = msis[i];
return max;
}
/* Driver program to test above function */
int main()
{
int arr[] = {1, 101, 2, 3, 100, 4, 5};
int n = sizeof(arr)/sizeof(arr[0]);
printf("Sum of maximum sum increasing subsequence is %d\n",
maxSumIS( arr, n ) );
return 0;
}

Python
# Dynamic Programming bsed Python implementation of Maximum Sum Increasing
# Subsequence (MSIS) problem
# maxSumIS() returns the maximum sum of increasing subsequence in arr[] of
# size n
def maxSumIS(arr, n):
max = 0
msis = [0 for x in range(n)]
# Initialize msis values for all indexes
for i in range(n):
msis[i] = arr[i]
# Compute maximum sum values in bottom up manner
for i in range(1, n):
for j in range(i):
if arr[i] > arr[j] and msis[i] < msis[j] + arr[i]:
msis[i] = msis[j] + arr[i]
# Pick maximum of all msis values
for i in range(n):
if max < msis[i]:

max = msis[i]
return max
# Driver program to test above function
arr = [1, 101, 2, 3, 100, 4, 5]
n = len(arr)
print("Sum of maximum sum increasing subsequence is " +
str(maxSumIS(arr, n)))
# This code is contributed by Bhavya Jain

Output:
Sum of maximum sum increasing subsequence is 106

Time Complexity: O(n^2)

Source: Maximum Sum Increasing Subsequence Problem

Dynamic Programming | Set 15 (Longest Bitonic Subsequence)

Given an array arr[0 n-1] containing n positive integers, a subsequence of arr[] is called Bitonic if it is first increasing, then decreasing. Write a
function that takes an array as argument and returns the length of the longest bitonic subsequence.
A sequence, sorted in increasing order is considered Bitonic with the decreasing part as empty. Similarly, decreasing order sequence is considered
Bitonic with the increasing part as empty.
Examples:
Input arr[] = {1, 11, 2, 10, 4, 5, 2, 1};
Output: 6 (A Longest Bitonic Subsequence of length 6 is 1, 2, 10, 4, 2, 1)
Input arr[] = {12, 11, 40, 5, 3, 1}
Output: 5 (A Longest Bitonic Subsequence of length 5 is 12, 11, 5, 3, 1)
Input arr[] = {80, 60, 30, 40, 20, 10}
Output: 5 (A Longest Bitonic Subsequence of length 5 is 80, 60, 30, 20, 10)

Source: Microsoft Interview Question

Solution
This problem is a variation of standard Longest Increasing Subsequence (LIS) problem. Let the input array be arr[] of length n. We need to
construct two arrays lis[] and lds[] using Dynamic Programming solution of LIS problem. lis[i] stores the length of the Longest Increasing
subsequence ending with arr[i]. lds[i] stores the length of the longest Decreasing subsequence starting from arr[i]. Finally, we need to return the
max value of lis[i] + lds[i] 1 where i is from 0 to n-1.
Following is C++ implementation of the above Dynamic Programming solution.

C++
/* Dynamic Programming implementation of longest bitonic subsequence problem */
#include<stdio.h>
#include<stdlib.h>
/* lbs() returns the length of the Longest Bitonic Subsequence in
arr[] of size n. The function mainly creates two temporary arrays
lis[] and lds[] and returns the maximum lis[i] + lds[i] - 1.
lis[i] ==> Longest Increasing subsequence ending with arr[i]
lds[i] ==> Longest decreasing subsequence starting with arr[i]
*/
int lbs( int arr[], int n )
{
int i, j;
/* Allocate memory for LIS[] and initialize LIS values as 1 for
all indexes */
int *lis = new int[n];
for (i = 0; i < n; i++)
lis[i] = 1;
/* Compute LIS values from left to right */
for (i = 1; i < n; i++)
for (j = 0; j < i; j++)
if (arr[i] > arr[j] && lis[i] < lis[j] + 1)
lis[i] = lis[j] + 1;
/* Allocate memory for lds and initialize LDS values for
all indexes */
int *lds = new int [n];
for (i = 0; i < n; i++)
lds[i] = 1;
/* Compute LDS values from right to left */
for (i = n-2; i >= 0; i--)
for (j = n-1; j > i; j--)
if (arr[i] > arr[j] && lds[i] < lds[j] + 1)
lds[i] = lds[j] + 1;
/* Return the maximum value
int max = lis[0] + lds[0] for (i = 1; i < n; i++)
if (lis[i] + lds[i] - 1 >
max = lis[i] + lds[i]
return max;

of lis[i] + lds[i] - 1*/

1;
max)
- 1;

}
/* Driver program to test above function */
int main()
{
int arr[] = {0, 8, 4, 12, 2, 10, 6, 14, 1, 9, 5,
13, 3, 11, 7, 15};
int n = sizeof(arr)/sizeof(arr[0]);
printf("Length of LBS is %d\n", lbs( arr, n ) );
return 0;
}

Java
/* Dynamic Programming implementation in Java for longest bitonic
subsequence problem */
import java.util.*;
import java.lang.*;
import java.io.*;
class LBS
{
/* lbs() returns the length of the Longest Bitonic Subsequence in
arr[] of size n. The function mainly creates two temporary arrays
lis[] and lds[] and returns the maximum lis[i] + lds[i] - 1.
lis[i] ==>
lds[i] ==>
*/
static int
{
int i,

Longest Increasing subsequence ending with arr[i]

Longest decreasing subsequence starting with arr[i]
lbs( int arr[], int n )
j;

/* Allocate memory for LIS[] and initialize LIS values as 1 for

all indexes */
int[] lis = new int[n];
for (i = 0; i < n; i++)
lis[i] = 1;
/* Compute LIS values from left to right */
for (i = 1; i < n; i++)
for (j = 0; j < i; j++)
if (arr[i] > arr[j] && lis[i] < lis[j] + 1)
lis[i] = lis[j] + 1;
/* Allocate memory for lds and initialize LDS values for
all indexes */
int[] lds = new int [n];
for (i = 0; i < n; i++)
lds[i] = 1;
/* Compute LDS values from right to left */
for (i = n-2; i >= 0; i--)
for (j = n-1; j > i; j--)
if (arr[i] > arr[j] && lds[i] < lds[j] + 1)
lds[i] = lds[j] + 1;
/* Return the maximum value of lis[i] + lds[i] - 1*/
int max = lis[0] + lds[0] - 1;
for (i = 1; i < n; i++)
if (lis[i] + lds[i] - 1 > max)
max = lis[i] + lds[i] - 1;
return max;
}
public static void main (String[] args)
{
int arr[] = {0, 8, 4, 12, 2, 10, 6, 14, 1, 9, 5,
13, 3, 11, 7, 15};
int n = arr.length;
System.out.println("Length of LBS is "+ lbs( arr, n ));
}
}

Length of LBS is 7

Time Complexity: O(n^2)

Auxiliary Space: O(n)

Dynamic Programming | Set 16 (Floyd Warshall Algorithm)

The Floyd Warshall Algorithm is for solving the All Pairs Shortest Path problem. The problem is to find shortest distances between every pair of
vertices in a given edge weighted directed Graph.
Example:
Input:
graph[][] = { {0, 5, INF, 10},
{INF, 0, 3, INF},
{INF, INF, 0, 1},
{INF, INF, INF, 0} }
which represents the following graph
10
(0)------->(3)
|
/|\
5 |
|
|
| 1
\|/
|
(1)------->(2)
3
Note that the value of graph[i][j] is 0 if i is equal to j
And graph[i][j] is INF (infinite) if there is no edge from vertex i to j.
Output:
Shortest distance matrix
0
5
8
INF
0
3
INF
INF
0
INF
INF
INF

9
4
1
0

Floyd Warshall Algorithm

We initialize the solution matrix same as the input graph matrix as a first step. Then we update the solution matrix by considering all vertices as an
intermediate vertex. The idea is to one by one pick all vertices and update all shortest paths which include the picked vertex as an intermediate
vertex in the shortest path. When we pick vertex number k as an intermediate vertex, we already have considered vertices {0, 1, 2, .. k-1} as
intermediate vertices. For every pair (i, j) of source and destination vertices respectively, there are two possible cases.
1) k is not an intermediate vertex in shortest path from i to j. We keep the value of dist[i][j] as it is.
2) k is an intermediate vertex in shortest path from i to j. We update the value of dist[i][j] as dist[i][k] + dist[k][j].
The following figure is taken from the Cormen book. It shows the above optimal substructure property in the all-pairs shortest path problem.

Following is implementations of the Floyd Warshall algorithm.

C/C++
// C Program for Floyd Warshall Algorithm
#include<stdio.h>
// Number of vertices in the graph
#define V 4
/* Define Infinite as a large enough value. This value will be used
for vertices not connected to each other */
#define INF 99999
// A function to print the solution matrix
void printSolution(int dist[][V]);
// Solves the all-pairs shortest path problem using Floyd Warshall algorithm
void floydWarshell (int graph[][V])
{
/* dist[][] will be the output matrix that will finally have the shortest
distances between every pair of vertices */
int dist[V][V], i, j, k;
/* Initialize the solution matrix same as input graph matrix. Or
we can say the initial values of shortest distances are based
on shortest paths considering no intermediate vertex. */

for (i = 0; i < V; i++)

for (j = 0; j < V; j++)
dist[i][j] = graph[i][j];
/* Add all vertices one by one to the set of intermediate vertices.
---> Before start of a iteration, we have shortest distances between all
pairs of vertices such that the shortest distances consider only the
vertices in set {0, 1, 2, .. k-1} as intermediate vertices.
----> After the end of a iteration, vertex no. k is added to the set of
intermediate vertices and the set becomes {0, 1, 2, .. k} */
for (k = 0; k < V; k++)
{
// Pick all vertices as source one by one
for (i = 0; i < V; i++)
{
// Pick all vertices as destination for the
// above picked source
for (j = 0; j < V; j++)
{
// If vertex k is on the shortest path from
// i to j, then update the value of dist[i][j]
if (dist[i][k] + dist[k][j] < dist[i][j])
dist[i][j] = dist[i][k] + dist[k][j];
}
}
}
// Print the shortest distance matrix
printSolution(dist);
}
/* A utility function to print solution */
void printSolution(int dist[][V])
{
printf ("Following matrix shows the shortest distances"
" between every pair of vertices \n");
for (int i = 0; i < V; i++)
{
for (int j = 0; j < V; j++)
{
if (dist[i][j] == INF)
printf("%7s", "INF");
else
printf ("%7d", dist[i][j]);
}
printf("\n");
}
}
// driver program to test above function
int main()
{
/* Let us create the following weighted graph
10
(0)------->(3)
|
/|\
5 |
|
|
| 1
\|/
|
(1)------->(2)
3
*/
int graph[V][V] = { {0, 5, INF, 10},
{INF, 0, 3, INF},
{INF, INF, 0, 1},
{INF, INF, INF, 0}
};
// Print the solution
floydWarshell(graph);
return 0;
}

Java
// A Java program for Floyd Warshall All Pairs Shortest
// Path algorithm.
import java.util.*;
import java.lang.*;
import java.io.*;

class AllPairShortestPath
{
final static int INF = 99999, V = 4;
void floydWarshall(int graph[][])
{
int dist[][] = new int[V][V];
int i, j, k;
/* Initialize the solution matrix same as input graph matrix.
Or we can say the initial values of shortest distances
are based on shortest paths considering no intermediate
vertex. */
for (i = 0; i < V; i++)
for (j = 0; j < V; j++)
dist[i][j] = graph[i][j];
/* Add all vertices one by one to the set of intermediate
vertices.
---> Before start of a iteration, we have shortest
distances between all pairs of vertices such that
the shortest distances consider only the vertices in
set {0, 1, 2, .. k-1} as intermediate vertices.
----> After the end of a iteration, vertex no. k is added
to the set of intermediate vertices and the set
becomes {0, 1, 2, .. k} */
for (k = 0; k < V; k++)
{
// Pick all vertices as source one by one
for (i = 0; i < V; i++)
{
// Pick all vertices as destination for the
// above picked source
for (j = 0; j < V; j++)
{
// If vertex k is on the shortest path from
// i to j, then update the value of dist[i][j]
if (dist[i][k] + dist[k][j] < dist[i][j])
dist[i][j] = dist[i][k] + dist[k][j];
}
}
}
// Print the shortest distance matrix
printSolution(dist);
}
void printSolution(int dist[][])
{
System.out.println("Following matrix shows the shortest "+
"distances between every pair of vertices");
for (int i=0; i<V; ++i)
{
for (int j=0; j<V; ++j)
{
if (dist[i][j]==INF)
System.out.print("INF ");
else
System.out.print(dist[i][j]+" ");
}
System.out.println();
}
}
// Driver program to test above function
public static void main (String[] args)
{
/* Let us create the following weighted graph
10
(0)------->(3)
|
/|\
5 |
|
|
| 1
\|/
|
(1)------->(2)
3
*/
int graph[][] = { {0, 5, INF, 10},
{INF, 0, 3, INF},
{INF, INF, 0, 1},
{INF, INF, INF, 0}

};
AllPairShortestPath a = new AllPairShortestPath();
// Print the solution
a.floydWarshall(graph);
}
}
// Contributed by Aakash Hasija

Output:
Following matrix shows the shortest distances between every pair of vertices
0
5
8
9
INF
0
3
4
INF
INF
0
1
INF
INF
INF
0

Time Complexity: O(V^3)

The above program only prints the shortest distances. We can modify the solution to print the shortest paths also by storing the predecessor
information in a separate 2D matrix.
Also, the value of INF can be taken as INT_MAX from limits.h to make sure that we handle maximum possible value. When we take INF as
INT_MAX, we need to change the if condition in the above program to avoid arithmatic overflow.
#include<limits.h>
#define INF INT_MAX
..........................
if (dist[i][k] != INF && dist[k][j] != INF && dist[i][k] + dist[k][j] < dist[i][j])
dist[i][j] = dist[i][k] + dist[k][j];
...........................

Dynamic Programming | Set 17 (Palindrome Partitioning)

Given a string, a partitioning of the string is a palindrome partitioning if every substring of the partition is a palindrome. For example,
aba|b|bbabb|a|b|aba is a palindrome partitioning of ababbbabbababa. Determine the fewest cuts needed for palindrome partitioning of a given
string. For example, minimum 3 cuts are needed for ababbbabbababa. The three cuts are a|babbbab|b|ababa. If a string is palindrome, then
minimum 0 cuts are needed. If a string of length n containing all different characters, then minimum n-1 cuts are needed.
Solution
This problem is a variation of Matrix Chain Multiplication problem. If the string is palindrome, then we simply return 0. Else, like the Matrix Chain
Multiplication problem, we try making cuts at all possible places, recursively calculate the cost for each cut and return the minimum value.
Let the given string be str and minPalPartion() be the function that returns the fewest cuts needed for palindrome partitioning. following is the
optimal substructure property.
// i is the starting index and j is the ending index. i must be passed as 0 and j as n-1
minPalPartion(str, i, j) = 0 if i == j. // When string is of length 1.
minPalPartion(str, i, j) = 0 if str[i..j] is palindrome.
// If none of the above conditions is true, then minPalPartion(str, i, j) can be
// calculated recursively using the following formula.
minPalPartion(str, i, j) = Min { minPalPartion(str, i, k) + 1 +
minPalPartion(str, k+1, j) }
where k varies from i to j-1

Following is Dynamic Programming solution. It stores the solutions to subproblems in two arrays P[][] and C[][], and reuses the calculated values.
// Dynamic Programming Solution for Palindrome Partitioning Problem
#include <stdio.h>
#include <string.h>
#include <limits.h>
// A utility function to get minimum of two integers
int min (int a, int b) { return (a < b)? a : b; }
// Returns the minimum number of cuts needed to partition a string
// such that every part is a palindrome
int minPalPartion(char *str)
{
// Get the length of the string
int n = strlen(str);
/* Create two arrays to build the solution in bottom up manner
C[i][j] = Minimum number of cuts needed for palindrome partitioning
of substring str[i..j]
P[i][j] = true if substring str[i..j] is palindrome, else false
Note that C[i][j] is 0 if P[i][j] is true */
int C[n][n];
bool P[n][n];
int i, j, k, L; // different looping variables
// Every substring of length 1 is a palindrome
for (i=0; i<n; i++)
{
P[i][i] = true;
C[i][i] = 0;
}
/* L is substring length. Build the solution in bottom up manner by
considering all substrings of length starting from 2 to n.
The loop structure is same as Matrx Chain Multiplication problem (
See http://www.geeksforgeeks.org/archives/15553 )*/
for (L=2; L<=n; L++)
{
// For substring of length L, set different possible starting indexes
for (i=0; i<n-L+1; i++)
{
j = i+L-1; // Set ending index
// If L is 2, then we just need to compare two characters. Else
// need to check two corner characters and value of P[i+1][j-1]
if (L == 2)
P[i][j] = (str[i] == str[j]);
else
P[i][j] = (str[i] == str[j]) && P[i+1][j-1];
// IF str[i..j] is palindrome, then C[i][j] is 0
if (P[i][j] == true)

C[i][j] = 0;
else
{
// Make a cut at every possible localtion starting from i to j,
// and get the minimum cost cut.
C[i][j] = INT_MAX;
for (k=i; k<=j-1; k++)
C[i][j] = min (C[i][j], C[i][k] + C[k+1][j]+1);
}
}
}
// Return the min cut value for complete string. i.e., str[0..n-1]
return C[0][n-1];
}
// Driver program to test above function
int main()
{
char str[] = "ababbbabbababa";
printf("Min cuts needed for Palindrome Partitioning is %d",
minPalPartion(str));
return 0;
}

Output:
Min cuts needed for Palindrome Partitioning is 3

Time Complexity: O(n3)

An optimization to above approach
In above approach, we can calculating minimum cut while finding all palindromic substring. If we finding all palindromic substring 1st and then we
calculate minimum cut, time complexity will reduce to O(n2).
Thanks for Vivek for suggesting this optimization.
// Dynamic Programming Solution for Palindrome Partitioning Problem
#include <stdio.h>
#include <string.h>
#include <limits.h>
// A utility function to get minimum of two integers
int min (int a, int b) { return (a < b)? a : b; }
// Returns the minimum number of cuts needed to partition a string
// such that every part is a palindrome
int minPalPartion(char *str)
{
// Get the length of the string
int n = strlen(str);
/* Create two arrays to build the solution in bottom up manner
C[i] = Minimum number of cuts needed for palindrome partitioning
of substring str[0..i]
P[i][j] = true if substring str[i..j] is palindrome, else false
Note that C[i] is 0 if P[0][i] is true */
int C[n];
bool P[n][n];
int i, j, k, L; // different looping variables
// Every substring of length 1 is a palindrome
for (i=0; i<n; i++)
{
P[i][i] = true;
}
/* L is substring length. Build the solution in bottom up manner by
considering all substrings of length starting from 2 to n. */
for (L=2; L<=n; L++)
{
// For substring of length L, set different possible starting indexes
for (i=0; i<n-L+1; i++)
{
j = i+L-1; // Set ending index
// If L is 2, then we just need to compare two characters. Else
// need to check two corner characters and value of P[i+1][j-1]
if (L == 2)

P[i][j] = (str[i] == str[j]);

else
P[i][j] = (str[i] == str[j]) && P[i+1][j-1];
}
}
for (i=0; i<n; i++)
{
if (P[0][i] == true)
C[i] = 0;
else
{
C[i] = INT_MAX;
for(j=0;j<i;j++)
{
if(P[j+1][i] == true && 1+C[j]<C[i])
C[i]=1+C[j];
}
}
}
// Return the min cut value for complete string. i.e., str[0..n-1]
return C[n-1];
}
// Driver program to test above function
int main()
{
char str[] = "ababbbabbababa";
printf("Min cuts needed for Palindrome Partitioning is %d",
minPalPartion(str));
return 0;
}

Output:
Min cuts needed for Palindrome Partitioning is 3

Time Complexity: O(n2)

Dynamic Programming | Set 18 (Partition problem)

Partition problem is to determine whether a given set can be partitioned into two subsets such that the sum of elements in both subsets is same.
Examples
arr[] = {1, 5, 11, 5}
Output: true
The array can be partitioned as {1, 5, 5} and {11}
arr[] = {1, 5, 3}
Output: false
The array cannot be partitioned into equal sum sets.

Following are the two main steps to solve this problem:

1) Calculate sum of the array. If sum is odd, there can not be two subsets with equal sum, so return false.
2) If sum of array elements is even, calculate sum/2 and find a subset of array with sum equal to sum/2.
The first step is simple. The second step is crucial, it can be solved either using recursion or Dynamic Programming.
Recursive Solution
Following is the recursive property of the second step mentioned above.
Let isSubsetSum(arr, n, sum/2) be the function that returns true if
there is a subset of arr[0..n-1] with sum equal to sum/2
The isSubsetSum problem can be divided into two subproblems
a) isSubsetSum() without considering last element
(reducing n to n-1)
b) isSubsetSum considering the last element
(reducing sum/2 by arr[n-1] and n to n-1)
If any of the above the above subproblems return true, then return true.
isSubsetSum (arr, n, sum/2) = isSubsetSum (arr, n-1, sum/2) ||
isSubsetSum (arr, n-1, sum/2 - arr[n-1])

C/C++
// A recursive C program for partition problem
#include <stdio.h>
// A utility function that returns true if there is
// a subset of arr[] with sun equal to given sum
bool isSubsetSum (int arr[], int n, int sum)
{
// Base Cases
if (sum == 0)
return true;
if (n == 0 && sum != 0)
return false;
// If last element is greater than sum, then
// ignore it
if (arr[n-1] > sum)
return isSubsetSum (arr, n-1, sum);
/* else, check if sum can be obtained by any of
the following
(a) including the last element
(b) excluding the last element
*/
return isSubsetSum (arr, n-1, sum) ||
isSubsetSum (arr, n-1, sum-arr[n-1]);
}
// Returns true if arr[] can be partitioned in two
// subsets of equal sum, otherwise false
bool findPartiion (int arr[], int n)
{
// Calculate sum of the elements in array
int sum = 0;
for (int i = 0; i < n; i++)
sum += arr[i];
// If sum is odd, there cannot be two subsets
// with equal sum
if (sum%2 != 0)
return false;

// Find if there is subset with sum equal to

// half of total sum
return isSubsetSum (arr, n, sum/2);
}
// Driver program to test above function
int main()
{
int arr[] = {3, 1, 5, 9, 12};
int n = sizeof(arr)/sizeof(arr[0]);
if (findPartiion(arr, n) == true)
printf("Can be divided into two subsets "
"of equal sum");
else
printf("Can not be divided into two subsets"
" of equal sum");
return 0;
}

Java
// A recursive Java solution for partition problem
import java.io.*;
class Partition
{
// A utility function that returns true if there is a
// subset of arr[] with sun equal to given sum
static boolean isSubsetSum (int arr[], int n, int sum)
{
// Base Cases
if (sum == 0)
return true;
if (n == 0 && sum != 0)
return false;
// If last element is greater than sum, then ignore it
if (arr[n-1] > sum)
return isSubsetSum (arr, n-1, sum);
/* else, check if sum can be obtained by any of
the following
(a) including the last element
(b) excluding the last element
*/
return isSubsetSum (arr, n-1, sum) ||
isSubsetSum (arr, n-1, sum-arr[n-1]);
}
// Returns true if arr[] can be partitioned in two
// subsets of equal sum, otherwise false
static boolean findPartition (int arr[], int n)
{
// Calculate sum of the elements in array
int sum = 0;
for (int i = 0; i < n; i++)
sum += arr[i];
// If sum is odd, there cannot be two subsets
// with equal sum
if (sum%2 != 0)
return false;
// Find if there is subset with sum equal to half
// of total sum
return isSubsetSum (arr, n, sum/2);
}
/*Driver function to check for above function*/
public static void main (String[] args)
{
int arr[] = {3, 1, 5, 9, 12};
int n = arr.length;
if (findPartition(arr, n) == true)
System.out.println("Can be divided into two "+
"subsets of equal sum");
else
System.out.println("Can not be divided into " +

"two subsets of equal sum");

}
}
/* This code is contributed by Devesh Agrawal */

Can be divided into two subsets of equal sum

Time Complexity: O(2^n) In worst case, this solution tries two possibilities (whether to include or exclude) for every element.

Dynamic Programming Solution

The problem can be solved using dynamic programming when the sum of the elements is not too big. We can create a 2D array part[][] of size
(sum/2)*(n+1). And we can construct the solution in bottom up manner such that every filled entry has following property
part[i][j] = true if a subset of {arr[0], arr[1], ..arr[j-1]} has sum
equal to i, otherwise false

C/C++
// A Dynamic Programming based C program to partition problem
#include <stdio.h>
// Returns true if arr[] can be partitioned in two subsets of
// equal sum, otherwise false
bool findPartiion (int arr[], int n)
{
int sum = 0;
int i, j;
// Caculcate sun of all elements
for (i = 0; i < n; i++)
sum += arr[i];
if (sum%2 != 0)
return false;
bool part[sum/2+1][n+1];
// initialize top row as true
for (i = 0; i <= n; i++)
part[0][i] = true;
// initialize leftmost column, except part[0][0], as 0
for (i = 1; i <= sum/2; i++)
part[i][0] = false;
// Fill the partition table in botton up manner
for (i = 1; i <= sum/2; i++)
{
for (j = 1; j <= n; j++)
{
part[i][j] = part[i][j-1];
if (i >= arr[j-1])
part[i][j] = part[i][j] || part[i - arr[j-1]][j-1];
}
}
/* // uncomment this part to print table
for (i = 0; i <= sum/2; i++)
{
for (j = 0; j <= n; j++)
printf ("%4d", part[i][j]);
printf("\n");
} */
return part[sum/2][n];
}
// Driver program to test above funtion
int main()
{
int arr[] = {3, 1, 1, 2, 2, 1};
int n = sizeof(arr)/sizeof(arr[0]);
if (findPartiion(arr, n) == true)
printf("Can be divided into two subsets of equal sum");
else

printf("Can not be divided into two subsets of equal sum");

getchar();
return 0;
}

Java
// A dynamic programming based Java program for partition problem
import java.io.*;
class Partition {
// Returns true if arr[] can be partitioned in two subsets of
// equal sum, otherwise false
static boolean findPartition (int arr[], int n)
{
int sum = 0;
int i, j;
// Caculcate sun of all elements
for (i = 0; i < n; i++)
sum += arr[i];
if (sum%2 != 0)
return false;
boolean part[][]=new boolean[sum/2+1][n+1];
// initialize top row as true
for (i = 0; i <= n; i++)
part[0][i] = true;
// initialize leftmost column, except part[0][0], as 0
for (i = 1; i <= sum/2; i++)
part[i][0] = false;
// Fill the partition table in botton up manner
for (i = 1; i <= sum/2; i++)
{
for (j = 1; j <= n; j++)
{
part[i][j] = part[i][j-1];
if (i >= arr[j-1])
part[i][j] = part[i][j] ||
part[i - arr[j-1]][j-1];
}
}
/* // uncomment this part to print table
for (i = 0; i <= sum/2; i++)
{
for (j = 0; j <= n; j++)
printf ("%4d", part[i][j]);
printf("\n");
} */
return part[sum/2][n];
}
/*Driver function to check for above function*/
public static void main (String[] args)
{
int arr[] = {3, 1, 1, 2, 2,1};
int n = arr.length;
if (findPartition(arr, n) == true)
System.out.println("Can be divided into two "
"subsets of equal sum");
else
System.out.println("Can not be divided into"
" two subsets of equal sum");
}
}
/* This code is contributed by Devesh Agrawal */

Can be divided into two subsets of equal sum

Following diagram shows the values in partition table. The diagram is taken form the wiki page of partition problem.

Time Complexity: O(sum*n)

Auxiliary Space: O(sum*n)
Please note that this solution will not be feasible for arrays with big sum.
References:
http://en.wikipedia.org/wiki/Partition_problem

Dynamic Programming | Set 19 (Word Wrap Problem)

Given a sequence of words, and a limit on the number of characters that can be put in one line (line width). Put line breaks in the given sequence
such that the lines are printed neatly. Assume that the length of each word is smaller than the line width.
The word processors like MS Word do task of placing line breaks. The idea is to have balanced lines. In other words, not have few lines with lots
of extra spaces and some lines with small amount of extra spaces.
The extra spaces includes spaces put at the end of every line except the last one.
The problem is to minimize the following total cost.
Cost of a line = (Number of extra spaces in the line)^3
Total Cost = Sum of costs for all lines
For example, consider the following string and line width M = 15
"Geeks for Geeks presents word wrap problem"
Following is the optimized arrangement of words in 3 lines
Geeks for Geeks
presents word
wrap problem
The total extra spaces in line 1, line 2 and line 3 are 0, 2 and 3 respectively.
So optimal value of total cost is 0 + 2*2 + 3*3 = 13

Please note that the total cost function is not sum of extra spaces, but sum of cubes (or square is also used) of extra spaces. The idea behind this
cost function is to balance the spaces among lines. For example, consider the following two arrangement of same set of words:
1) There are 3 lines. One line has 3 extra spaces and all other lines have 0 extra spaces. Total extra spaces = 3 + 0 + 0 = 3. Total cost = 3*3*3 +
0*0*0 + 0*0*0 = 27.
2) There are 3 lines. Each of the 3 lines has one extra space. Total extra spaces = 1 + 1 + 1 = 3. Total cost = 1*1*1 + 1*1*1 + 1*1*1 = 3.
Total extra spaces are 3 in both scenarios, but second arrangement should be preferred because extra spaces are balanced in all three lines. The
cost function with cubic sum serves the purpose because the value of total cost in second scenario is less.
Method 1 (Greedy Solution)
The greedy solution is to place as many words as possible in the first line. Then do the same thing for the second line and so on until all words are
placed. This solution gives optimal solution for many cases, but doesnt give optimal solution in all cases. For example, consider the following string
aaa bb cc ddddd and line width as 6. Greedy method will produce following output.
aaa bb
cc
ddddd

Extra spaces in the above 3 lines are 0, 4 and 1 respectively. So total cost is 0 + 64 + 1 = 65.
But the above solution is not the best solution. Following arrangement has more balanced spaces. Therefore less value of total cost function.
aaa
bb cc
ddddd

Extra spaces in the above 3 lines are 3, 1 and 1 respectively. So total cost is 27 + 1 + 1 = 29.
Despite being sub-optimal in some cases, the greedy approach is used by many word processors like MS Word and OpenOffice.org Writer.
Method 2 (Dynamic Programming)
The following Dynamic approach strictly follows the algorithm given in solution of Cormen book. First we compute costs of all possible lines in a
2D table lc[][]. The value lc[i][j] indicates the cost to put words from i to j in a single line where i and j are indexes of words in the input
sequences. If a sequence of words from i to j cannot fit in a single line, then lc[i][j] is considered infinite (to avoid it from being a part of the
solution). Once we have the lc[][] table constructed, we can calculate total cost using following recursive formula. In the following formula, C[j] is
the optimized total cost for arranging words from 1 to j.

The above recursion has overlapping subproblem property. For example, the solution of subproblem c(2) is used by c(3), C(4) and so on. So
Dynamic Programming is used to store the results of subproblems. The array c[] can be computed from left to right, since each value depends only
on earlier values.
To print the output, we keep track of what words go on what lines, we can keep a parallel p array that points to where each c value came from.

The last line starts at word p[n] and goes through word n. The previous line starts at word p[p[n]] and goes through word p[n] 1, etc. The function
printSolution() uses p[] to print the solution.
In the below program, input is an array l[] that represents lengths of words in a sequence. The value l[i] indicates length of the ith word (i starts
from 1) in theinput sequence.
// A Dynamic programming solution for Word Wrap Problem
#include <limits.h>
#include <stdio.h>
#define INF INT_MAX
// A utility function to print the solution
int printSolution (int p[], int n);
// l[] represents lengths of different words in input sequence. For example,
// l[] = {3, 2, 2, 5} is for a sentence like "aaa bb cc ddddd". n is size of
// l[] and M is line width (maximum no. of characters that can fit in a line)
void solveWordWrap (int l[], int n, int M)
{
// For simplicity, 1 extra space is used in all below arrays
// extras[i][j] will have number of extra spaces if words from i
// to j are put in a single line
int extras[n+1][n+1];
// lc[i][j] will have cost of a line which has words from
// i to j
int lc[n+1][n+1];
// c[i] will have total cost of optimal arrangement of words
// from 1 to i
int c[n+1];
// p[] is used to print the solution.
int p[n+1];
int i, j;
// calculate extra spaces in a single line. The value extra[i][j]
// indicates extra spaces if words from word number i to j are
// placed in a single line
for (i = 1; i <= n; i++)
{
extras[i][i] = M - l[i-1];
for (j = i+1; j <= n; j++)
extras[i][j] = extras[i][j-1] - l[j-1] - 1;
}
// Calculate line cost corresponding to the above calculated extra
// spaces. The value lc[i][j] indicates cost of putting words from
// word number i to j in a single line
for (i = 1; i <= n; i++)
{
for (j = i; j <= n; j++)
{
if (extras[i][j] < 0)
lc[i][j] = INF;
else if (j == n && extras[i][j] >= 0)
lc[i][j] = 0;
else
lc[i][j] = extras[i][j]*extras[i][j];
}
}
// Calculate minimum cost and find minimum cost arrangement.
// The value c[j] indicates optimized cost to arrange words
// from word number 1 to j.
c[0] = 0;
for (j = 1; j <= n; j++)
{
c[j] = INF;
for (i = 1; i <= j; i++)
{
if (c[i-1] != INF && lc[i][j] != INF && (c[i-1] + lc[i][j] < c[j]))
{
c[j] = c[i-1] + lc[i][j];
p[j] = i;
}
}
}

printSolution(p, n);
}
int printSolution (int p[], int n)
{
int k;
if (p[n] == 1)
k = 1;
else
k = printSolution (p, p[n]-1) + 1;
printf ("Line number %d: From word no. %d to %d \n", k, p[n], n);
return k;
}
// Driver program to test above functions
int main()
{
int l[] = {3, 2, 2, 5};
int n = sizeof(l)/sizeof(l[0]);
int M = 6;
solveWordWrap (l, n, M);
return 0;
}

Output:
Line number 1: From word no. 1 to 1
Line number 2: From word no. 2 to 3
Line number 3: From word no. 4 to 4

Time Complexity: O(n^2)

Auxiliary Space: O(n^2) The auxiliary space used in the above program cane be optimized to O(n) (See the reference 2 for details)
References:
http://en.wikipedia.org/wiki/Word_wrap

Dynamic Programming | Set 20 (Maximum Length Chain of Pairs)

You are given n pairs of numbers. In every pair, the first number is always smaller than the second number. A pair (c, d) can follow another pair (a,
b) if b < c. Chain of pairs can be formed in this fashion. Find the longest chain which can be formed from a given set of pairs. Source: Amazon
Interview | Set 2
For example, if the given pairs are {{5, 24}, {39, 60}, {15, 28}, {27, 40}, {50, 90} }, then the longest chain that can be formed is of length 3,
and the chain is {{5, 24}, {27, 40}, {50, 90}}
This problem is a variation of standard Longest Increasing Subsequence problem. Following is a simple two step process.
1) Sort given pairs in increasing order of first (or smaller) element.
2) Now run a modified LIS process where we compare the second element of already finalized LIS with the first element of new LIS being
constructed.
The following code is a slight modification of method 2 of this post.
#include<stdio.h>
#include<stdlib.h>
// Structure for a pair
struct pair
{
int a;
int b;
};
// This function assumes that arr[] is sorted in increasing order
// according the first (or smaller) values in pairs.
int maxChainLength( struct pair arr[], int n)
{
int i, j, max = 0;
int *mcl = (int*) malloc ( sizeof( int ) * n );
/* Initialize MCL (max chain length) values for all indexes */
for ( i = 0; i < n; i++ )
mcl[i] = 1;
/* Compute optimized chain length values in bottom up manner */
for ( i = 1; i < n; i++ )
for ( j = 0; j < i; j++ )
if ( arr[i].a > arr[j].b && mcl[i] < mcl[j] + 1)
mcl[i] = mcl[j] + 1;
// mcl[i] now stores the maximum chain length ending with pair i
/* Pick maximum of all MCL values */
for ( i = 0; i < n; i++ )
if ( max < mcl[i] )
max = mcl[i];
/* Free memory to avoid memory leak */
free( mcl );
return max;
}
/* Driver program to test above function */
int main()
{
struct pair arr[] = { {5, 24}, {15, 25},
{27, 40}, {50, 60} };
int n = sizeof(arr)/sizeof(arr[0]);
printf("Length of maximum size chain is %d\n",
maxChainLength( arr, n ));
return 0;
}

Output:
Length of maximum size chain is 3

Time Complexity: O(n^2) where n is the number of pairs.

The given problem is also a variation of Activity Selection problem and can be solved in (nLogn) time. To solve it as a activity selection problem,
consider the first element of a pair as start time in activity selection problem, and the second element of pair as end time. Thanks to Palash for
suggesting this approach.

Dynamic Programming | Set 22 (Box Stacking Problem)

You are given a set of n types of rectangular 3-D boxes, where the i^th box has height h(i), width w(i) and depth d(i) (all real numbers). You want
to create a stack of boxes which is as tall as possible, but you can only stack a box on top of another box if the dimensions of the 2-D base of the
lower box are each strictly larger than those of the 2-D base of the higher box. Of course, you can rotate a box so that any side functions as its
base. It is also allowable to use multiple instances of the same type of box.
Source: http://people.csail.mit.edu/bdean/6.046/dp/. The link also has video for explanation of solution.

The Box Stacking problem is a variation of LIS problem. We need to build a maximum height stack.
Following are the key points to note in the problem statement:
1) A box can be placed on top of another box only if both width and depth of the upper placed box are smaller than width and depth of the lower
box respectively.
2) We can rotate boxes. For example, if there is a box with dimensions {1x2x3} where 1 is height, 23 is base, then there can be three possibilities,
{1x2x3}, {2x1x3} and {3x1x2}.
3) We can use multiple instances of boxes. What it means is, we can have two different rotations of a box as part of our maximum height stack.
Following is the solution based on DP solution of LIS problem.
1) Generate all 3 rotations of all boxes. The size of rotation array becomes 3 times the size of original array. For simplicity, we consider depth as
always smaller than or equal to width.
2) Sort the above generated 3n boxes in decreasing order of base area.
3) After sorting the boxes, the problem is same as LIS with following optimal substructure property.
MSH(i) = Maximum possible Stack Height with box i at top of stack
MSH(i) = { Max ( MSH(j) ) + height(i) } where j < i and width(j) > width(i) and depth(j) > depth(i).
If there is no such j then MSH(i) = height(i)
4) To get overall maximum height, we return max(MSH(i)) where 0 < i < n Following is C++ implementation of the above solution.
/* Dynamic Programming implementation of Box Stacking problem */
#include<stdio.h>
#include<stdlib.h>
/* Representation of a box */
struct Box
{
// h > height, w > width, d > depth
int h, w, d; // for simplicity of solution, always keep w <= d
};
// A utility function to get minimum of two intgers
int min (int x, int y)
{ return (x < y)? x : y; }
// A utility function to get maximum of two intgers
int max (int x, int y)
{ return (x > y)? x : y; }
/* Following function is needed for library function qsort(). We
use qsort() to sort boxes in decreasing order of base area.
Refer following link for help of qsort() and compare()
http://www.cplusplus.com/reference/clibrary/cstdlib/qsort/ */
int compare (const void *a, const void * b)
{
return ( (*(Box *)b).d * (*(Box *)b).w )
( (*(Box *)a).d * (*(Box *)a).w );
}
/* Returns the height of the tallest stack that can be formed with give type of boxes */
int maxStackHeight( Box arr[], int n )

{
/* Create an array of all rotations of given boxes
For example, for a box {1, 2, 3}, we consider three
instances{{1, 2, 3}, {2, 1, 3}, {3, 1, 2}} */
Box rot[3*n];
int index = 0;
for (int i = 0; i < n; i++)
{
// Copy the original box
rot[index] = arr[i];
index++;
// First rotation of box
rot[index].h = arr[i].w;
rot[index].d = max(arr[i].h, arr[i].d);
rot[index].w = min(arr[i].h, arr[i].d);
index++;
// Second rotation of box
rot[index].h = arr[i].d;
rot[index].d = max(arr[i].h, arr[i].w);
rot[index].w = min(arr[i].h, arr[i].w);
index++;
}
// Now the number of boxes is 3n
n = 3*n;
/* Sort the array rot[] in decreasing order, using library
function for quick sort */
qsort (rot, n, sizeof(rot[0]), compare);
// Uncomment following two lines to print all rotations
// for (int i = 0; i < n; i++ )
//
printf("%d x %d x %d\n", rot[i].h, rot[i].w, rot[i].d);
/* Initialize msh values for all indexes
msh[i] > Maximum possible Stack Height with box i on top */
int msh[n];
for (int i = 0; i < n; i++ )
msh[i] = rot[i].h;
/* Compute optimized msh values in bottom up manner */
for (int i = 1; i < n; i++ )
for (int j = 0; j < i; j++ )
if ( rot[i].w < rot[j].w &&
rot[i].d < rot[j].d &&
msh[i] < msh[j] + rot[i].h
)
{
msh[i] = msh[j] + rot[i].h;
}
/* Pick maximum of all msh values */
int max = -1;
for ( int i = 0; i < n; i++ )
if ( max < msh[i] )
max = msh[i];
return max;
}
/* Driver program to test above function */
int main()
{
Box arr[] = { {4, 6, 7}, {1, 2, 3}, {4, 5, 6}, {10, 12, 32} };
int n = sizeof(arr)/sizeof(arr[0]);
printf("The maximum possible height of stack is %d\n",
maxStackHeight (arr, n) );
return 0;
}

Output:
The maximum possible height of stack is 60

In the above program, given input boxes are {4, 6, 7}, {1, 2, 3}, {4, 5, 6}, {10, 12, 32}. Following are all rotations of the boxes in decreasing
order of base area.

10 x 12
12 x 10
32 x 10
4 x 6 x
4 x 5 x
6 x 4 x
5 x 4 x
7 x 4 x
6 x 4 x
1 x 2 x
2 x 1 x
3 x 1 x

x 32
x 32
x 12
7
6
7
6
6
5
3
3
2

The height 60 is obtained by boxes { {3, 1, 2}, {1, 2, 3}, {6, 4, 5}, {4, 5, 6}, {4, 6, 7}, {32, 10, 12}, {10, 12, 32}}
Time Complexity: O(n^2)
Auxiliary Space: O(n)

Program for Fibonacci numbers

The Fibonacci numbers are the numbers in the following integer sequence.
0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89, 141, ..
In mathematical terms, the sequence Fn of Fibonacci numbers is defined by the recurrence relation
Fn = Fn-1 + Fn-2

with seed values

F0 = 0 and F1 = 1.

Write a function int fib(int n) that returns Fn. For example, if n = 0, then fib() should return 0. If n = 1, then it should return 1. For n > 1, it should
return Fn-1 + Fn-2
Following are different methods to get the nth Fibonacci number.
Method 1 ( Use recursion )
A simple method that is a direct recusrive implementation mathematical recurance relation given above.
#include<stdio.h>
int fib(int n)
{
if (n <= 1)
return n;
return fib(n-1) + fib(n-2);
}
int main ()
{
int n = 9;
printf("%d", fib(n));
getchar();
return 0;
}

Time Complexity: T(n) = T(n-1) + T(n-2) which is exponential.

We can observe that this implementation does a lot of repeated work (see the following recursion tree). So this is a bad implementation for nth
Fibonacci number.
fib(5)
/
\
fib(4)
fib(3)
/
\
/
\
fib(3)
fib(2)
fib(2)
fib(1)
/
\
/
\
/
\
fib(2) fib(1) fib(1) fib(0) fib(1) fib(0)
/
\
fib(1) fib(0)

Extra Space: O(n) if we consider the function call stack size, otherwise O(1).
Method 2 ( Use Dynamic Programming )
We can avoid the repeated work done is the method 1 by storing the Fibonacci numbers calculated so far.
#include<stdio.h>
int fib(int n)
{
/* Declare an array to store Fibonacci numbers. */
int f[n+1];
int i;
/* 0th and 1st number of the series are 0 and 1*/
f[0] = 0;
f[1] = 1;
for (i = 2; i <= n; i++)
{
/* Add the previous 2 numbers in the series
and store it */
f[i] = f[i-1] + f[i-2];
}

return f[n];
}
int main ()
{
int n = 9;
printf("%d", fib(n));
getchar();
return 0;
}

Time Complexity: O(n)

Extra Space: O(n)
Method 3 ( Space Otimized Method 2 )
We can optimize the space used in method 2 by storing the previous two numbers only because that is all we need to get the next Fibannaci
number in series.
#include<stdio.h>
int fib(int n)
{
int a = 0, b = 1, c, i;
if( n == 0)
return a;
for (i = 2; i <= n; i++)
{
c = a + b;
a = b;
b = c;
}
return b;
}
int main ()
{
int n = 9;
printf("%d", fib(n));
getchar();
return 0;
}

Time Complexity: O(n)

Extra Space: O(1)
Method 4 ( Using power of the matrix {{1,1},{1,0}} )
This another O(n) which relies on the fact that if we n times multiply the matrix M = {{1,1},{1,0}} to itself (in other words calculate power(M, n
)), then we get the (n+1)th Fibonacci number as the element at row and column (0, 0) in the resultant matrix.
The matrix representation gives the following closed expression for the Fibonacci numbers:

#include <stdio.h>
/* Helper function that multiplies 2 matricies F and M of size 2*2, and
puts the multiplication result back to F[][] */
void multiply(int F[2][2], int M[2][2]);
/* Helper function that calculates F[][] raise to the power n and puts the
result in F[][]
Note that this function is desinged only for fib() and won't work as general
power function */
void power(int F[2][2], int n);
int fib(int n)
{
int F[2][2] = {{1,1},{1,0}};
if (n == 0)
return 0;
power(F, n-1);
return F[0][0];
}
void multiply(int F[2][2], int M[2][2])
{
int x = F[0][0]*M[0][0] + F[0][1]*M[1][0];
int y = F[0][0]*M[0][1] + F[0][1]*M[1][1];

int z = F[1][0]*M[0][0] + F[1][1]*M[1][0];

int w = F[1][0]*M[0][1] + F[1][1]*M[1][1];
F[0][0]
F[0][1]
F[1][0]
F[1][1]

=
=
=
=

x;
y;
z;
w;

}
void power(int F[2][2], int n)
{
int i;
int M[2][2] = {{1,1},{1,0}};
// n - 1 times multiply the matrix to {{1,0},{0,1}}
for (i = 2; i <= n; i++)
multiply(F, M);
}
/* Driver program to test above function */
int main()
{
int n = 9;
printf("%d", fib(n));
getchar();
return 0;
}

Time Complexity: O(n)

Extra Space: O(1)
Method 5 ( Optimized Method 4 )
The method 4 can be optimized to work in O(Logn) time complexity. We can do recursive multiplication to get power(M, n) in the prevous
method (Similar to the optimization done in this post)
#include <stdio.h>
void multiply(int F[2][2], int M[2][2]);
void power(int F[2][2], int n);
/* function that returns nth Fibonacci number */
int fib(int n)
{
int F[2][2] = {{1,1},{1,0}};
if (n == 0)
return 0;
power(F, n-1);
return F[0][0];
}
/* Optimized version of power() in method 4 */
void power(int F[2][2], int n)
{
if( n == 0 || n == 1)
return;
int M[2][2] = {{1,1},{1,0}};
power(F, n/2);
multiply(F, F);
if (n%2 != 0)
multiply(F, M);
}
void multiply(int F[2][2],
{
int x = F[0][0]*M[0][0]
int y = F[0][0]*M[0][1]
int z = F[1][0]*M[0][0]
int w = F[1][0]*M[0][1]
F[0][0]
F[0][1]
F[1][0]
F[1][1]

=
=
=
=

int M[2][2])
+
+
+
+

F[0][1]*M[1][0];
F[0][1]*M[1][1];
F[1][1]*M[1][0];
F[1][1]*M[1][1];

x;
y;
z;
w;

}
/* Driver program to test above function */

int main()
{
int n = 9;
printf("%d", fib(9));
getchar();
return 0;
}

Time Complexity: O(Logn)

Extra Space: O(Logn) if we consider the function call stack size, otherwise O(1).
References:
http://en.wikipedia.org/wiki/Fibonacci_number
http://www.ics.uci.edu/~eppstein/161/960109.html

Minimum number of jumps to reach end

Given an array of integers where each element represents the max number of steps that can be made forward from that element. Write a function
to return the minimum number of jumps to reach the end of the array (starting from the first element). If an element is 0, then cannot move through
that element.
Example:
Input: arr[] = {1, 3, 5, 8, 9, 2, 6, 7, 6, 8, 9}
Output: 3 (1-> 3 -> 8 ->9)

First element is 1, so can only go to 3. Second element is 3, so can make at most 3 steps eg to 5 or 8 or 9.
Method 1 (Naive Recursive Approach)
A naive approach is to start from the first element and recursively call for all the elements reachable from first element. The minimum number of
jumps to reach end from first can be calculated using minimum number of jumps needed to reach end from the elements reachable from first.
minJumps(start, end) = Min ( minJumps(k, end) ) for all k reachable from start
#include <stdio.h>
#include <limits.h>
// Returns minimum number of jumps to reach arr[h] from arr[l]
int minJumps(int arr[], int l, int h)
{
// Base case: when source and destination are same
if (h == l)
return 0;
// When nothing is reachable from the given source
if (arr[l] == 0)
return INT_MAX;
// Traverse through all the points reachable from arr[l]. Recursively
// get the minimum number of jumps needed to reach arr[h] from these
// reachable points.
int min = INT_MAX;
for (int i = l+1; i <= h && i <= l + arr[l]; i++)
{
int jumps = minJumps(arr, i, h);
if(jumps != INT_MAX && jumps + 1 < min)
min = jumps + 1;
}
return min;
}
// Driver program to test above function
int main()
{
int arr[] = {1, 3, 6, 3, 2, 3, 6, 8, 9, 5};
int n = sizeof(arr)/sizeof(arr[0]);
printf("Minimum number of jumps to reach end is %d ", minJumps(arr, 0, n-1));
return 0;
}

If we trace the execution of this method, we can see that there will be overlapping subproblems. For example, minJumps(3, 9) will be called two
times as arr[3] is reachable from arr[1] and arr[2]. So this problem has both properties (optimal substructure and overlapping subproblems) of
Dynamic Programming.

Method 2 (Dynamic Programming)

In this method, we build a jumps[] array from left to right such that jumps[i] indicates the minimum number of jumps needed to reach arr[i] from
arr[0]. Finally, we return jumps[n-1].
#include <stdio.h>
#include <limits.h>
int min(int x, int y) { return (x < y)? x: y; }
// Returns minimum number of jumps to reach arr[n-1] from arr[0]
int minJumps(int arr[], int n)
{
int *jumps = new int[n]; // jumps[n-1] will hold the result
int i, j;

if (n == 0 || arr[0] == 0)
return INT_MAX;
jumps[0] = 0;
// Find the minimum number of jumps to reach arr[i]
// from arr[0], and assign this value to jumps[i]
for (i = 1; i < n; i++)
{
jumps[i] = INT_MAX;
for (j = 0; j < i; j++)
{
if (i <= j + arr[j] && jumps[j] != INT_MAX)
{
jumps[i] = min(jumps[i], jumps[j] + 1);
break;
}
}
}
return jumps[n-1];
}
// Driver program to test above function
int main()
{
int arr[] = {1, 3, 6, 1, 0, 9};
int size = sizeof(arr)/sizeof(int);
printf("Minimum number of jumps to reach end is %d ", minJumps(arr,size));
return 0;
}

Output:
Minimum number of jumps to reach end is 3

Thanks to paras for suggesting this method.

Time Complexity: O(n^2)

Method 3 (Dynamic Programming)

In this method, we build jumps[] array from right to left such that jumps[i] indicates the minimum number of jumps needed to reach arr[n-1] from
arr[i]. Finally, we return arr[0].
int minJumps(int arr[], int n)
{
int *jumps = new int[n]; // jumps[0] will hold the result
int min;
// Minimum number of jumps needed to reach last element
// from last elements itself is always 0
jumps[n-1] = 0;
int i, j;
// Start from the second element, move from right to left
// and construct the jumps[] array where jumps[i] represents
// minimum number of jumps needed to reach arr[m-1] from arr[i]
for (i = n-2; i >=0; i--)
{
// If arr[i] is 0 then arr[n-1] can't be reached from here
if (arr[i] == 0)
jumps[i] = INT_MAX;
// If we can direcly reach to the end point from here then
// jumps[i] is 1
else if (arr[i] >= n - i - 1)
jumps[i] = 1;
// Otherwise, to find out the minimum number of jumps needed
// to reach arr[n-1], check all the points reachable from here
// and jumps[] value for those points
else
{
min = INT_MAX; // initialize min value
// following loop checks with all reachable points and

// takes the minimum

for (j = i+1; j < n && j <= arr[i] + i; j++)
{
if (min > jumps[j])
min = jumps[j];
}
// Handle overflow
if (min != INT_MAX)
jumps[i] = min + 1;
else
jumps[i] = min; // or INT_MAX
}
}
return jumps[0];
}

Time Complexity: O(n^2) in worst case.

Thanks to Ashish for suggesting this solution.

Maximum size square sub-matrix with all 1s

Given a binary matrix, find out the maximum size square sub-matrix with all 1s.
For example, consider the below binary matrix.
0
1
0
1
1
0

1
1
1
1
1
0

1
0
1
1
1
0

0
1
1
1
1
0

1
0
0
0
1
0

The maximum square sub-matrix with all set bits is

1 1 1
1 1 1
1 1 1

Algorithm:
Let the given binary matrix be M[R][C]. The idea of the algorithm is to construct an auxiliary size matrix S[][] in which each entry S[i][j] represents
size of the square sub-matrix with all 1s including M[i][j] where M[i][j] is the rightmost and bottommost entry in sub-matrix.
1) Construct a sum matrix S[R][C] for the given M[R][C].
a) Copy first row and first columns as it is from M[][] to S[][]
b) For other entries, use following expressions to construct S[][]
If M[i][j] is 1 then
S[i][j] = min(S[i][j-1], S[i-1][j], S[i-1][j-1]) + 1
Else /*If M[i][j] is 0*/
S[i][j] = 0
2) Find the maximum entry in S[R][C]
3) Using the value and coordinates of maximum entry in S[i], print
sub-matrix of M[][]

For the given M[R][C] in above example, constructed S[R][C] would be:
0
1
0
1
1
0

1
1
1
1
2
0

1
0
1
2
2
0

0
1
1
2
3
0

1
0
0
0
1
0

The value of maximum entry in above matrix is 3 and coordinates of the entry are (4, 3). Using the maximum value and its coordinates, we can find
out the required sub-matrix.
#include<stdio.h>
#define bool int
#define R 6
#define C 5
void printMaxSubSquare(bool M[R][C])
{
int i,j;
int S[R][C];
int max_of_s, max_i, max_j;
/* Set first column of S[][]*/
for(i = 0; i < R; i++)
S[i][0] = M[i][0];
/* Set first row of S[][]*/
for(j = 0; j < C; j++)
S[0][j] = M[0][j];
/* Construct other entries of S[][]*/
for(i = 1; i < R; i++)
{
for(j = 1; j < C; j++)
{
if(M[i][j] == 1)
S[i][j] = min(S[i][j-1], S[i-1][j], S[i-1][j-1]) + 1;
else
S[i][j] = 0;
}
}

/* Find the maximum entry, and indexes of maximum entry

in S[][] */
max_of_s = S[0][0]; max_i = 0; max_j = 0;
for(i = 0; i < R; i++)
{
for(j = 0; j < C; j++)
{
if(max_of_s < S[i][j])
{
max_of_s = S[i][j];
max_i = i;
max_j = j;
}
}
}
printf("\n Maximum size sub-matrix is: \n");
for(i = max_i; i > max_i - max_of_s; i--)
{
for(j = max_j; j > max_j - max_of_s; j--)
{
printf("%d ", M[i][j]);
}
printf("\n");
}
}
/* UTILITY FUNCTIONS */
/* Function to get minimum of three values */
int min(int a, int b, int c)
{
int m = a;
if (m > b)
m = b;
if (m > c)
m = c;
return m;
}
/* Driver function to test above functions */
int main()
{
bool M[R][C] = {{0, 1, 1, 0, 1},
{1, 1, 0, 1, 0},
{0, 1, 1, 1, 0},
{1, 1, 1, 1, 0},
{1, 1, 1, 1, 1},
{0, 0, 0, 0, 0}};
printMaxSubSquare(M);
getchar();
}

Time Complexity: O(m*n) where m is number of rows and n is number of columns in the given matrix.
Auxiliary Space: O(m*n) where m is number of rows and n is number of columns in the given matrix.
Algorithmic Paradigm: Dynamic Programming

Ugly Numbers
Ugly numbers are numbers whose only prime factors are 2, 3 or 5. The sequence
1, 2, 3, 4, 5, 6, 8, 9, 10, 12, 15,
shows the first 11 ugly numbers. By convention, 1 is included.
Write a program to find and print the 150th ugly number.
METHOD 1 (Simple)
Thanks to Nedylko Draganov for suggesting this solution.
Algorithm:
Loop for all positive integers until ugly number count is smaller than n, if an integer is ugly than increment ugly number count.
To check if a number is ugly, divide the number by greatest divisible powers of 2, 3 and 5, if the number becomes 1 then it is an ugly number
otherwise not.
For example, let us see how to check for 300 is ugly or not. Greatest divisible power of 2 is 4, after dividing 300 by 4 we get 75. Greatest
divisible power of 3 is 3, after dividing 75 by 3 we get 25. Greatest divisible power of 5 is 25, after dividing 25 by 25 we get 1. Since we get 1
finally, 300 is ugly number.
Implementation:
# include<stdio.h>
# include<stdlib.h>
/*This function divides a by greatest divisible
power of b*/
int maxDivide(int a, int b)
{
while (a%b == 0)
a = a/b;
return a;
}
/* Function to check
int isUgly(int no)
{
no = maxDivide(no,
no = maxDivide(no,
no = maxDivide(no,

if a number is ugly or not */

2);
3);
5);

return (no == 1)? 1 : 0;

}
/* Function to get the nth ugly number*/
int getNthUglyNo(int n)
{
int i = 1;
int count = 1; /* ugly number count */
/*Check for all integers untill ugly count
becomes n*/
while (n > count)
{
i++;
if (isUgly(i))
count++;
}
return i;
}
/* Driver program to test above functions */
int main()
{
unsigned no = getNthUglyNo(150);
printf("150th ugly no. is %d ", no);
getchar();
return 0;
}

This method is not time efficient as it checks for all integers until ugly number count becomes n, but space complexity of this method is O(1)
METHOD 2 (Use Dynamic Programming)
Here is a time efficient solution with O(n) extra space. The ugly-number sequence is 1, 2, 3, 4, 5, 6, 8, 9, 10, 12, 15,

because every number can only be divided by 2, 3, 5, one way to look at the sequence is to split the sequence to three groups as below:
(1) 12, 22, 32, 42, 52,
(2) 13, 23, 33, 43, 53,
(3) 15, 25, 35, 45, 55,
We can find that every subsequence is the ugly-sequence itself (1, 2, 3, 4, 5, ) multiply 2, 3, 5. Then we use similar merge method as merge sort,
to get every ugly number from the three subsequence. Every step we choose the smallest one, and move one step after.
Algorithm:
1 Declare an array for ugly numbers: ugly[150]
2 Initialize first ugly no: ugly[0] = 1
3 Initialize three array index variables i2, i3, i5 to point to
1st element of the ugly array:
i2 = i3 = i5 =0;
4 Initialize 3 choices for the next ugly no:
next_mulitple_of_2 = ugly[i2]*2;
next_mulitple_of_3 = ugly[i3]*3
next_mulitple_of_5 = ugly[i5]*5;
5 Now go in a loop to fill all ugly numbers till 150:
For (i = 1; i < 150; i++ )
{
/* These small steps are not optimized for good
readability. Will optimize them in C program */
next_ugly_no = Min(next_mulitple_of_2,
next_mulitple_of_3,
next_mulitple_of_5);
if (next_ugly_no == next_mulitple_of_2)
{
i2 = i2 + 1;
next_mulitple_of_2 = ugly[i2]*2;
}
if (next_ugly_no == next_mulitple_of_3)
{
i3 = i3 + 1;
next_mulitple_of_3 = ugly[i3]*3;
}
if (next_ugly_no == next_mulitple_of_5)
{
i5 = i5 + 1;
next_mulitple_of_5 = ugly[i5]*5;
}
ugly[i] = next_ugly_no
}/* end of for loop */
6.return next_ugly_no

Example:
Let us see how it works
initialize
ugly[] = | 1 |
i2 = i3 = i5 = 0;
First iteration
ugly[1] = Min(ugly[i2]*2, ugly[i3]*3, ugly[i5]*5)
= Min(2, 3, 5)
= 2
ugly[] = | 1 | 2 |
i2 = 1, i3 = i5 = 0 (i2 got incremented )
Second iteration
ugly[2] = Min(ugly[i2]*2, ugly[i3]*3, ugly[i5]*5)
= Min(4, 3, 5)
= 3
ugly[] = | 1 | 2 | 3 |
i2 = 1, i3 = 1, i5 = 0 (i3 got incremented )
Third iteration
ugly[3] = Min(ugly[i2]*2, ugly[i3]*3, ugly[i5]*5)
= Min(4, 6, 5)
= 4
ugly[] = | 1 | 2 | 3 | 4 |
i2 = 2, i3 = 1, i5 = 0 (i2 got incremented )
Fourth iteration
ugly[4] = Min(ugly[i2]*2, ugly[i3]*3, ugly[i5]*5)
= Min(6, 6, 5)
= 5
ugly[] = | 1 | 2 | 3 | 4 | 5 |

i2 = 2, i3 = 1, i5 = 1 (i5 got incremented )

Fifth iteration
ugly[4] = Min(ugly[i2]*2, ugly[i3]*3, ugly[i5]*5)
= Min(6, 6, 10)
= 6
ugly[] = | 1 | 2 | 3 | 4 | 5 | 6 |
i2 = 3, i3 = 2, i5 = 1 (i2 and i3 got incremented )
Will continue same way till I < 150

Program:
# include<stdio.h>
# include<stdlib.h>
# define bool int
/* Function to find minimum of 3 numbers */
unsigned min(unsigned , unsigned , unsigned );
/* Function to get the nth ugly number*/
unsigned getNthUglyNo(unsigned n)
{
unsigned *ugly =
(unsigned *)(malloc (sizeof(unsigned)*n));
unsigned i2 = 0, i3 = 0, i5 = 0;
unsigned i;
unsigned next_multiple_of_2 = 2;
unsigned next_multiple_of_3 = 3;
unsigned next_multiple_of_5 = 5;
unsigned next_ugly_no = 1;
*(ugly+0) = 1;
for(i=1; i<n; i++)
{
next_ugly_no = min(next_multiple_of_2,
next_multiple_of_3,
next_multiple_of_5);
*(ugly+i) = next_ugly_no;
if(next_ugly_no == next_multiple_of_2)
{
i2 = i2+1;
next_multiple_of_2 = *(ugly+i2)*2;
}
if(next_ugly_no == next_multiple_of_3)
{
i3 = i3+1;
next_multiple_of_3 = *(ugly+i3)*3;
}
if(next_ugly_no == next_multiple_of_5)
{
i5 = i5+1;
next_multiple_of_5 = *(ugly+i5)*5;
}
} /*End of for loop (i=1; i<n; i++) */
return next_ugly_no;
}
/* Function to find minimum of 3 numbers */
unsigned min(unsigned a, unsigned b, unsigned c)
{
if(a <= b)
{
if(a <= c)
return a;
else
return c;
}
if(b <= c)
return b;
else
return c;
}
/* Driver program to test above functions */
int main()
{
unsigned no = getNthUglyNo(150);
printf("%dth ugly no. is %d ", 150, no);
getchar();
return 0;

Algorithmic Paradigm: Dynamic Programming

Time Complexity: O(n)
Storage Complexity: O(n)

Largest Sum Contiguous Subarray

Write an efficient C program to find the sum of contiguous subarray within a one-dimensional array of numbers which has the largest sum.
Kadanes Algorithm:
Initialize:
max_so_far = 0
max_ending_here = 0
Loop for each element of the array
(a) max_ending_here = max_ending_here + a[i]
(b) if(max_ending_here < 0)
max_ending_here = 0
(c) if(max_so_far < max_ending_here)
max_so_far = max_ending_here
return max_so_far

Explanation:
Simple idea of the Kadane's algorithm is to look for all positive contiguous segments of the array (max_ending_here is used for this). And keep
track of maximum sum contiguous segment among all positive segments (max_so_far is used for this). Each time we get a positive sum compare it
with max_so_far and update max_so_far if it is greater than max_so_far
Lets take the example:
{-2, -3, 4, -1, -2, 1, 5, -3}
max_so_far = max_ending_here = 0
for i=0, a[0] = -2
max_ending_here = max_ending_here + (-2)
Set max_ending_here = 0 because max_ending_here < 0
for i=1, a[1] = -3
max_ending_here = max_ending_here + (-3)
Set max_ending_here = 0 because max_ending_here < 0
for i=2, a[2] = 4
max_ending_here = max_ending_here + (4)
max_ending_here = 4
max_so_far is updated to 4 because max_ending_here greater
than max_so_far which was 0 till now
for i=3, a[3] = -1
max_ending_here = max_ending_here + (-1)
max_ending_here = 3
for i=4, a[4] = -2
max_ending_here = max_ending_here + (-2)
max_ending_here = 1
for i=5, a[5] = 1
max_ending_here = max_ending_here + (1)
max_ending_here = 2
for i=6, a[6] = 5
max_ending_here = max_ending_here + (5)
max_ending_here = 7
max_so_far is updated to 7 because max_ending_here is
greater than max_so_far
for i=7, a[7] = -3
max_ending_here = max_ending_here + (-3)
max_ending_here = 4

Program:

C++
// C++ program to print largest contiguous array sum
#include<iostream>
using namespace std;
int maxSubArraySum(int a[], int size)
{
int max_so_far = 0, max_ending_here = 0;
for (int i = 0; i < size; i++)
{

max_ending_here = max_ending_here + a[i];

if (max_ending_here < 0)
max_ending_here = 0;
if (max_so_far < max_ending_here)
max_so_far = max_ending_here;
}
return max_so_far;
}
/*Driver program to test maxSubArraySum*/
int main()
{
int a[] = {-2, -3, 4, -1, -2, 1, 5, -3};
int n = sizeof(a)/sizeof(a[0]);
int max_sum = maxSubArraySum(a, n);
cout << "Maximum contiguous sum is \n" << max_sum;
return 0;
}

Python
# Python program to find maximum contiguous subarray
# Function to find the maximum contiguous subarray
def maxSubArraySum(a,size):
max_so_far = 0
max_ending_here = 0
for i in range(0, size):
max_ending_here = max_ending_here + a[i]
if max_ending_here < 0:
max_ending_here = 0
if (max_so_far < max_ending_here):
max_so_far = max_ending_here
return max_so_far
# Driver function to check the above function
a = [-2, -3, 4, -1, -2, 1, 5, -3]
print"Maximum contiguous sum is", maxSubArraySum(a,len(a))
#This code is contributed by _Devesh Agrawal_

Maximum contiguous sum is 7

Notes:
Algorithm doesn't work for all negative numbers. It simply returns 0 if all numbers are negative. For handling this we can add an extra phase before
actual implementation. The phase will look if all numbers are negative, if they are it will return maximum of them (or smallest in terms of absolute
value). There may be other ways to handle it though.
Above program can be optimized further, if we compare max_so_far with max_ending_here only if max_ending_here is greater than 0.

C++
int maxSubArraySum(int a[], int size)
{
int max_so_far = 0, max_ending_here = 0;
for (int i = 0; i < size; i++)
{
max_ending_here = max_ending_here + a[i];
if (max_ending_here < 0)
max_ending_here = 0;
/* Do not compare for all elements. Compare only
when max_ending_here > 0 */
else if (max_so_far < max_ending_here)
max_so_far = max_ending_here;
}
return max_so_far;
}

Python

def maxSubArraySum(a,size):
max_so_far = 0
max_ending_here = 0
for i in range(0, size):
max_ending_here = max_ending_here + a[i]
if max_ending_here < 0:
max_ending_here = 0
# Do not compare for all elements. Compare only
# when max_ending_here > 0
elif (max_so_far < max_ending_here):
max_so_far = max_ending_here
return max_so_far

Time Complexity: O(n)

Algorithmic Paradigm: Dynamic Programming
Following is another simple implementation suggested by Mohit Kumar. The implementation handles the case when all numbers in array are
negative.

Longest Palindromic Substring | Set 1

Given a string, find the longest substring which is palindrome. For example, if the given string is forgeeksskeegfor, the output should be
geeksskeeg.
Method 1 ( Brute Force )
The simple approach is to check each substring whether the substring is a palindrome or not. We can run three loops, the outer two loops pick all
substrings one by one by fixing the corner characters, the inner loop checks whether the picked substring is palindrome or not.
Time complexity: O ( n^3 )
Auxiliary complexity: O ( 1 )
Method 2 ( Dynamic Programming )
The time complexity can be reduced by storing results of subproblems. The idea is similar to this post. We maintain a boolean table[n][n] that is
filled in bottom up manner. The value of table[i][j] is true, if the substring is palindrome, otherwise false. To calculate table[i][j], we first check the
value of table[i+1][j-1], if the value is true and str[i] is same as str[j], then we make table[i][j] true. Otherwise, the value of table[i][j] is made false.
// A dynamic programming solution for longest palindr.
// This code is adopted from following link
// http://www.leetcode.com/2011/11/longest-palindromic-substring-part-i.html
#include <stdio.h>
#include <string.h>
// A utility function to print a substring str[low..high]
void printSubStr( char* str, int low, int high )
{
for( int i = low; i <= high; ++i )
printf("%c", str[i]);
}
// This function prints the longest palindrome substring
// of str[].
// It also returns the length of the longest palindrome
int longestPalSubstr( char *str )
{
int n = strlen( str ); // get length of input string
// table[i][j] will be false if substring str[i..j]
// is not palindrome.
// Else table[i][j] will be true
bool table[n][n];
memset(table, 0, sizeof(table));
// All substrings of length 1 are palindromes
int maxLength = 1;
for (int i = 0; i < n; ++i)
table[i][i] = true;
// check for sub-string of length 2.
int start = 0;
for (int i = 0; i < n-1; ++i)
{
if (str[i] == str[i+1])
{
table[i][i+1] = true;
start = i;
maxLength = 2;
}
}
// Check for lengths greater than 2. k is length
// of substring
for (int k = 3; k <= n; ++k)
{
// Fix the starting index
for (int i = 0; i < n-k+1 ; ++i)
{
// Get the ending index of substring from
// starting index i and length k
int j = i + k - 1;
//
//
//
if
{

checking for sub-string from ith index to

jth index iff str[i+1] to str[j-1] is a
palindrome
(table[i+1][j-1] && str[i] == str[j])
table[i][j] = true;

if (k > maxLength)
{
start = i;
maxLength = k;
}
}
}
}
printf("Longest palindrome substring is: ");
printSubStr( str, start, start + maxLength - 1 );
return maxLength; // return length of LPS
}
// Driver program to test above functions
int main()
{
char str[] = "forgeeksskeegfor";
printf("\nLength is: %d\n", longestPalSubstr( str ) );
return 0;
}

Output:
Longest palindrome substring is: geeksskeeg
Length is: 10

Time complexity: O ( n^2 )

Auxiliary Space: O ( n^2 )
We will soon be adding more optimized methods as separate posts.

Dynamic Programming | Set 23 (BellmanFord Algorithm)

Given a graph and a source vertex src in graph, find shortest paths from src to all vertices in the given graph. The graph may contain negative
weight edges.
We have discussed Dijkstras algorithm for this problem. Dijksras algorithm is a Greedy algorithm and time complexity is O(VLogV) (with the use
of Fibonacci heap). Dijkstra doesnt work for Graphs with negative weight edges, Bellman-Ford works for such graphs. Bellman-Ford is
also simpler than Dijkstra and suites well for distributed systems. But time complexity of Bellman-Ford is O(VE), which is more than
Dijkstra.
Algorithm
Following are the detailed steps.
Input: Graph and a source vertex src
Output: Shortest distance to all vertices from src. If there is a negative weight cycle, then shortest distances are not calculated, negative weight
cycle is reported.
1) This step initializes distances from source to all vertices as infinite and distance to source itself as 0. Create an array dist[] of size |V| with all
values as infinite except dist[src] where src is source vertex.
2) This step calculates shortest distances. Do following |V|-1 times where |V| is the number of vertices in given graph.
..a) Do following for each edge u-v
If dist[v] > dist[u] + weight of edge uv, then update dist[v]
.dist[v] = dist[u] + weight of edge uv
3) This step reports if there is a negative weight cycle in graph. Do following for each edge u-v
If dist[v] > dist[u] + weight of edge uv, then Graph contains negative weight cycle
The idea of step 3 is, step 2 guarantees shortest distances if graph doesnt contain negative weight cycle. If we iterate through all edges one more
time and get a shorter path for any vertex, then there is a negative weight cycle
How does this work? Like other Dynamic Programming Problems, the algorithm calculate shortest paths in bottom-up manner. It first calculates
the shortest distances for the shortest paths which have at-most one edge in the path. Then, it calculates shortest paths with at-nost 2 edges, and so
on. After the ith iteration of outer loop, the shortest paths with at most i edges are calculated. There can be maximum |V| 1 edges in any simple
path, that is why the outer loop runs |v| 1 times. The idea is, assuming that there is no negative weight cycle, if we have calculated shortest paths
with at most i edges, then an iteration over all edges guarantees to give shortest path with at-most (i+1) edges (Proof is simple, you can refer this or
MIT Video Lecture)
Example
Let us understand the algorithm with following example graph. The images are taken from this source.
Let the given source vertex be 0. Initialize all distances as infinite, except the distance to source itself. Total number of vertices in the graph is 5, so
all edges must be processed 4 times.

Let all edges are processed in following order: (B,E), (D,B), (B,D), (A,B), (A,C), (D,C), (B,C), (E,D). We get following distances when all edges
are processed first time. The first row in shows initial distances. The second row shows distances when edges (B,E), (D,B), (B,D) and (A,B) are
processed. The third row shows distances when (A,C) is processed. The fourth row shows when (D,C), (B,C) and (E,D) are processed.

The first iteration guarantees to give all shortest paths which are at most 1 edge long. We get following distances when all edges are processed
second time (The last row shows final values).

The second iteration guarantees to give all shortest paths which are at most 2 edges long. The algorithm processes all edges 2 more times. The
distances are minimized after the second iteration, so third and fourth iterations dont update the distances.
Implementation:

C++
// A C / C++ program for Bellman-Ford's single source
// shortest path algorithm.
#include
#include
#include
#include

<stdio.h>
<stdlib.h>
<string.h>
<limits.h>

// a structure to represent a weighted edge in graph

struct Edge
{
int src, dest, weight;
};
// a structure to represent a connected, directed and
// weighted graph
struct Graph
{
// V-> Number of vertices, E-> Number of edges
int V, E;
// graph is represented as an array of edges.
struct Edge* edge;
};
// Creates a graph with V vertices and E edges
struct Graph* createGraph(int V, int E)
{
struct Graph* graph =
(struct Graph*) malloc( sizeof(struct Graph) );
graph->V = V;
graph->E = E;
graph->edge =
(struct Edge*) malloc( graph->E * sizeof( struct Edge ) );
return graph;
}
// A utility function used to print the solution
void printArr(int dist[], int n)
{
printf("Vertex Distance from Source\n");
for (int i = 0; i < n; ++i)
printf("%d \t\t %d\n", i, dist[i]);
}
// The main function that finds shortest distances from src to
// all other vertices using Bellman-Ford algorithm. The function
// also detects negative weight cycle
void BellmanFord(struct Graph* graph, int src)
{
int V = graph->V;
int E = graph->E;
int dist[V];
// Step 1: Initialize distances from src to all other vertices
// as INFINITE
for (int i = 0; i < V; i++)
dist[i] = INT_MAX;
dist[src] = 0;

// Step 2: Relax all edges |V| - 1 times. A simple shortest

// path from src to any other vertex can have at-most |V| - 1
// edges
for (int i = 1; i <= V-1; i++)
{
for (int j = 0; j < E; j++)
{
int u = graph->edge[j].src;
int v = graph->edge[j].dest;
int weight = graph->edge[j].weight;
if (dist[u] != INT_MAX && dist[u] + weight < dist[v])
dist[v] = dist[u] + weight;
}
}
// Step 3: check for negative-weight cycles. The above step
// guarantees shortest distances if graph doesn't contain
// negative weight cycle. If we get a shorter path, then there
// is a cycle.
for (int i = 0; i < E; i++)
{
int u = graph->edge[i].src;
int v = graph->edge[i].dest;
int weight = graph->edge[i].weight;
if (dist[u] != INT_MAX && dist[u] + weight < dist[v])
printf("Graph contains negative weight cycle");
}
printArr(dist, V);
return;
}
// Driver program to test above functions
int main()
{
/* Let us create the graph given in above example */
int V = 5; // Number of vertices in graph
int E = 8; // Number of edges in graph
struct Graph* graph = createGraph(V, E);
// add edge 0-1 (or A-B in above figure)
graph->edge[0].src = 0;
graph->edge[0].dest = 1;
graph->edge[0].weight = -1;
// add edge 0-2 (or A-C in above figure)
graph->edge[1].src = 0;
graph->edge[1].dest = 2;
graph->edge[1].weight = 4;
// add edge 1-2 (or B-C in above figure)
graph->edge[2].src = 1;
graph->edge[2].dest = 2;
graph->edge[2].weight = 3;
// add edge 1-3 (or B-D in above figure)
graph->edge[3].src = 1;
graph->edge[3].dest = 3;
graph->edge[3].weight = 2;
// add edge 1-4 (or A-E in above figure)
graph->edge[4].src = 1;
graph->edge[4].dest = 4;
graph->edge[4].weight = 2;
// add edge 3-2 (or D-C in above figure)
graph->edge[5].src = 3;
graph->edge[5].dest = 2;
graph->edge[5].weight = 5;
// add edge 3-1 (or D-B in above figure)
graph->edge[6].src = 3;
graph->edge[6].dest = 1;
graph->edge[6].weight = 1;
// add edge 4-3 (or E-D in above figure)
graph->edge[7].src = 4;
graph->edge[7].dest = 3;
graph->edge[7].weight = -3;

BellmanFord(graph, 0);
return 0;
}

Java
// A Java program for Bellman-Ford's single source shortest path
// algorithm.
import java.util.*;
import java.lang.*;
import java.io.*;
// A class to represent a connected, directed and weighted graph
class Graph
{
// A class to represent a weighted edge in graph
class Edge {
int src, dest, weight;
Edge() {
src = dest = weight = 0;
}
};
int V, E;
Edge edge[];
// Creates a graph with V vertices and E edges
Graph(int v, int e)
{
V = v;
E = e;
edge = new Edge[e];
for (int i=0; i<e; ++i)
edge[i] = new Edge();
}
// The main function that finds shortest distances from src
// to all other vertices using Bellman-Ford algorithm. The
// function also detects negative weight cycle
void BellmanFord(Graph graph,int src)
{
int V = graph.V, E = graph.E;
int dist[] = new int[V];
// Step 1: Initialize distances from src to all other
// vertices as INFINITE
for (int i=0; i<V; ++i)
dist[i] = Integer.MAX_VALUE;
dist[src] = 0;
// Step 2: Relax all edges |V| - 1 times. A simple
// shortest path from src to any other vertex can
// have at-most |V| - 1 edges
for (int i=1; i<V; ++i)
{
for (int j=0; j<E; ++j)
{
int u = graph.edge[j].src;
int v = graph.edge[j].dest;
int weight = graph.edge[j].weight;
if (dist[u]!=Integer.MAX_VALUE &&
dist[u]+weight<dist[v])
dist[v]=dist[u]+weight;
}
}
// Step 3: check for negative-weight cycles. The above
// step guarantees shortest distances if graph doesn't
// contain negative weight cycle. If we get a shorter
// path, then there is a cycle.
for (int j=0; j<E; ++j)
{
int u = graph.edge[j].src;
int v = graph.edge[j].dest;
int weight = graph.edge[j].weight;
if (dist[u]!=Integer.MAX_VALUE &&
dist[u]+weight<dist[v])
System.out.println("Graph contains negative weight cycle");

}
printArr(dist, V);
}
// A utility function used to print the solution
void printArr(int dist[], int V)
{
System.out.println("Vertex Distance from Source");
for (int i=0; i<V; ++i)
System.out.println(i+"\t\t"+dist[i]);
}
// Driver method to test above function
public static void main(String[] args)
{
int V = 5; // Number of vertices in graph
int E = 8; // Number of edges in graph
Graph graph = new Graph(V, E);
// add edge 0-1 (or A-B in above figure)
graph.edge[0].src = 0;
graph.edge[0].dest = 1;
graph.edge[0].weight = -1;
// add edge 0-2 (or A-C in above figure)
graph.edge[1].src = 0;
graph.edge[1].dest = 2;
graph.edge[1].weight = 4;
// add edge 1-2 (or B-C in above figure)
graph.edge[2].src = 1;
graph.edge[2].dest = 2;
graph.edge[2].weight = 3;
// add edge 1-3 (or B-D in above figure)
graph.edge[3].src = 1;
graph.edge[3].dest = 3;
graph.edge[3].weight = 2;
// add edge 1-4 (or A-E in above figure)
graph.edge[4].src = 1;
graph.edge[4].dest = 4;
graph.edge[4].weight = 2;
// add edge 3-2 (or D-C in above figure)
graph.edge[5].src = 3;
graph.edge[5].dest = 2;
graph.edge[5].weight = 5;
// add edge 3-1 (or D-B in above figure)
graph.edge[6].src = 3;
graph.edge[6].dest = 1;
graph.edge[6].weight = 1;
// add edge 4-3 (or E-D in above figure)
graph.edge[7].src = 4;
graph.edge[7].dest = 3;
graph.edge[7].weight = -3;
graph.BellmanFord(graph, 0);
}
}
// Contributed by Aakash Hasija

Vertex
0
1
2
3
4

Distance from Source

0
-1
2
-2
1

Notes
1) Negative weights are found in various applications of graphs. For example, instead of paying cost for a path, we may get some advantage if we
follow the path.
2) Bellman-Ford works better (better than Dijksras) for distributed systems. Unlike Dijksras where we need to find minimum value of all vertices,
in Bellman-Ford, edges are considered one by one.

Exercise
1) The standard Bellman-Ford algorithm reports shortest path only if there is no negative weight cycles. Modify it so that it reports minimum
distances even if there is a negative weight cycle.
2) Can we use Dijksras algorithm for shortest paths for graphs with negative weights one idea can be, calculate the minimum weight value, add a
positive value (equal to absolute value of minimum weight value) to all weights and run the Dijksras algorithm for the modified graph. Will this
algorithm work?
References:
http://www.youtube.com/watch?v=Ttezuzs39nk
http://en.wikipedia.org/wiki/Bellman%E2%80%93Ford_algorithm
http://www.cs.arizona.edu/classes/cs445/spring07/ShortestPath2.prn.pdf

Dynamic Programming | Set 24 (Optimal Binary Search Tree)

Given a sorted array keys[0.. n-1] of search keys and an array freq[0.. n-1] of frequency counts, where freq[i] is the number of searches to
keys[i]. Construct a binary search tree of all keys such that the total cost of all the searches is as small as possible.
Let us first define the cost of a BST. The cost of a BST node is level of that node multiplied by its frequency. Level of root is 1.
Example 1
Input: keys[] = {10, 12}, freq[] = {34, 50}
There can be following two possible BSTs
10
12
\
/
12
10
I
II
Frequency of searches of 10 and 12 are 34 and 50 respectively.
The cost of tree I is 34*1 + 50*2 = 134
The cost of tree II is 50*1 + 34*2 = 118
Example 2
Input: keys[] = {10, 12, 20}, freq[] = {34, 8, 50}
There can be following possible BSTs
10
12
20
10
\
/
\
/
\
12
10
20
12
20
\
/
/
20
10
12
I
II
III
IV
Among all possible BSTs, cost of the fifth BST is minimum.
Cost of the fifth BST is 1*50 + 2*34 + 3*8 = 142

20
/
10
\
12
V

1) Optimal Substructure:
The optimal cost for freq[i..j] can be recursively calculated using following formula.

We need to calculate optCost(0, n-1) to find the result.

The idea of above formula is simple, we one by one try all nodes as root (r varies from i to j in second term). When we make rth node as root, we
recursively calculate optimal cost from i to r-1 and r+1 to j.
We add sum of frequencies from i to j (see first term in the above formula), this is added because every search will go through root and one
comparison will be done for every search.
2) Overlapping Subproblems
Following is recursive implementation that simply follows the recursive structure mentioned above.
// A naive recursive implementation of optimal binary search tree problem
#include <stdio.h>
#include <limits.h>
// A utility function to get sum of array elements freq[i] to freq[j]
int sum(int freq[], int i, int j);
// A recursive function to calculate cost of optimal binary search tree
int optCost(int freq[], int i, int j)
{
// Base cases
if (j < i)
// If there are no elements in this subarray
return 0;
if (j == i)
// If there is one element in this subarray
return freq[i];
// Get sum of freq[i], freq[i+1], ... freq[j]
int fsum = sum(freq, i, j);
// Initialize minimum value
int min = INT_MAX;
// One by one consider all elements as root and recursively find cost
// of the BST, compare the cost with min and update min if needed
for (int r = i; r <= j; ++r)
{
int cost = optCost(freq, i, r-1) + optCost(freq, r+1, j);
if (cost < min)
min = cost;
}
// Return minimum value

return min + fsum;

}
// The main function that calculates minimum cost of a Binary Search Tree.
// It mainly uses optCost() to find the optimal cost.
int optimalSearchTree(int keys[], int freq[], int n)
{
// Here array keys[] is assumed to be sorted in increasing order.
// If keys[] is not sorted, then add code to sort keys, and rearrange
// freq[] accordingly.
return optCost(freq, 0, n-1);
}
// A utility function to get sum of array elements freq[i] to freq[j]
int sum(int freq[], int i, int j)
{
int s = 0;
for (int k = i; k <=j; k++)
s += freq[k];
return s;
}
// Driver program to test above functions
int main()
{
int keys[] = {10, 12, 20};
int freq[] = {34, 8, 50};
int n = sizeof(keys)/sizeof(keys[0]);
printf("Cost of Optimal BST is %d ", optimalSearchTree(keys, freq, n));
return 0;
}

Output:
Cost of Optimal BST is 142

Time complexity of the above naive recursive approach is exponential. It should be noted that the above function computes the same subproblems
again and again. We can see many subproblems being repeated in the following recursion tree for freq[1..4].

Since same suproblems are called again, this problem has Overlapping Subprolems property. So optimal BST problem has both properties (see
this and this) of a dynamic programming problem. Like other typical Dynamic Programming(DP) problems, recomputations of same subproblems
can be avoided by constructing a temporary array cost[][] in bottom up manner.
Dynamic Programming Solution
Following is C/C++ implementation for optimal BST problem using Dynamic Programming. We use an auxiliary array cost[n][n] to store the
solutions of subproblems. cost[0][n-1] will hold the final result. The challenge in implementation is, all diagonal values must be filled first, then the
values which lie on the line just above the diagonal. In other words, we must first fill all cost[i][i] values, then all cost[i][i+1] values, then all cost[i]
[i+2] values. So how to fill the 2D array in such manner> The idea used in the implementation is same as Matrix Chain Multiplication problem, we
use a variable L for chain length and increment L, one by one. We calculate column number j using the values of i and L.
// Dynamic Programming code for Optimal Binary Search Tree Problem
#include <stdio.h>
#include <limits.h>
// A utility function to get sum of array elements freq[i] to freq[j]
int sum(int freq[], int i, int j);
/* A Dynamic Programming based function that calculates minimum cost of
a Binary Search Tree. */
int optimalSearchTree(int keys[], int freq[], int n)
{
/* Create an auxiliary 2D matrix to store results of subproblems */
int cost[n][n];
/* cost[i][j] = Optimal cost of binary search tree that can be
formed from keys[i] to keys[j].

cost[0][n-1] will store the resultant cost */

// For a single key, cost is equal to frequency of the key
for (int i = 0; i < n; i++)
cost[i][i] = freq[i];
// Now we need to consider chains of length 2, 3, ... .
// L is chain length.
for (int L=2; L<=n; L++)
{
// i is row number in cost[][]
for (int i=0; i<=n-L+1; i++)
{
// Get column number j from row number i and chain length L
int j = i+L-1;
cost[i][j] = INT_MAX;
// Try making all keys in interval keys[i..j] as root
for (int r=i; r<=j; r++)
{
// c = cost when keys[r] becomes root of this subtree
int c = ((r > i)? cost[i][r-1]:0) +
((r < j)? cost[r+1][j]:0) +
sum(freq, i, j);
if (c < cost[i][j])
cost[i][j] = c;
}
}
}
return cost[0][n-1];
}
// A utility function to get sum of array elements freq[i] to freq[j]
int sum(int freq[], int i, int j)
{
int s = 0;
for (int k = i; k <=j; k++)
s += freq[k];
return s;
}
// Driver program to test above functions
int main()
{
int keys[] = {10, 12, 20};
int freq[] = {34, 8, 50};
int n = sizeof(keys)/sizeof(keys[0]);
printf("Cost of Optimal BST is %d ", optimalSearchTree(keys, freq, n));
return 0;
}

Output:
Cost of Optimal BST is 142

Notes
1) The time complexity of the above solution is O(n^4). The time complexity can be easily reduced to O(n^3) by pre-calculating sum of
frequencies instead of calling sum() again and again.
2) In the above solutions, we have computed optimal cost only. The solutions can be easily modified to store the structure of BSTs also. We can
create another auxiliary array of size n to store the structure of tree. All we need to do is, store the chosen r in the innermost loop.

Dynamic Programming | Set 26 (Largest Independent Set Problem)

Given a Binary Tree, find size of the Largest Independent Set(LIS) in it. A subset of all tree nodes is an independent set if there is no edge
between any two nodes of the subset.
For example, consider the following binary tree. The largest independent set(LIS) is {10, 40, 60, 70, 80} and size of the LIS is 5.

A Dynamic Programming solution solves a given problem using solutions of subproblems in bottom up manner. Can the given problem be solved
using solutions to subproblems? If yes, then what are the subproblems? Can we find largest independent set size (LISS) for a node X if we know
LISS for all descendants of X? If a node is considered as part of LIS, then its children cannot be part of LIS, but its grandchildren can be.
Following is optimal substructure property.
1) Optimal Substructure:
Let LISS(X) indicates size of largest independent set of a tree with root X.
LISS(X) = MAX { (1 + sum of LISS for all grandchildren of X),
(sum of LISS for all children of X) }

The idea is simple, there are two possibilities for every node X, either X is a member of the set or not a member. If X is a member, then the value
of LISS(X) is 1 plus LISS of all grandchildren. If X is not a member, then the value is sum of LISS of all children.
2) Overlapping Subproblems
Following is recursive implementation that simply follows the recursive structure mentioned above.
// A naive recursive implementation of Largest Independent Set problem
#include <stdio.h>
#include <stdlib.h>
// A utility function to find max of two integers
int max(int x, int y) { return (x > y)? x: y; }
/* A binary tree node has data, pointer to left child and a pointer to
right child */
struct node
{
int data;
struct node *left, *right;
};
// The function returns size of the largest independent set in a given
// binary tree
int LISS(struct node *root)
{
if (root == NULL)
return 0;
// Caculate size excluding the current node
int size_excl = LISS(root->left) + LISS(root->right);
// Calculate size including the current node
int size_incl = 1;
if (root->left)
size_incl += LISS(root->left->left) + LISS(root->left->right);
if (root->right)
size_incl += LISS(root->right->left) + LISS(root->right->right);
// Return the maximum of two sizes
return max(size_incl, size_excl);
}
// A utility function to create a node
struct node* newNode( int data )

{
struct node* temp = (struct node *) malloc( sizeof(struct node) );
temp->data = data;
temp->left = temp->right = NULL;
return temp;
}
// Driver program to test above functions
int main()
{
// Let us construct the tree given in the above diagram
struct node *root
= newNode(20);
root->left
= newNode(8);
root->left->left
= newNode(4);
root->left->right
= newNode(12);
root->left->right->left = newNode(10);
root->left->right->right = newNode(14);
root->right
= newNode(22);
root->right->right
= newNode(25);
printf ("Size of the Largest Independent Set is %d ", LISS(root));
return 0;
}

Output:
Size of the Largest Independent Set is 5

Time complexity of the above naive recursive approach is exponential. It should be noted that the above function computes the same subproblems
again and again. For example, LISS of node with value 50 is evaluated for node with values 10 and 20 as 50 is grandchild of 10 and child of 20.
Since same suproblems are called again, this problem has Overlapping Subprolems property. So LISS problem has both properties (see this and
this) of a dynamic programming problem. Like other typical Dynamic Programming(DP) problems, recomputations of same subproblems can be
avoided by storing the solutions to subproblems and solving problems in bottom up manner.
Following is C implementation of Dynamic Programming based solution. In the following solution, an additional field liss is added to tree nodes.
The initial value of liss is set as 0 for all nodes. The recursive function LISS() calculates liss for a node only if it is not already set.
/* Dynamic programming based program for Largest Independent Set problem */
#include <stdio.h>
#include <stdlib.h>
// A utility function to find max of two integers
int max(int x, int y) { return (x > y)? x: y; }
/* A binary tree node has data, pointer to left child and a pointer to
right child */
struct node
{
int data;
int liss;
struct node *left, *right;
};
// A memoization function returns size of the largest independent set in
// a given binary tree
int LISS(struct node *root)
{
if (root == NULL)
return 0;
if (root->liss)
return root->liss;
if (root->left == NULL && root->right == NULL)
return (root->liss = 1);
// Calculate size excluding the current node
int liss_excl = LISS(root->left) + LISS(root->right);
// Calculate size including the current node
int liss_incl = 1;
if (root->left)
liss_incl += LISS(root->left->left) + LISS(root->left->right);
if (root->right)
liss_incl += LISS(root->right->left) + LISS(root->right->right);
// Maximum of two sizes is LISS, store it for future uses.
root->liss = max(liss_incl, liss_excl);

return root->liss;
}
// A utility function to create a node
struct node* newNode(int data)
{
struct node* temp = (struct node *) malloc( sizeof(struct node) );
temp->data = data;
temp->left = temp->right = NULL;
temp->liss = 0;
return temp;
}
// Driver program to test above functions
int main()
{
// Let us construct the tree given in the above diagram
struct node *root
= newNode(20);
root->left
= newNode(8);
root->left->left
= newNode(4);
root->left->right
= newNode(12);
root->left->right->left = newNode(10);
root->left->right->right = newNode(14);
root->right
= newNode(22);
root->right->right
= newNode(25);
printf ("Size of the Largest Independent Set is %d ", LISS(root));
return 0;
}

Output
Size of the Largest Independent Set is 5

Time Complexity: O(n) where n is the number of nodes in given Binary tree.
Following extensions to above solution can be tried as an exercise.
1) Extend the above solution for n-ary tree.
2) The above solution modifies the given tree structure by adding an additional field liss to tree nodes. Extend the solution so that it doesnt modify
the tree structure.
3) The above solution only returns size of LIS, it doesnt print elements of LIS. Extend the solution to print all nodes that are part of LIS.

Dynamic Programming | Set 25 (Subset Sum Problem)

Given a set of non-negative integers, and a value sum, determine if there is a subset of the given set with sum equal to given sum.
Examples: set[] = {3, 34, 4, 12, 5, 2}, sum = 9
Output: True //There is a subset (4, 5) with sum 9.

Let isSubSetSum(int set[], int n, int sum) be the function to find whether there is a subset of set[] with sum equal to sum. n is the number of
elements in set[].
The isSubsetSum problem can be divided into two subproblems
a) Include the last element, recur for n = n-1, sum = sum set[n-1]
b) Exclude the last element, recur for n = n-1.
If any of the above the above subproblems return true, then return true.
Following is the recursive formula for isSubsetSum() problem.
isSubsetSum(set, n, sum) = isSubsetSum(set, n-1, sum) ||
isSubsetSum(arr, n-1, sum-set[n-1])
Base Cases:
isSubsetSum(set, n, sum) = false, if sum > 0 and n == 0
isSubsetSum(set, n, sum) = true, if sum == 0

Following is naive recursive implementation that simply follows the recursive structure mentioned above.
// A recursive solution for subset sum problem
#include <stdio.h>
// Returns true if there is a subset of set[] with sun equal to given sum
bool isSubsetSum(int set[], int n, int sum)
{
// Base Cases
if (sum == 0)
return true;
if (n == 0 && sum != 0)
return false;
// If last element is greater than sum, then ignore it
if (set[n-1] > sum)
return isSubsetSum(set, n-1, sum);
/* else, check if sum can be obtained by any of the following
(a) including the last element
(b) excluding the last element */
return isSubsetSum(set, n-1, sum) || isSubsetSum(set, n-1, sum-set[n-1]);
}
// Driver program to test above function
int main()
{
int set[] = {3, 34, 4, 12, 5, 2};
int sum = 9;
int n = sizeof(set)/sizeof(set[0]);
if (isSubsetSum(set, n, sum) == true)
printf("Found a subset with given sum");
else
printf("No subset with given sum");
return 0;
}

Output:
Found a subset with given sum

The above solution may try all subsets of given set in worst case. Therefore time complexity of the above solution is exponential. The problem is infact NP-Complete (There is no known polynomial time solution for this problem).
We can solve the problem in Pseudo-polynomial time using Dynamic programming. We create a boolean 2D table subset[][] and fill it in
bottom up manner. The value of subset[i][j] will be true if there is a subset of set[0..j-1] with sum equal to i., otherwise false. Finally, we return
subset[sum][n]
// A Dynamic Programming solution for subset sum problem
#include <stdio.h>
// Returns true if there is a subset of set[] with sun equal to given sum
bool isSubsetSum(int set[], int n, int sum)
{

// The value of subset[i][j] will be true if there is a subset of set[0..j-1]

// with sum equal to i
bool subset[sum+1][n+1];
// If sum is 0, then answer is true
for (int i = 0; i <= n; i++)
subset[0][i] = true;
// If sum is not 0 and set is empty, then answer is false
for (int i = 1; i <= sum; i++)
subset[i][0] = false;
// Fill the subset table in botton up manner
for (int i = 1; i <= sum; i++)
{
for (int j = 1; j <= n; j++)
{
subset[i][j] = subset[i][j-1];
if (i >= set[j-1])
subset[i][j] = subset[i][j] || subset[i - set[j-1]][j-1];
}
}
/* // uncomment this code to print table
for (int i = 0; i <= sum; i++)
{
for (int j = 0; j <= n; j++)
printf ("%4d", subset[i][j]);
printf("\n");
} */
return subset[sum][n];
}
// Driver program to test above function
int main()
{
int set[] = {3, 34, 4, 12, 5, 2};
int sum = 9;
int n = sizeof(set)/sizeof(set[0]);
if (isSubsetSum(set, n, sum) == true)
printf("Found a subset with given sum");
else
printf("No subset with given sum");
return 0;
}

Output:
Found a subset with given sum

Time complexity of the above solution is O(sum*n).

Dynamic Programming | Set 27 (Maximum sum rectangle in a 2D matrix)

Given a 2D array, find the maximum sum subarray in it. For example, in the following 2D array, the maximum sum subarray is highlighted with blue
rectangle and sum of this subarray is 29.

This problem is mainly an extension of Largest Sum Contiguous Subarray for 1D array.
The naive solution for this problem is to check every possible rectangle in given 2D array. This solution requires 4 nested loops and time
complexity of this solution would be O(n^4).
Kadanes algorithm for 1D array can be used to reduce the time complexity to O(n^3). The idea is to fix the left and right columns one by one
and find the maximum sum contiguous rows for every left and right column pair. We basically find top and bottom row numbers (which have
maximum sum) for every fixed left and right column pair. To find the top and bottom row numbers, calculate sun of elements in every row from left
to right and store these sums in an array say temp[]. So temp[i] indicates sum of elements from left to right in row i. If we apply Kadanes 1D
algorithm on temp[], and get the maximum sum subarray of temp, this maximum sum would be the maximum possible sum with left and right as
boundary columns. To get the overall maximum sum, we compare this sum with the maximum sum so far.
// Program to find maximum sum subarray in a given 2D array
#include <stdio.h>
#include <string.h>
#include <limits.h>
#define ROW 4
#define COL 5
// Implementation of Kadane's algorithm for 1D array. The function returns the
// maximum sum and stores starting and ending indexes of the maximum sum subarray
// at addresses pointed by start and finish pointers respectively.
int kadane(int* arr, int* start, int* finish, int n)
{
// initialize sum, maxSum and
int sum = 0, maxSum = INT_MIN, i;
// Just some initial value to check for all negative values case
*finish = -1;
// local variable
int local_start = 0;
for (i = 0; i < n; ++i)
{
sum += arr[i];
if (sum < 0)
{
sum = 0;
local_start = i+1;
}
else if (sum > maxSum)
{
maxSum = sum;
*start = local_start;
*finish = i;
}
}
// There is at-least one non-negative number
if (*finish != -1)
return maxSum;
// Special Case: When all numbers in arr[] are negative
maxSum = arr[0];
*start = *finish = 0;

// Find the maximum element in array

for (i = 1; i < n; i++)
{
if (arr[i] > maxSum)
{
maxSum = arr[i];
*start = *finish = i;
}
}
return maxSum;
}
// The main function that finds maximum sum rectangle in M[][]
void findMaxSum(int M[][COL])
{
// Variables to store the final output
int maxSum = INT_MIN, finalLeft, finalRight, finalTop, finalBottom;
int left, right, i;
int temp[ROW], sum, start, finish;
// Set the left column
for (left = 0; left < COL; ++left)
{
// Initialize all elements of temp as 0
memset(temp, 0, sizeof(temp));
// Set the right column for the left column set by outer loop
for (right = left; right < COL; ++right)
{
// Calculate sum between current left and right for every row 'i'
for (i = 0; i < ROW; ++i)
temp[i] += M[i][right];
// Find the maximum sum subarray in temp[]. The kadane() function
// also sets values of start and finish. So 'sum' is sum of
// rectangle between (start, left) and (finish, right) which is the
// maximum sum with boundary columns strictly as left and right.
sum = kadane(temp, &start, &finish, ROW);
// Compare sum with maximum sum so far. If sum is more, then update
// maxSum and other output values
if (sum > maxSum)
{
maxSum = sum;
finalLeft = left;
finalRight = right;
finalTop = start;
finalBottom = finish;
}
}
}
// Print final values
printf("(Top, Left) (%d, %d)\n", finalTop, finalLeft);
printf("(Bottom, Right) (%d, %d)\n", finalBottom, finalRight);
printf("Max sum is: %d\n", maxSum);
}
// Driver program to test above functions
int main()
{
int M[ROW][COL] = {{1, 2, -1, -4, -20},
{-8, -3, 4, 2, 1},
{3, 8, 10, 1, 3},
{-4, -1, 1, 7, -6}
};
findMaxSum(M);
return 0;
}

Output:
(Top, Left) (1, 1)
(Bottom, Right) (3, 3)
Max sum is: 29

Time Complexity: O(n^3)

Count number of binary strings without consecutive 1s

Given a positive integer N, count all possible distinct binary strings of length N such that there are no consecutive 1s.
Examples:
Input: N = 2
Output: 3
// The 3 strings are 00, 01, 10
Input: N = 3
Output: 5
// The 5 strings are 000, 001, 010, 100, 101

This problem can be solved using Dynamic Programming. Let a[i] be the number of binary strings of length i which do not contain any two
consecutive 1s and which end in 0. Similarly, let b[i] be the number of such strings which end in 1. We can append either 0 or 1 to a string ending
in 0, but we can only append 0 to a string ending in 1. This yields the recurrence relation:
a[i] = a[i - 1] + b[i - 1]
b[i] = a[i - 1]

The base cases of above recurrence are a[1] = b[1] = 1. The total number of strings of length i is just a[i] + b[i].
Following is C++ implementation of above solution. In the following implementation, indexes start from 0. So a[i] represents the number of binary
strings for input length i+1. Similarly, b[i] represents binary strings for input length i+1.
// C++ program to count all distinct binary strings
// without two consecutive 1's
#include <iostream>
using namespace std;
int countStrings(int n)
{
int a[n], b[n];
a[0] = b[0] = 1;
for (int i = 1; i < n; i++)
{
a[i] = a[i-1] + b[i-1];
b[i] = a[i-1];
}
return a[n-1] + b[n-1];
}
// Driver program to test above functions
int main()
{
cout << countStrings(3) << endl;
return 0;
}

Output:
5

Source:
courses.csail.mit.edu/6.006/oldquizzes/solutions/q2-f2009-sol.pdf

Dynamic Programming | Set 37 (Boolean Parenthesization Problem)

Given a boolean expression with following symbols.
Symbols
'T' ---> true
'F' ---> false

And following operators filled between symbols

Operators
& ---> boolean AND
| ---> boolean OR
^ ---> boolean XOR

Count the number of ways we can parenthesize the expression so that the value of expression evaluates to true.
Let the input be in form of two arrays one contains the symbols (T and F) in order and other contains operators (&, | and ^}
Examples:
Input: symbol[]
= {T, F, T}
operator[] = {^, &}
Output: 2
The given expression is "T ^ F & T", it evaluates true
in two ways "((T ^ F) & T)" and "(T ^ (F & T))"
Input: symbol[]
= {T, F, F}
operator[] = {^, |}
Output: 2
The given expression is "T ^ F | F", it evaluates true
in two ways "( (T ^ F) | F )" and "( T ^ (F | F) )".
Input: symbol[]
= {T, T, F, T}
operator[] = {|, &, ^}
Output: 4
The given expression is "T | T & F ^ T", it evaluates true
in 4 ways ((T|T)&(F^T)), (T|(T&(F^T))), (((T|T)&F)^T)
and (T|((T&F)^T)).

Solution:
Let T(i, j) represents the number of ways to parenthesize the symbols between i and j (both inclusive) such that the subexpression between i and j
evaluates to true.

Let F(i, j) represents the number of ways to parenthesize the symbols between i and j (both inclusive) such that the subexpression between i and j
evaluates to false.

Base Cases:
T(i, i) = 1 if symbol[i] = 'T'
T(i, i) = 0 if symbol[i] = 'F'
F(i, i) = 1 if symbol[i] = 'F'
F(i, i) = 0 if symbol[i] = 'T'

If we draw recursion tree of above recursive solution, we can observe that it many overlapping subproblems. Like other dynamic programming
problems, it can be solved by filling a table in bottom up manner. Following is C++ implementation of dynamic programming solution.
#include<iostream>
#include<cstring>
using namespace std;
// Returns count of all possible parenthesizations that lead to
// result true for a boolean expression with symbols like true
// and false and operators like &, | and ^ filled between symbols

int countParenth(char symb[], char oper[], int n)

{
int F[n][n], T[n][n];
// Fill diaginal entries first
// All diagonal entries in T[i][i] are 1 if symbol[i]
// is T (true). Similarly, all F[i][i] entries are 1 if
// symbol[i] is F (False)
for (int i = 0; i < n; i++)
{
F[i][i] = (symb[i] == 'F')? 1: 0;
T[i][i] = (symb[i] == 'T')? 1: 0;
}
// Now fill T[i][i+1], T[i][i+2], T[i][i+3]... in order
// And F[i][i+1], F[i][i+2], F[i][i+3]... in order
for (int gap=1; gap<n; ++gap)
{
for (int i=0, j=gap; j<n; ++i, ++j)
{
T[i][j] = F[i][j] = 0;
for (int g=0; g<gap; g++)
{
// Find place of parenthesization using current value
// of gap
int k = i + g;
// Store Total[i][k] and Total[k+1][j]
int tik = T[i][k] + F[i][k];
int tkj = T[k+1][j] + F[k+1][j];
// Follow the recursive formulas according to the current
// operator
if (oper[k] == '&')
{
T[i][j] += T[i][k]*T[k+1][j];
F[i][j] += (tik*tkj - T[i][k]*T[k+1][j]);
}
if (oper[k] == '|')
{
F[i][j] += F[i][k]*F[k+1][j];
T[i][j] += (tik*tkj - F[i][k]*F[k+1][j]);
}
if (oper[k] == '^')
{
T[i][j] += F[i][k]*T[k+1][j] + T[i][k]*F[k+1][j];
F[i][j] += T[i][k]*T[k+1][j] + F[i][k]*F[k+1][j];
}
}
}
}
return T[0][n-1];
}
// Driver program to test above function
int main()
{
char symbols[] = "TTFT";
char operators[] = "|&^";
int n = strlen(symbols);
// There are 4 ways
// ((T|T)&(F^T)), (T|(T&(F^T))), (((T|T)&F)^T) and (T|((T&F)^T))
cout << countParenth(symbols, operators, n);
return 0;
}

Output:
4

Time Complexity: O(n3)

Auxiliary Space: O(n2)
References:
http://people.cs.clemson.edu/~bcdean/dp_practice/dp_9.swf

Count ways to reach the nth stair

There are n stairs, a person standing at the bottom wants to reach the top. The person can climb either 1 stair or 2 stairs at a time. Count the
number of ways, the person can reach the top.

Consider the example shown in diagram. The value of n is 3. There are 3 ways to reach the top. The diagram is taken from Easier Fibonacci
puzzles

More Examples:
Input: n = 1
Output: 1
There is only one way to climb 1 stair
Input: n = 2
Output: 2
There are two ways: (1, 1) and (2)
Input: n = 4
Output: 5
(1, 1, 1, 1), (1, 1, 2), (2, 1, 1), (1, 2, 1), (2, 2)

We can easily find recursive nature in above problem. The person can reach nth stair from either (n-1)th stair or from (n-2)th stair. Let the total
number of ways to reach nt stair be ways(n). The value of ways(n) can be written as following.
ways(n) = ways(n-1) + ways(n-2)

The above expression is actually the expression for Fibonacci numbers, but there is one thing to notice, the value of ways(n) is equal to
fibonacci(n+1).
ways(1) = fib(2) = 1
ways(2) = fib(3) = 2
ways(3) = fib(4) = 3
So we can use function for fibonacci numbers to find the value of ways(n). Following is C++ implementation of the above idea.
// A C program to count number of ways to reach n't stair when
// a person can climb 1, 2, ..m stairs at a time.
#include<stdio.h>
// A simple recursive program to find n'th fibonacci number
int fib(int n)
{
if (n <= 1)
return n;
return fib(n-1) + fib(n-2);
}
// Returns number of ways to reach s'th stair
int countWays(int s)
{
return fib(s + 1);
}
// Driver program to test above functions

int main ()
{
int s = 4;
printf("Number of ways = %d", countWays(s));
getchar();
return 0;
}

Output:
Number of ways = 5

The time complexity of the above implementation is exponential (golden ratio raised to power n). It can be optimized to work in O(Logn) time
using the previously discussed Fibonacci function optimizations.
Generalization of the above problem
How to count number of ways if the person can climb up to m stairs for a given value m? For example if m is 4, the person can climb 1 stair or 2
stairs or 3 stairs or 4 stairs at a time.
We can write the recurrence as following.
ways(n, m) = ways(n-1, m) + ways(n-2, m) + ... ways(n-m, m)

Following is C++ implementation of above recurrence.

// A C program to count number of ways to reach n't stair when
// a person can climb either 1 or 2 stairs at a time
#include<stdio.h>
// A recursive function used by countWays
int countWaysUtil(int n, int m)
{
if (n <= 1)
return n;
int res = 0;
for (int i = 1; i<=m && i<=n; i++)
res += countWaysUtil(n-i, m);
return res;
}
// Returns number of ways to reach s'th stair
int countWays(int s, int m)
{
return countWaysUtil(s+1, m);
}
// Driver program to test above functions
int main ()
{
int s = 4, m = 2;
printf("Nuber of ways = %d", countWays(s, m));
return 0;
}

Output:
Number of ways = 5

The time complexity of above solution is exponential. It can be optimized to O(mn) by using dynamic programming. Following is dynamic
programming based solution. We build a table res[] in bottom up manner.
// A C program to count number of ways to reach n't stair when
// a person can climb 1, 2, ..m stairs at a time
#include<stdio.h>
// A recursive function used by countWays
int countWaysUtil(int n, int m)
{
int res[n];
res[0] = 1; res[1] = 1;
for (int i=2; i<n; i++)
{
res[i] = 0;
for (int j=1; j<=m && j<=i; j++)
res[i] += res[i-j];
}
return res[n-1];
}

// Returns number of ways to reach s'th stair

int countWays(int s, int m)
{
return countWaysUtil(s+1, m);
}
// Driver program to test above functions
int main ()
{
int s = 4, m = 2;
printf("Nuber of ways = %d", countWays(s, m));
return 0;
}

Output:
Number of ways = 5

Minimum Cost Polygon Triangulation

A triangulation of a convex polygon is formed by drawing diagonals between non-adjacent vertices (corners) such that the diagonals never
intersect. The problem is to find the cost of triangulation with the minimum cost. The cost of a triangulation is sum of the weights of its component
triangles. Weight of each triangle is its perimeter (sum of lengths of all sides)
See following example taken from this source.

Two triangulations of the same convex pentagon. The triangulation on the left has a cost of 8 + 2?2 + 2?5 (approximately 15.30), the
one on the right has a cost of 4 + 2?2 + 4?5 (approximately 15.77).
This problem has recursive substructure. The idea is to divide the polygon into three parts: a single triangle, the sub-polygon to the left, and the
sub-polygon to the right. We try all possible divisions like this and find the one that minimizes the cost of the triangle plus the cost of the
triangulation of the two sub-polygons.
Let Minimum Cost of triangulation of vertices from i to j be minCost(i, j)
If j <= i + 2 Then
minCost(i, j) = 0
Else
minCost(i, j) = Min { minCost(i, k) + minCost(k, j) + cost(i, k, j) }
Here k varies from 'i+1' to 'j-1'
Cost of a triangle formed by edges (i, j), (j, k) and (k, j) is
cost(i, j, k) = dist(i, j) + dist(j, k) + dist(k, j)

Following is C++ implementation of above naive recursive formula.

// Recursive implementation for minimum cost convex polygon triangulation
#include <iostream>
#include <cmath>
#define MAX 1000000.0
using namespace std;
// Structure of a point in 2D plane
struct Point
{
int x, y;
};
// Utility function to find minimum of two double values
double min(double x, double y)
{
return (x <= y)? x : y;
}
// A utility function
double dist(Point p1,
{
return sqrt((p1.x
(p1.y
}

to find distance between two points in a plane

Point p2)
- p2.x)*(p1.x - p2.x) +
- p2.y)*(p1.y - p2.y));

// A utility function to find cost of a triangle. The cost is considered

// as perimeter (sum of lengths of all edges) of the triangle
double cost(Point points[], int i, int j, int k)
{
Point p1 = points[i], p2 = points[j], p3 = points[k];
return dist(p1, p2) + dist(p2, p3) + dist(p3, p1);
}
// A recursive function to find minimum cost of polygon triangulation
// The polygon is represented by points[i..j].
double mTC(Point points[], int i, int j)
{
// There must be at least three points between i and j
// (including i and j)
if (j < i+2)

return 0;
// Initialize result as infinite
double res = MAX;
// Find minimum triangulation by considering all
for (int k=i+1; k<j; k++)
res = min(res, (mTC(points, i, k) + mTC(points, k, j) +
cost(points, i, k, j)));
return res;
}
// Driver program to test above functions
int main()
{
Point points[] = {{0, 0}, {1, 0}, {2, 1}, {1, 2}, {0, 2}};
int n = sizeof(points)/sizeof(points[0]);
cout << mTC(points, 0, n-1);
return 0;
}

Output:
15.3006

The above problem is similar to Matrix Chain Multiplication. The following is recursion tree for mTC(points[], 0, 4).

It can be easily seen in the above recursion tree that the problem has many overlapping subproblems. Since the problem has both properties:
Optimal Substructure and Overlapping Subproblems, it can be efficiently solved using dynamic programming.
Following is C++ implementation of dynamic programming solution.
// A Dynamic Programming based program to find minimum cost of convex
// polygon triangulation
#include <iostream>
#include <cmath>
#define MAX 1000000.0
using namespace std;
// Structure of a point in 2D plane
struct Point
{
int x, y;
};
// Utility function to find minimum of two double values
double min(double x, double y)
{
return (x <= y)? x : y;
}
// A utility function
double dist(Point p1,
{
return sqrt((p1.x
(p1.y
}

to find distance between two points in a plane

Point p2)
- p2.x)*(p1.x - p2.x) +
- p2.y)*(p1.y - p2.y));

// A utility function to find cost of a triangle. The cost is considered

// as perimeter (sum of lengths of all edges) of the triangle
double cost(Point points[], int i, int j, int k)
{

Point p1 = points[i], p2 = points[j], p3 = points[k];

return dist(p1, p2) + dist(p2, p3) + dist(p3, p1);
}
// A Dynamic programming based function to find minimum cost for convex
// polygon triangulation.
double mTCDP(Point points[], int n)
{
// There must be at least 3 points to form a triangle
if (n < 3)
return 0;
// table to store results of subproblems. table[i][j] stores cost of
// triangulation of points from i to j. The entry table[0][n-1] stores
// the final result.
double table[n][n];
// Fill table using above recursive formula. Note that the table
// is filled in diagonal fashion i.e., from diagonal elements to
// table[0][n-1] which is the result.
for (int gap = 0; gap < n; gap++)
{
for (int i = 0, j = gap; j < n; i++, j++)
{
if (j < i+2)
table[i][j] = 0.0;
else
{
table[i][j] = MAX;
for (int k = i+1; k < j; k++)
{
double val = table[i][k] + table[k][j] + cost(points,i,j,k);
if (table[i][j] > val)
table[i][j] = val;
}
}
}
}
return table[0][n-1];
}
// Driver program to test above functions
int main()
{
Point points[] = {{0, 0}, {1, 0}, {2, 1}, {1, 2}, {0, 2}};
int n = sizeof(points)/sizeof(points[0]);
cout << mTCDP(points, n);
return 0;
}

Output:
15.3006

Time complexity of the above dynamic programming solution is O(n3).

Please note that the above implementations assume that the points of covnvex polygon are given in order (either clockwise or anticlockwise)
Exercise:
Extend the above solution to print triangulation also. For the above example, the optimal triangulation is 0 3 4, 0 1 3, and 1 2 3.
Sources:
http://www.cs.utexas.edu/users/djimenez/utsa/cs3343/lecture12.html
http://www.cs.utoronto.ca/~heap/Courses/270F02/A4/chains/node2.html

Mobile Numeric Keypad Problem

Given the mobile numeric keypad. You can only press buttons that are up, left, right or down to the current
button. You are not allowed to press bottom row corner buttons (i.e. * and # ).
Given a number N, find out the number of possible numbers of given length.
Examples:
For N=1, number of possible numbers would be 10 (0, 1, 2, 3, ., 9)
For N=2, number of possible numbers would be 36
Possible numbers: 00,08 11,12,14 22,21,23,25 and so on.
If we start with 0, valid numbers will be 00, 08 (count: 2)
If we start with 1, valid numbers will be 11, 12, 14 (count: 3)
If we start with 2, valid numbers will be 22, 21, 23,25 (count: 4)
If we start with 3, valid numbers will be 33, 32, 36 (count: 3)
If we start with 4, valid numbers will be 44,41,45,47 (count: 4)
If we start with 5, valid numbers will be 55,54,52,56,58 (count: 5)
We need to print the count of possible numbers.
N = 1 is trivial case, number of possible numbers would be 10 (0, 1, 2, 3, ., 9)
For N > 1, we need to start from some button, then move to any of the four direction (up, left, right or down) which takes to a valid button (should
not go to *, #). Keep doing this until N length number is obtained (depth first traversal).
Recursive Solution:
Mobile Keypad is a rectangular grid of 4X3 (4 rows and 3 columns)
Lets say Count(i, j, N) represents the count of N length numbers starting from position (i, j)
If N = 1
Count(i, j, N) = 10
Else
Count(i, j, N) = Sum of all Count(r, c, N-1) where (r, c) is new
position after valid move of length 1 from current
position (i, j)

Following is C implementation of above recursive formula.

// A Naive Recursive C program to count number of possible numbers
// of given length
#include <stdio.h>
// left, up, right, down move from current location
int row[] = {0, 0, -1, 0, 1};
int col[] = {0, -1, 0, 1, 0};
// Returns count of numbers of length n starting from key position
// (i, j) in a numeric keyboard.
int getCountUtil(char keypad[][3], int i, int j, int n)
{
if (keypad == NULL || n <= 0)
return 0;
// From a given key, only one number is possible of length 1
if (n == 1)
return 1;
int k=0, move=0, ro=0, co=0, totalCount = 0;
// move left, up, right, down from current location and if
// new location is valid, then get number count of length
// (n-1) from that new position and add in count obtained so far
for (move=0; move<5; move++)
{
ro = i + row[move];
co = j + col[move];
if (ro >= 0 && ro <= 3 && co >=0 && co <= 2 &&

keypad[ro][co] != '*' && keypad[ro][co] != '#')

{
totalCount += getCountUtil(keypad, ro, co, n-1);
}
}
return totalCount;
}
// Return count of all possible numbers of length n
// in a given numeric keyboard
int getCount(char keypad[][3], int n)
{
// Base cases
if (keypad == NULL || n <= 0)
return 0;
if (n == 1)
return 10;
int i=0, j=0, totalCount = 0;
for (i=0; i<4; i++) // Loop on keypad row
{
for (j=0; j<3; j++) // Loop on keypad column
{
// Process for 0 to 9 digits
if (keypad[i][j] != '*' && keypad[i][j] != '#')
{
// Get count when number is starting from key
// position (i, j) and add in count obtained so far
totalCount += getCountUtil(keypad, i, j, n);
}
}
}
return totalCount;
}
// Driver program to test above function
int main(int argc, char *argv[])
{
char keypad[4][3] = {{'1','2','3'},
{'4','5','6'},
{'7','8','9'},
{'*','0','#'}};
printf("Count for numbers of length %d:
printf("Count for numbers of length %d:
printf("Count for numbers of length %d:
printf("Count for numbers of length %d:
printf("Count for numbers of length %d:

%d\n",
%d\n",
%d\n",
%d\n",
%d\n",

1,
2,
3,
4,
5,

getCount(keypad,
getCount(keypad,
getCount(keypad,
getCount(keypad,
getCount(keypad,

1));
2));
3));
4));
5));

return 0;
}

Output:
Count
Count
Count
Count
Count

for
for
for
for
for

numbers
numbers
numbers
numbers
numbers

of
of
of
of
of

length
length
length
length
length

1:
2:
3:
4:
5:

10
36
138
532
2062

Dynamic Programming
There are many repeated traversal on smaller paths (traversal for smaller N) to find all possible longer paths (traversal for bigger N). See following
two diagrams for example. In this traversal, for N = 4 from two starting positions (buttons 4 and 8), we can see there are few repeated traversals
for N = 2 (e.g. 4 -> 1, 6 -> 3, 8 -> 9, 8 -> 7 etc).

Since the problem has both properties: Optimal Substructure and Overlapping Subproblems, it can be efficiently solved using dynamic
programming.
Following is C program for dynamic programming implementation.
// A Dynamic Programming based C program to count number of
// possible numbers of given length
#include <stdio.h>
// Return count of all possible numbers of length n
// in a given numeric keyboard
int getCount(char keypad[][3], int n)
{
if(keypad == NULL || n <= 0)
return 0;
if(n == 1)
return 10;
// left, up, right, down move from current location
int row[] = {0, 0, -1, 0, 1};
int col[] = {0, -1, 0, 1, 0};
// taking n+1 for simplicity - count[i][j] will store
// number count starting with digit i and length j
int count[10][n+1];
int i=0, j=0, k=0, move=0, ro=0, co=0, num = 0;
int nextNum=0, totalCount = 0;
// count numbers starting with digit i and of lengths 0 and 1
for (i=0; i<=9; i++)
{
count[i][0] = 0;
count[i][1] = 1;
}
// Bottom up - Get number count of length 2, 3, 4, ... , n
for (k=2; k<=n; k++)
{
for (i=0; i<4; i++) // Loop on keypad row
{
for (j=0; j<3; j++) // Loop on keypad column
{
// Process for 0 to 9 digits
if (keypad[i][j] != '*' && keypad[i][j] != '#')
{
// Here we are counting the numbers starting with

// digit keypad[i][j] and of length k keypad[i][j]

// will become 1st digit, and we need to look for
// (k-1) more digits
num = keypad[i][j] - '0';
count[num][k] = 0;
// move left, up, right, down from current location
// and if new location is valid, then get number
// count of length (k-1) from that new digit and
// add in count we found so far
for (move=0; move<5; move++)
{
ro = i + row[move];
co = j + col[move];
if (ro >= 0 && ro <= 3 && co >=0 && co <= 2 &&
keypad[ro][co] != '*' && keypad[ro][co] != '#')
{
nextNum = keypad[ro][co] - '0';
count[num][k] += count[nextNum][k-1];
}
}
}
}
}
}
// Get count of all possible numbers of length "n" starting
// with digit 0, 1, 2, ..., 9
totalCount = 0;
for (i=0; i<=9; i++)
totalCount += count[i][n];
return totalCount;
}
// Driver program to test above function
int main(int argc, char *argv[])
{
char keypad[4][3] = {{'1','2','3'},
{'4','5','6'},
{'7','8','9'},
{'*','0','#'}};
printf("Count for numbers of length %d:
printf("Count for numbers of length %d:
printf("Count for numbers of length %d:
printf("Count for numbers of length %d:
printf("Count for numbers of length %d:

%d\n",
%d\n",
%d\n",
%d\n",
%d\n",

1,
2,
3,
4,
5,

getCount(keypad,
getCount(keypad,
getCount(keypad,
getCount(keypad,
getCount(keypad,

1));
2));
3));
4));
5));

return 0;
}

Output:
Count
Count
Count
Count
Count

for
for
for
for
for

numbers
numbers
numbers
numbers
numbers

of
of
of
of
of

length
length
length
length
length

1:
2:
3:
4:
5:

10
36
138
532
2062

A Space Optimized Solution:

The above dynamic programming approach also runs in O(n) time and requires O(n) auxiliary space, as only one for loop runs n times, other for
loops runs for constant time. We can see that nth iteration needs data from (n-1)th iteration only, so we need not keep the data from older
iterations. We can have a space efficient dynamic programming approach with just two arrays of size 10. Thanks to Nik for suggesting this
solution.
// A Space Optimized C program to count number of possible numbers
// of given length
#include <stdio.h>
// Return count of all possible numbers of length n
// in a given numeric keyboard
int getCount(char keypad[][3], int n)
{
if(keypad == NULL || n <= 0)
return 0;
if(n == 1)
return 10;
// odd[i], even[i] arrays represent count of numbers starting
// with digit i for any length j
int odd[10], even[10];

int i = 0, j = 0, useOdd = 0, totalCount = 0;

for (i=0; i<=9; i++)
odd[i] = 1; // for j = 1
for (j=2; j<=n; j++) // Bottom Up calculation from j = 2 to n
{
useOdd = 1 - useOdd;
// Here we are explicitly writing lines for each number 0
// to 9. But it can always be written as DFS on 4X3 grid
// using row, column array valid moves
if(useOdd == 1)
{
even[0] = odd[0] + odd[8];
even[1] = odd[1] + odd[2] + odd[4];
even[2] = odd[2] + odd[1] + odd[3] + odd[5];
even[3] = odd[3] + odd[2] + odd[6];
even[4] = odd[4] + odd[1] + odd[5] + odd[7];
even[5] = odd[5] + odd[2] + odd[4] + odd[8] + odd[6];
even[6] = odd[6] + odd[3] + odd[5] + odd[9];
even[7] = odd[7] + odd[4] + odd[8];
even[8] = odd[8] + odd[0] + odd[5] + odd[7] + odd[9];
even[9] = odd[9] + odd[6] + odd[8];
}
else
{
odd[0] = even[0] + even[8];
odd[1] = even[1] + even[2] + even[4];
odd[2] = even[2] + even[1] + even[3] + even[5];
odd[3] = even[3] + even[2] + even[6];
odd[4] = even[4] + even[1] + even[5] + even[7];
odd[5] = even[5] + even[2] + even[4] + even[8] + even[6];
odd[6] = even[6] + even[3] + even[5] + even[9];
odd[7] = even[7] + even[4] + even[8];
odd[8] = even[8] + even[0] + even[5] + even[7] + even[9];
odd[9] = even[9] + even[6] + even[8];
}
}
// Get count of all possible numbers of length "n" starting
// with digit 0, 1, 2, ..., 9
totalCount = 0;
if(useOdd == 1)
{
for (i=0; i<=9; i++)
totalCount += even[i];
}
else
{
for (i=0; i<=9; i++)
totalCount += odd[i];
}
return totalCount;
}
// Driver program to test above function
int main()
{
char keypad[4][3] = {{'1','2','3'},
{'4','5','6'},
{'7','8','9'},
{'*','0','#'}
};
printf("Count for numbers of length %d:
printf("Count for numbers of length %d:
printf("Count for numbers of length %d:
printf("Count for numbers of length %d:
printf("Count for numbers of length %d:
return 0;
}

Output:
Count
Count
Count
Count
Count

for
for
for
for
for

numbers
numbers
numbers
numbers
numbers

of
of
of
of
of

length
length
length
length
length

1:
2:
3:
4:
5:

10
36
138
532
2062

%d\n",
%d\n",
%d\n",
%d\n",
%d\n",

1,
2,
3,
4,
5,

getCount(keypad,
getCount(keypad,
getCount(keypad,
getCount(keypad,
getCount(keypad,

1));
2));
3));
4));
5));

Count of n digit numbers whose sum of digits equals to given sum

Given two integers n and sum, find count of all n digit numbers with sum of digits as sum. Leading 0s are not counted as digits.
1 <= n <= 100 and 1 <= sum <= 50000
Example:
Input: n = 2, sum = 2
Output: 2
Explanation: Numbers are 11 and 20
Input: n = 2, sum = 5
Output: 5
Explanation: Numbers are 14, 23, 32, 41 and 50
Input: n = 3, sum = 6
Output: 21

The idea is simple, we subtract all values from 0 to 9 from given sum and recur for sum minus that digit. Below is recursive formula.
countRec(n, sum) = ?finalCount(n-1, sum-x)
where 1 =< x <= 9 and
sum-x >= 0
One important observation is, leading 0's must be
handled explicitly as they are not counted as digits.
So our final count can be written as below.
finalCount(n, sum) = ?finalCount(n-1, sum-x)
where 0 =< x <= 9 and
sum-x >= 0

Below is a simple recursive solution based on above recursive formula.

// A recursive program to count numbers with sum
// of digits as given 'sum'
#include<bits/stdc++.h>
using namespace std;
// Recursive function to count 'n' digit numbers
// with sum of digits as 'sum'. This function
// considers leading 0's also as digits, that is
// why not directly called
unsigned long long int countRec(int n, int sum)
{
// Base case
if (n == 0)
return sum == 0;
// Initialize answer
unsigned long long int ans = 0;
// Traverse through every digit and count
// numbers beginning with it using recursion
for (int i=0; i<=9; i++)
if (sum-i >= 0)
ans += countRec(n-1, sum-i);
return ans;
}
// This is mainly a wrapper over countRec. It
// explicitly handles leading digit and calls
// countRec() for remaining digits.
unsigned long long int finalCount(int n, int sum)
{
// Initialize final answer
unsigned long long int ans = 0;
// Traverse through every digit from 1 to
// 9 and count numbers beginning with it
for (int i = 1; i <= 9; i++)
if (sum-i >= 0)
ans += countRec(n-1, sum-i);
return ans;
}
// Driver program
int main()

{
int n = 2, sum = 5;
cout << finalCount(n, sum);
return 0;
}

Output:
5

The time complexity of above solution is exponential. If we draw the complete recursion tree, we can observer that many subproblems are solved
again and again. For example, if we start with n = 3 and sum = 10, we can reach n = 1, sum = 8, by considering digit sequences 1,1 or 2, 0.
Since same suproblems are called again, this problem has Overlapping Subprolems property. So min square sum problem has both properties (see
this and this) of a dynamic programming problem.
Below is Memoization based C++ implementation.
// A memoization based recursive program to count
// numbers with sum of n as given 'sum'
#include<bits/stdc++.h>
using namespace std;
// A lookup table used for memoization
unsigned long long int lookup[101][50001];
// Memoizatiob based implementation of recursive
// function
unsigned long long int countRec(int n, int sum)
{
// Base case
if (n == 0)
return sum == 0;
// If this subproblem is already evaluated,
// return the evaluated value
if (lookup[n][sum] != -1)
return lookup[n][sum];
// Initialize answer
unsigned long long int ans = 0;
// Traverse through every digit and
// recursively count numbers beginning
// with it
for (int i=0; i<10; i++)
if (sum-i >= 0)
ans += countRec(n-1, sum-i);
return lookup[n][sum] = ans;
}
// This is mainly a wrapper over countRec. It
// explicitly handles leading digit and calls
// countRec() for remaining n.
unsigned long long int finalCount(int n, int sum)
{
// Initialize all entries of lookup table
memset(lookup, -1, sizeof lookup);
// Initialize final answer
unsigned long long int ans = 0;
// Traverse through every digit from 1 to
// 9 and count numbers beginning with it
for (int i = 1; i <= 9; i++)
if (sum-i >= 0)
ans += countRec(n-1, sum-i);
return ans;
}
// Driver program
int main()
{
int n = 3, sum = 5;
cout << finalCount(n, sum);
return 0;
}

Output:

Thanks to Gaurav Ahirwar for suggesting above solution.

Minimum Initial Points to Reach Destination

Given a grid with each cell consisting of positive, negative or no points i.e, zero points. We can move across a cell only if we have positive points (
> 0 ). Whenever we pass through a cell, points in that cell are added to our overall points. We need to find minimum initial points to reach cell (m1, n-1) from (0, 0).
Constraints :

Total number of non-decreasing numbers with n digits

A number is non-decreasing if every digit (except the first one) is greater than or equal to previous digit. For example, 223, 4455567, 899, are
non-decreasing numbers.
So, given the number of digits n, you are required to find the count of total non-decreasing numbers with n digits.
Examples:
Input: n = 1
Output: count = 10
Input: n = 2
Output: count = 55
Input: n = 3
Output: count = 220

One way to look at the problem is, count of numbers is equal to count n digit number ending with 9 plus count of ending with digit 8 plus count for
7 and so on. How to get count ending with a particular digit? We can recur for n-1 length and digits smaller than or equal to the last digit. So below
is recursive formula.
Count of n digit numbers = (Count of (n-1) digit numbers Ending with digit 9) +
(Count of (n-1) digit numbers Ending with digit 8) +
.............................................+
.............................................+
(Count of (n-1) digit numbers Ending with digit 0)

Let count ending with digit d and length n be count(n, d)

count(n, d) = ? (count(n-1, i)) where i varies from 0 to d
Total count = ? count(n-1, d) where d varies from 0 to n-1

The above recursive solution is going to have many overlapping subproblems. Therefore, we can use Dynamic Programming to build a table in
bottom up manner. Below is Dynamic programming based C++ program.
// C++ program to count non-decreasing number with n digits
#include<bits/stdc++.h>
using namespace std;
long long int countNonDecreasing(int n)
{
// dp[i][j] contains total count of non decreasing
// numbers ending with digit i and of length j
long long int dp[10][n+1];
memset(dp, 0, sizeof dp);
// Fill table for non decreasing numbers of length 1
// Base cases 0, 1, 2, 3, 4, 5, 6, 7, 8, 9
for (int i = 0; i < 10; i++)
dp[i][1] = 1;
// Fill the table in bottom-up manner
for (int digit = 0; digit <= 9; digit++)
{
// Compute total numbers of non decreasing
// numbers of length 'len'
for (int len = 2; len <= n; len++)
{
// sum of all numbers of length of len-1
// in which last digit x is <= 'digit'
for (int x = 0; x <= digit; x++)
dp[digit][len] += dp[x][len-1];
}
}
long long int count = 0;
// There total nondecreasing numbers of length n
// wiint be dp[0][n] + dp[1][n] ..+ dp[9][n]
for (int i = 0; i < 10; i++)
count += dp[i][n];
return count;
}
// Driver program

int main()
{
int n = 3;
cout << countNonDecreasing(n);
return 0;
}

Output:
220

Thanks to Gaurav Ahirwar for suggesting above method.

Another method is based on below direct formula
Count of non-decreasing numbers with n digits =
N*(N+1)/2*(N+2)/3* ....*(N+n-1)/n
Where N = 10

Below is a C++ program to compute count using above formula.

// C++ program to count non-decreasing numner with n digits
#include<bits/stdc++.h>
using namespace std;
long long int countNonDecreasing(int n)
{
int N = 10;
// Compute value of N*(N+1)/2*(N+2)/3* ....*(N+n-1)/n
long long count = 1;
for (int i=1; i<=n; i++)
{
count *= (N+i-1);
count /= i;
}
return count;
}
// Driver program
int main()
{
int n = 3;
cout << countNonDecreasing(n);
return 0;
}

Output:
220

Thanks to Abhishek Somani for suggesting this method.

How does this formula work?
N * (N+1)/2 * (N+2)/3 * .... * (N+n-1)/n
Where N = 10

Let us try for different values of n.

For n = 1, the value is N from formula.
Which is true as for n = 1, we have all single digit
numbers, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9.
For n = 2, the value is N(N+1)/2 from formula
We can have N numbers beginning with 0, (N-1) numbers
beginning with 1, and so on.
So sum is N + (N-1) + .... + 1 = N(N+1)/2
For n = 3, the value is N(N+1)/2(N+2)/3 from formula
We can have N(N+1)/2 numbers beginning with 0, (N-1)N/2
numbers beginning with 1 (Note that when we begin with 1,
we have N-1 digits left to consider for remaining places),
(N-2)(N-1)/2 beginning with 2, and so on.
Count = N(N+1)/2 + (N-1)N/2 + (N-2)(N-1)/2 +
(N-3)(N-2)/2 .... 3 + 1
[Combining first 2 terms, next 2 terms and so on]
= 1/2[N2 + (N-2)2 + .... 4]

= N*(N+1)*(N+2)/6 [Refer this , putting n=N/2 in the

even sum formula]

For general n digit case, we can apply Mathematical Induction. The count would be equal to count n-1 digit beginning with 0, i.e., N*(N+1)/2*
(N+2)/3* .*(N+n-1-1)/(n-1). Plus count of n-1 digit numbers beginning with 1, i.e., (N-1)*(N)/2*(N+1)/3* .*(N-1+n-1-1)/(n-1) (Note that N is
replaced by N-1) and so on.

Find length of the longest consecutive path from a given starting character
Given a matrix of characters. Find length of the longest path from a given character, such that all characters in the path are consecutive to each
other, i.e., every character in path is next to previous in alphabetical order. It is allowed to move in all 8 directions from a cell.

Example
Input: mat[][] = { {a,
{h,
{i,
Starting Point =

c, d},
b, e},
g, f}}
'e'

Output: 5
If starting point is 'e', then longest path with consecutive
characters is "e f g h i".
Input: mat[R][C] = { {b, e, f},
{h, d, a},
{i, c, a}};
Starting Point = 'b'
Output: 1
'c' is not present in all adjacent cells of 'b'

The idea is to first search given starting character in the given matrix. Do Depth First Search (DFS) from all occurrences to find all consecutive
paths. While doing DFS, we may encounter many subproblems again and again. So we use dynamic programming to store results of subproblems.
Below is C++ implementation of above idea.
// C++ program to find the longest consecutive path
#include<bits/stdc++.h>
#define R 3
#define C 3
using namespace std;
// tool matrices to recur for adjacent cells.
int x[] = {0, 1, 1, -1, 1, 0, -1, -1};
int y[] = {1, 0, 1, 1, -1, -1, 0, -1};
// dp[i][j] Stores length of longest consecutive path
// starting at arr[i][j].
int dp[R][C];
// check whether mat[i][j] is a valid cell or not.
bool isvalid(int i, int j)
{
if (i < 0 || j < 0 || i >= R || j >= C)
return false;
return true;
}
// Check whether current character is adjacent to previous
// character (character processed in parent call) or not.
bool isadjacent(char prev, char curr)
{
return ((curr - prev) == 1);
}

// i, j are the indices of the current cell and prev is the

// character processed in the parent call.. also mat[i][j]
// is our current character.
int getLenUtil(char mat[R][C], int i, int j, char prev)
{
// If this cell is not valid or current character is not
// adjacent to previous one (e.g. d is not adjacent to b )
// or if this cell is already included in the path than return 0.
if (!isvalid(i, j) || !isadjacent(prev, mat[i][j]))
return 0;
// If this subproblem is already solved , return the answer
if (dp[i][j] != -1)
return dp[i][j];
int ans = 0; // Initialize answer
// recur for paths with differnt adjacent cells and store
// the length of longest path.
for (int k=0; k<8; k++)
ans = max(ans, 1 + getLenUtil(mat, i + x[k],
j + y[k], mat[i][j]));
// save the answer and return
return dp[i][j] = ans;
}
// Returns length of the longest path with all characters consecutive
// to each other. This function first initializes dp array that
// is used to store results of subproblems, then it calls
// recursive DFS based function getLenUtil() to find max length path
int getLen(char mat[R][C], char s)
{
memset(dp, -1, sizeof dp);
int ans = 0;
for (int i=0; i<R; i++)
{
for (int j=0; j<C; j++)
{
// check for each possible starting point
if (mat[i][j] == s) {
// recur for all eight adjacent cells
for (int k=0; k<8; k++)
ans = max(ans, 1 + getLenUtil(mat,
i + x[k], j + y[k], s));
}
}
}
return ans;
}
// Driver program
int main() {
char mat[R][C] = { {'a','c','d'},
{ 'h','b','a'},
{ 'i','g','f'}};
cout << getLen(mat,
cout << getLen(mat,
cout << getLen(mat,
cout << getLen(mat,
return 0;

'a')
'e')
'b')
'f')

<<
<<
<<
<<

endl;
endl;
endl;
endl;

Output:
4
0
3
4

Thanks to Gaurav Ahirwar for above solution.

Find minimum number of coins that make a given value

Given a value V, if we want to make change for V cents, and we have infinite supply of each of C = { C1, C2, .. , Cm} valued coins, what is the
minimum number of coins to make the change?
Examples:
Input: coins[] = {25, 10, 5}, V = 30
Output: Minimum 2 coins required
We can use one coin of 25 cents and one of 5 cents
Input: coins[] = {9, 6, 5, 1}, V = 11
Output: Minimum 2 coins required
We can use one coin of 6 cents and 1 coin of 5 cents

This problem is a variation of the problem discussed Coin Change Problem. Here instead of finding total number of possible solutions, we need to
find the solution with minimum number of coins.
The minimum number of coins for a value V can be computed using below recursive formula.
If V == 0, then 0 coins required.
If V > 0
minCoin(coins[0..m-1], V) = min {1 + minCoins(V-coin[i])}
where i varies from 0 to m-1
and coin[i] <= V

Below is recursive solution based on above recursive formula.

// A Naive recursive C++ program to find minimum of coins
// to make a given change V
#include<bits/stdc++.h>
using namespace std;
// m is size of coins array (number of different coins)
int minCoins(int coins[], int m, int V)
{
// base case
if (V == 0) return 0;
// Initialize result
int res = INT_MAX;
// Try every coin that has smaller value than V
for (int i=0; i<m; i++)
{
if (coins[i] <= V)
{
int sub_res = minCoins(coins, m, V-coins[i]);
// Check for INT_MAX to avoid overflow and see if
// result can minimized
if (sub_res != INT_MAX && sub_res + 1 < res)
res = sub_res + 1;
}
}
return res;
}
// Driver program to test above function
int main()
{
int coins[] = {9, 6, 5, 1};
int m = sizeof(coins)/sizeof(coins[0]);
int V = 11;
cout << "Minimum coins required is "
<< minCoins(coins, m, V);
return 0;
}

Output:
Minimum coins required is 2

The time complexity of above solution is exponential. If we draw the complete recursion tree, we can observer that many subproblems are solved
again and again. For example, when we start from V = 11, we can reach 6 by subtracting one 5 times and by subtracting 5 one times. So the
subproblem for 6 is called twice.
Since same suproblems are called again, this problem has Overlapping Subprolems property. So the min coins problem has both properties (see

this and this) of a dynamic programming problem. Like other typical Dynamic Programming(DP) problems, recomputations of same subproblems
can be avoided by constructing a temporary array table[][] in bottom up manner. Below is Dynamic Programming based solution.
// A Dynamic Programming based C++ program to find minimum of coins
// to make a given change V
#include<bits/stdc++.h>
using namespace std;
// m is size of coins array (number of different coins)
int minCoins(int coins[], int m, int V)
{
// table[i] will be storing the minimum number of coins
// required for i value. So table[V] will have result
int table[V+1];
// Base case (If given value V is 0)
table[0] = 0;
// Initialize all table values as Infinite
for (int i=1; i<=V; i++)
table[i] = INT_MAX;
// Compute minimum coins required for all
// values from 1 to V
for (int i=1; i<=V; i++)
{
// Go through all coins smaller than i
for (int j=0; j<m; j++)
if (coins[j] <= i)
{
int sub_res = table[i-coins[j]];
if (sub_res != INT_MAX && sub_res + 1 < table[i])
table[i] = sub_res + 1;
}
}
return table[V];
}
// Driver program to test above function
int main()
{
int coins[] = {9, 6, 5, 1};
int m = sizeof(coins)/sizeof(coins[0]);
int V = 11;
cout << "Minimum coins required is "
<< minCoins(coins, m, V);
return 0;
}

Output:
Minimum coins required is 2

Time complexity of the above solution is O(mV).

Thanks to Goku for suggesting above solution in a comment here and thanks to Vignesh Mohan for suggesting this problem and initial solution.

Collect maximum points in a grid using two traversals

Given a matrix where every cell represents points. How to collect maximum points using two traversals under following conditions?
Let the dimensions of given grid be R x C.
1) The first traversal starts from top left corner, i.e., (0, 0) and should reach left bottom corner, i.e., (R-1, 0). The second traversal starts from top
right corner, i.e., (0, C-1) and should reach bottom right corner, i.e., (R-1, C-1)/
2) From a point (i, j), we can move to (i+1, j+1) or (i+1, j-1) or (i+1, j)
3) A traversal gets all points of a particular cell through which it passes. If one traversal has already collected points of a cell, then the other
traversal gets no points if goes through that cell again.
Input :
int arr[R][C] = {{3,
{5,
{1,
{1,
{1,
};

6,
2,
1,
1,
1,

8, 2},
4, 3},
20, 10},
20, 10},
20, 10},

Output: 73
Explanation :

First traversal collects total points of value 3 + 2 + 20 + 1 + 1 = 27

Second traversal collects total points of value 2 + 4 + 10 + 20 + 10 = 46.
Total Points collected = 27 + 46 = 73.

Source: http://qa.geeksforgeeks.org/1485/running-through-the-grid-to-get-maximum-nutritional-value
Both traversals always move forward along x
Base Cases:
// If destinations reached
if (x == R-1 && y1 == 0 && y2 == C-1)
maxPoints(arr, x, y1, y2) = arr[x][y1] + arr[x][y2];
// If any of the two locations is invalid (going out of grid)
if input is not valid
maxPoints(arr, x, y1, y2) = -INF (minus infinite)
// If both traversals are at same cell, then we count the value of cell
// only once.
If y1 and y2 are same
result = arr[x][y1]
Else
result = arr[x][y1] + arr[x][y2]
result += max { // Max of 9 cases
maxPoints(arr, x+1,
maxPoints(arr, x+1,
maxPoints(arr, x+1,
maxPoints(arr, x+1,
maxPoints(arr, x+1,
maxPoints(arr, x+1,
maxPoints(arr, x+1,
maxPoints(arr, x+1,
maxPoints(arr, x+1,
}

y1+1, y2),
y1+1, y2+1),
y1+1, y2-1),
y1-1, y2),
y1-1, y2+1),
y1-1, y2-1),
y1, y2),
y1, y2+1),
y1, y2-1)

The above recursive solution has many subproblems that are solved again and again. Therefore, we can use Dynamic Programming to solve the
above problem more efficiently. Below is memoization (Memoization is alternative to table based iterative solution in Dynamic Programming) based
implementation. In below implementation, we use a memoization table mem to keep track of already solved problems.
// A Memoization based program to find maximum collection
// using two traversals of a grid

#include<bits/stdc++.h>
using namespace std;
#define R 5
#define C 4
// checks whether a given input is valid or not
bool isValid(int x, int y1, int y2)
{
return (x >= 0 && x < R && y1 >=0 &&
y1 < C && y2 >=0 && y2 < C);
}
// Driver function to collect max value
int getMaxUtil(int arr[R][C], int mem[R][C][C], int x, int y1, int y2)
{
/*---------- BASE CASES -----------*/
// if P1 or P2 is at an invalid cell
if (!isValid(x, y1, y2)) return INT_MIN;
// if both traversals reach their destinations
if (x == R-1 && y1 == 0 && y2 == C-1)
return arr[x][y1] + arr[x][y2];
// If both traversals are at last row but not at their destination
if (x == R-1) return INT_MIN;
// If subproblem is already solved
if (mem[x][y1][y2] != -1) return mem[x][y1][y2];
// Initialize answer for this subproblem
int ans = INT_MIN;
// this variable is used to store gain of current cell(s)
int temp = (y1 == y2)? arr[x][y1]: arr[x][y1] + arr[x][y2];
/* Recur for all possible cases, then
one with max value */
ans = max(ans, temp + getMaxUtil(arr,
ans = max(ans, temp + getMaxUtil(arr,
ans = max(ans, temp + getMaxUtil(arr,

store and return the

mem, x+1, y1, y2-1));
mem, x+1, y1, y2+1));
mem, x+1, y1, y2));

ans = max(ans, temp + getMaxUtil(arr, mem, x+1, y1-1, y2));

ans = max(ans, temp + getMaxUtil(arr, mem, x+1, y1-1, y2-1));
ans = max(ans, temp + getMaxUtil(arr, mem, x+1, y1-1, y2+1));
ans = max(ans, temp + getMaxUtil(arr, mem, x+1, y1+1, y2));
ans = max(ans, temp + getMaxUtil(arr, mem, x+1, y1+1, y2-1));
ans = max(ans, temp + getMaxUtil(arr, mem, x+1, y1+1, y2+1));
return (mem[x][y1][y2] = ans);
}
// This is mainly a wrapper over recursive function getMaxUtil().
// This function creates a table for memoization and calls
// getMaxUtil()
int geMaxCollection(int arr[R][C])
{
// Create a memoization table and initialize all entries as -1
int mem[R][C][C];
memset(mem, -1, sizeof(mem));
// Calculation maximum value using memoization based function
// getMaxUtil()
return getMaxUtil(arr, mem, 0, 0, C-1);
}
// Driver program to test above functions
int main()
{
int arr[R][C] = {{3, 6, 8, 2},
{5, 2, 4, 3},
{1, 1, 20, 10},
{1, 1, 20, 10},
{1, 1, 20, 10},
};
cout << "Maximum collection is " << geMaxCollection(arr);
return 0;
}

Output:

Maximum collection is 73

Thanks to Gaurav Ahirwar for suggesting above problem and solution here.

Shortest Common Supersequence

Given two strings str1 and str2, find the shortest string that has both str1 and str2 as subsequences.
Examples:
Input: str1 = "geek", str2 = "eke"
Output: "geeke"
Input: str1 = "AGGTAB", str2 = "GXTXAYB"
Output: "AGXGTXAYB"

This problem is closely related to longest common subsequence problem. Below are steps.
1) Find Longest Common Subsequence (lcs) of two given strings. For example, lcs of geek and eke is ek.
2) Insert non-lcs characters (in their original order in strings) to the lcs found above, and return the result. So ek becomes geeke which is shortest
common supersequence.
Let us consider another example, str1 = AGGTAB and str2 = GXTXAYB. LCS of str1 and str2 is GTAB. Once we find LCS, we insert
characters of both strings in order and we get AGXGTXAYB
How does this work?
We need to find a string that has both strings as subsequences and is shortest such string. If both strings have all characters different, then result is
sum of lengths of two given strings. If there are common characters, then we dont want them multiple times as the task is to minimize length.
Therefore, we fist find the longest common subsequence, take one occurrence of this subsequence and add extra characters.
Length of the shortest supersequence = (Sum of lengths of given two strings) (Length of LCS of two given strings)

Below is C implementation of above idea. The below implementation only finds length of the shortest supersequence.
/* C program to find length of the shortest supersequence */
#include<stdio.h>
#include<string.h>
/* Utility function to get max of 2 integers */
int max(int a, int b) { return (a > b)? a : b; }
/* Returns length of LCS for X[0..m-1], Y[0..n-1] */
int lcs( char *X, char *Y, int m, int n);
// Function to find length of the shortest supersequence
// of X and Y.
int shortestSuperSequence(char *X, char *Y)
{
int m = strlen(X), n = strlen(Y);
int l = lcs(X, Y, m, n); // find lcs
// Result is sum of input string lengths - length of lcs
return (m + n - l);
}
/* Returns length of LCS for X[0..m-1], Y[0..n-1] */
int lcs( char *X, char *Y, int m, int n)
{
int L[m+1][n+1];
int i, j;
/* Following steps build L[m+1][n+1] in bottom up fashion.
Note that L[i][j] contains length of LCS of X[0..i-1]
and Y[0..j-1] */
for (i=0; i<=m; i++)
{
for (j=0; j<=n; j++)
{
if (i == 0 || j == 0)
L[i][j] = 0;
else if (X[i-1] == Y[j-1])
L[i][j] = L[i-1][j-1] + 1;
else
L[i][j] = max(L[i-1][j], L[i][j-1]);
}
}

/* L[m][n] contains length of LCS for X[0..n-1] and

Y[0..m-1] */
return L[m][n];
}
/* Driver program to test above function */
int main()
{
char X[] = "AGGTAB";
char Y[] = "GXTXAYB";
printf("Length of the shortest supersequence is %d\n",
shortestSuperSequence(X, Y));
return 0;
}

Output:
Length of the shortest supersequence is 9

Below is Another Method to solve the above problem.

A simple analysis yields below simple recursive solution.
Let X[0..m-1] and Y[0..n-1] be two strings and m and be respective
lengths.
if (m == 0) return n;
if (n == 0) return m;
// If last characters are same, then add 1 to result and
// recur for X[]
if (X[m-1] == Y[n-1])
return 1 + SCS(X, Y, m-1, n-1);
// Else find shortest of following two
// a) Remove last character from X and recur
// b) Remove last character from Y and recur
else return 1 + min( SCS(X, Y, m-1, n), SCS(X, Y, m, n-1) );

Below is simple naive recursive solution based on above recursive formula.

/* A Naive recursive C++ program to find length
of the shortest supersequence */
#include<bits/stdc++.h>
using namespace std;
int superSeq(char* X, char* Y, int m, int n)
{
if (!m) return n;
if (!n) return m;
if (X[m-1] == Y[n-1])
return 1 + superSeq(X, Y, m-1, n-1);
return 1 + min(superSeq(X, Y, m-1, n),
superSeq(X, Y, m, n-1));
}
// Driver program to test above function
int main()
{
char X[] = "AGGTAB";
char Y[] = "GXTXAYB";
cout << "Length of the shortest supersequence is "
<< superSeq(X, Y, strlen(X), strlen(Y));
return 0;
}

Output:
Length of the shortest supersequence is 9

Time complexity of the above solution exponential O(2min(m, n)). Since there are overlapping subproblems, we can efficiently solve this recursive
problem using Dynamic Programming. Below is Dynamic Programming based implementation. Time complexity of this solution is O(mn).
/* A dynamic programming based C program to find length
of the shortest supersequence */

#include<bits/stdc++.h>
using namespace std;
// Returns length of the shortest supersequence of X and Y
int superSeq(char* X, char* Y, int m, int n)
{
int dp[m+1][n+1];
// Fill table in bottom up manner
for (int i = 0; i <= m; i++)
{
for (int j = 0; j <= n; j++)
{
// Below steps follow above recurrence
if (!i)
dp[i][j] = j;
else if (!j)
dp[i][j] = i;
else if (X[i-1] == Y[j-1])
dp[i][j] = 1 + dp[i-1][j-1];
else
dp[i][j] = 1 + min(dp[i-1][j], dp[i][j-1]);
}
}
return dp[m][n];
}
// Driver program to test above function
int main()
{
char X[] = "AGGTAB";
char Y[] = "GXTXAYB";
cout << "Length of the shortest supersequence is "
<< superSeq(X, Y, strlen(X), strlen(Y));
return 0;
}

Output:
Length of the shortest supersequence is 9

Thanks to Gaurav Ahirwar for suggesting this solution.

Exercise:
Extend the above program to print shortest supersequence also using function to print LCS.
References:
https://en.wikipedia.org/wiki/Shortest_common_supersequence

Compute sum of digits in all numbers from 1 to n

Given a number x, find sum of digits in all numbers from 1 to n.
Examples:
Input: n = 5
Output: Sum of digits in numbers from 1 to 5 = 15
Input: n = 12
Output: Sum of digits in numbers from 1 to 12 = 51
Input: n = 328
Output: Sum of digits in numbers from 1 to 328 = 3241

Naive Solution:
A naive solution is to go through every number x from 1 to n, and compute sum in x by traversing all digits of x. Below is C++ implementation of
this idea.
// A Simple C++ program to compute sum of digits in numbers from 1 to n
#include<iostream>
using namespace std;
int sumOfDigits(int );
// Returns sum of all digits in numbers from 1 to n
int sumOfDigitsFrom1ToN(int n)
{
int result = 0; // initialize result
// One by one compute sum of digits in every number from
// 1 to n
for (int x=1; x<=n; x++)
result += sumOfDigits(x);
return result;
}
// A utility function to compute sum of digits in a
// given number x
int sumOfDigits(int x)
{
int sum = 0;
while (x != 0)
{
sum += x %10;
x = x /10;
}
return sum;
}
// Driver Program
int main()
{
int n = 328;
cout << "Sum of digits in numbers from 1 to " << n << " is "
<< sumOfDigitsFrom1ToN(n);
return 0;
}

Output
Sum of digits in numbers from 1 to 328 is 3241

Efficient Solution:
Above is a naive solution. We can do it more efficiently by finding a pattern.
Let us take few examples.
sum(9) = 1 + 2 + 3 + 4 ........... + 9
= 9*10/2
= 45
sum(99) =
=
=
=

45 + (10 + 45) + (20 + 45) + ..... (90 + 45)

45*10 + (10 + 20 + 30 ... 90)
45*10 + 10(1 + 2 + ... 9)
45*10 + 45*10

= sum(9)*10 + 45*10
sum(999) = sum(99)*10 + 45*100

In general, we can compute sum(10d 1) using below formula

sum(10d - 1) = sum(10d-1 - 1) * 10 + 45*(10d-1)

In below implementation, the above formula is implemented using dynamic programming as there are overlapping subproblems.
The above formula is one core step of the idea. Below is complete algorithm
Algorithm: sum(n)
1) Find number of digits minus one in n. Let this value be 'd'.
For 328, d is 2.
2) Compute some of digits in numbers from 1 to 10d - 1.
Let this sum be w. For 328, we compute sum of digits from 1 to
99 using above formula.
3) Find Most significant digit (msd) in n. For 328, msd is 3.
4) Overall sum is sum of following terms
a) Sum of digits in 1 to "msd * 10d - 1". For 328, sum of
digits in numbers from 1 to 299.
For 328, we compute 3*sum(99) + (1 + 2)*100. Note that sum of
sum(299) is sum(99) + sum of digits from 100 to 199 + sum of digits
from 200 to 299.
Sum of 100 to 199 is sum(99) + 1*100 and sum of 299 is sum(99) + 2*100.
In general, this sum can be computed as w*msd + (msd*(msd-1)/2)*10d
b) Sum of digits in msd * 10d to n. For 328, sum of digits in
300 to 328.
For 328, this sum is computed as 3*29 + recursive call "sum(28)"
In general, this sum can be computed as msd * (n % (msd*10d) + 1)
+ sum(n % (10d))

Below is C++ implementation of above aglorithm.

// C++ program to compute sum of digits in numbers from 1 to n
#include<bits/stdc++.h>
using namespace std;
// Function to computer sum of digits in numbers from 1 to n
// Comments use example of 328 to explain the code
int sumOfDigitsFrom1ToN(int n)
{
// base case: if n<10 return sum of
// first n natural numbers
if (n<10)
return n*(n+1)/2;
// d = number of digits minus one in n. For 328, d is 2
int d = log10(n);
// computing sum of digits from 1 to 10^d-1,
// d=1 a[0]=0;
// d=2 a[1]=sum of digit from 1 to 9 = 45
// d=3 a[2]=sum of digit from 1 to 99 = a[1]*10 + 45*10^1 = 900
// d=4 a[3]=sum of digit from 1 to 999 = a[2]*10 + 45*10^2 = 13500
int *a = new int[d+1];
a[0] = 0, a[1] = 45;
for (int i=2; i<=d; i++)
a[i] = a[i-1]*10 + 45*ceil(pow(10,i-1));
// computing 10^d
int p = ceil(pow(10, d));
// Most significant digit (msd) of n,
// For 328, msd is 3 which can be obtained using 328/100
int msd = n/p;
//
//
//
//
//

EXPLANATION FOR FIRST and SECOND TERMS IN BELOW LINE OF CODE

First two terms compute sum of digits from 1 to 299
(sum of digits in range 1-99 stored in a[d]) +
(sum of digits in range 100-199, can be calculated as 1*100 + a[d]
(sum of digits in range 200-299, can be calculated as 2*100 + a[d]

// The above sum can be written as 3*a[d] + (1+2)*100

// EXPLANATION FOR THIRD AND FOURTH TERMS IN BELOW LINE OF CODE
// The last two terms compute sum of digits in number from 300 to 328
// The third term adds 3*29 to sum as digit 3 occurs in all numbers
//
from 300 to 328
// The fourth term recursively calls for 28
return msd*a[d] + (msd*(msd-1)/2)*p +
msd*(1+n%p) + sumOfDigitsFrom1ToN(n%p);
}
// Driver Program
int main()
{
int n = 328;
cout << "Sum of digits in numbers from 1 to " << n << " is "
<< sumOfDigitsFrom1ToN(n);
return 0;
}

Output
Sum of digits in numbers from 1 to 328 is 3241

The efficient algorithm has one more advantage that we need to compute the array a[] only once even when we are given multiple inputs.

Count possible ways to construct buildings

Given an input number of sections and each section has 2 plots on either sides of the road. Find all possible ways to construct buildings in the plots
such that there is a space between any 2 buildings.
Example:
N = 1
Output = 4
Place a building
Place a building
Do not place any
Place a building

on one side.
on other side
building.
on both sides.

N = 3
Output = 25
3 sections, which means possible ways for one side are
BSS, BSB, SSS, SBS, SSB where B represents a building
and S represents an empty space
Total possible ways are 25, because a way to place on
one side can correspond to any of 5 ways on other side.
N = 4
Output = 64

We can simplify the problem to first calculate for one side only. If we know the result for one side, we can always do square of the result and get
result for two sides.
A new building can be placed on a section if section just before it has space. A space can be placed anywhere (it doesnt matter whether the
previous section has a building or not).
Let countB(i) be count of
ending with
countS(i) be count of
ending with

possible ways with i sections

a building.
possible ways with i sections
a space.

// A space can be added after a building or after a space.

countS(N) = countB(N-1) + countS(N-1)
// A building can only be added after a space.
countB[N] = countS(N-1)
// Result for one side is sum of the above two counts.
result1(N) = countS(N) + countB(N)
// Result for two sides is square of result1(N)
result2(N) = result1(N) * result1(N)

Below is C++ implementation of above idea.

// C++ program to count all possible way to construct buildings
#include<iostream>
using namespace std;
// Returns count of possible ways for N sections
int countWays(int N)
{
// Base case
if (N == 1)
return 4; // 2 for one side and 4 for two sides
//
//
//
//

countB is count of ways with a building at the end

countS is count of ways with a space at the end
prev_countB and prev_countS are previous values of
countB and countS respectively.

// Initialize countB and countS for one side

int countB=1, countS=1, prev_countB, prev_countS;
// Use the above recursive formula for calculating
// countB and countS using previous values
for (int i=2; i<=N; i++)
{
prev_countB = countB;
prev_countS = countS;
countS = prev_countB + prev_countS;
countB = prev_countS;

}
// Result for one side is sum of ways ending with building
// and ending with space
int result = countS + countB;
// Result for 2 sides is square of result for one side
return (result*result);
}
// Driver program
int main()
{
int N = 3;
cout << "Count of ways for " << N
<< " sections is " << countWays(N);
return 0;
}

Output:
25

Time complexity: O(N)

Auxiliary Space: O(1)
Algorithmic Paradigm: Dynamic Programming
Optimized Solution:
Note that the above solution can be further optimized. If we take closer look at the results, for different values, we can notice that the results for
two sides are squares of Fibonacci Numbers.
N = 1, result = 4 [result for one side = 2]
N = 2, result = 9 [result for one side = 3]
N = 3, result = 25 [result for one side = 5]
N = 4, result = 64 [result for one side = 8]
N = 5, result = 169 [result for one side = 13]
.
.
In general, we can say
result(N) = fib(N+2)2
fib(N) is a function that returns N'th
Fibonacci Number.

Therefore, we can use O(LogN) implementation of Fibonacci Numbers to find number of ways in O(logN) time.

Maximum profit by buying and selling a share at most twice

In a daily share trading, a buyer buys shares in the morning and sells it on same day. If the trader is allowed to make at most 2 transactions in a
day, where as second transaction can only start after first one is complete (Sell->buy->sell->buy). Given stock prices throughout day, find out
maximum profit that a share trader could have made.
Examples:
Input: price[] = {10, 22, 5, 75, 65, 80}
Output: 87
Trader earns 87 as sum of 12 and 75
Buy at price 10, sell at 22, buy at 5 and sell at 80
Input: price[] = {2, 30, 15, 10, 8, 25, 80}
Output: 100
Trader earns 100 as sum of 28 and 72
Buy at price 2, sell at 30, buy at 8 and sell at 80
Input: price[] = {100, 30, 15, 10, 8, 25, 80};
Output: 72
Buy at price 8 and sell at 80.
Input: price[] = {90, 80, 70, 60, 50}
Output: 0
Not possible to earn.

A Simple Solution is to to consider every index i and do following

Max profit with at most two transactions =
MAX {max profit with one transaction and subarray price[0..i] +
max profit with one transaction and aubarray price[i+1..n-1] }
i varies from 0 to n-1.

Maximum possible using one transaction can be calculated using following O(n) algorithm
Maximum difference between two elements such that larger element appears after the smaller number
Time complexity of above simple solution is O(n2).
We can do this O(n) using following Efficient Solution. The idea is to store maximum possible profit of every subarray and solve the problem in
following two phases.
1) Create a table profit[0..n-1] and initialize all values in it 0.
2) Traverse price[] from right to left and update profit[i] such that profit[i] stores maximum profit achievable from one transaction in subarray
price[i..n-1]
3) Traverse price[] from left to right and update profit[i] such that profit[i] stores maximum profit such that profit[i] contains maximum achievable
profit from two transactions in subarray price[0..i].
4) Return profit[n-1]
To do step 1, we need to keep track of maximum price from right to left side and to do step 2, we need to keep track of minimum price from left
to right. Why we traverse in reverse directions? The idea is to save space, in second step, we use same array for both purposes, maximum with 1
transaction and maximum with 2 transactions. After an iteration i, the array profit[0..i] contains maximum profit with 2 transactions and
profit[i+1..n-1] contains profit with two transactions.
Below are implementations of above idea.

C++
// C++ program to find maximum possible profit with at most
// two transactions
#include<iostream>
using namespace std;
// Returns maximum profit with two transactions on a given
// list of stock prices, price[0..n-1]
int maxProfit(int price[], int n)
{
// Create profit array and initialize it as 0
int *profit = new int[n];
for (int i=0; i<n; i++)
profit[i] = 0;

/* Get the maximum profit with only one transaction

allowed. After this loop, profit[i] contains maximum
profit from price[i..n-1] using at most one trans. */
int max_price = price[n-1];
for (int i=n-2;i>=0;i--)
{
// max_price has maximum of price[i..n-1]
if (price[i] > max_price)
max_price = price[i];
// we can get profit[i] by taking maximum of:
// a) previous maximum, i.e., profit[i+1]
// b) profit by buying at price[i] and selling at
//
max_price
profit[i] = max(profit[i+1], max_price-price[i]);
}
/* Get the maximum profit with two transactions allowed
After this loop, profit[n-1] contains the result */
int min_price = price[0];
for (int i=1; i<n; i++)
{
// min_price is minimum price in price[0..i]
if (price[i] < min_price)
min_price = price[i];
// Maximum profit is maximum of:
// a) previous maximum, i.e., profit[i-1]
// b) (Buy, Sell) at (min_price, price[i]) and add
//
profit of other trans. stored in profit[i]
profit[i] = max(profit[i-1], profit[i] +
(price[i]-min_price) );
}
int result = profit[n-1];
delete [] profit; // To avoid memory leak
return result;
}
// Drive program
int main()
{
int price[] = {2, 30, 15, 10, 8, 25, 80};
int n = sizeof(price)/sizeof(price[0]);
cout << "Maximum Profit = " << maxProfit(price, n);
return 0;
}

Python
# Returns maximum profit with two transactions on a given
# list of stock prices price[0..n-1]
def maxProfit(price,n):
# Create profit array and initialize it as 0
profit = [0]*n
# Get the maximum profit with only one transaction
# allowed. After this loop, profit[i] contains maximum
# profit from price[i..n-1] using at most one trans.
max_price=price[n-1]
for i in range( n-2, 0 ,-1):
if price[i]> max_price:
max_price = price[i]
# we can get profit[i] by taking maximum of:
# a) previous maximum, i.e., profit[i+1]
# b) profit by buying at price[i] and selling at
#
max_price
profit[i] = max(profit[i+1], max_price - price[i])
# Get the maximum profit with two transactions allowed
# After this loop, profit[n-1] contains the result
min_price=price[0]
for i in range(1,n):

if price[i] < min_price:

min_price = price[i]
# Maximum profit is maximum of:
# a) previous maximum, i.e., profit[i-1]
# b) (Buy, Sell) at (min_price, A[i]) and add
#
profit of other trans. stored in profit[i]
profit[i] = max(profit[i-1], profit[i]+(price[i]-min_price))
result = profit[n-1]
return result
# Driver function
price = [2, 30, 15, 10, 8, 25, 80]
print "Maximum profit is", maxProfit(price, len(price))
# This code is contributed by __Devesh Agrawal__

Maximum Profit = 100

Time complexity of the above solution is O(n).

Algorithmic Paradigm: Dynamic Programming

How to print maximum number of As using given four keys

This is a famous interview question asked in Google, Paytm and many other company interviews.
Below is the problem statement.
Imagine you have a special keyboard with the following keys:
Key 1: Prints 'A' on screen
Key 2: (Ctrl-A): Select screen
Key 3: (Ctrl-C): Copy selection to buffer
Key 4: (Ctrl-V): Print buffer on screen appending it
after what has already been printed.
If you can only press the keyboard for N times (with the above four
keys), write a program to produce maximum numbers of A's. That is to
say, the input parameter is N (No. of keys that you can press), the
output is M (No. of As that you can produce).

Examples:
Input: N = 3
Output: 3
We can at most get 3 A's on screen by pressing
following key sequence.
A, A, A
Input: N = 7
Output: 9
We can at most get 9 A's on screen by pressing
following key sequence.
A, A, A, Ctrl A, Ctrl C, Ctrl V, Ctrl V
Input: N = 11
Output: 27
We can at most get 27 A's on screen by pressing
following key sequence.
A, A, A, Ctrl A, Ctrl C, Ctrl V, Ctrl V, Ctrl A,
Ctrl C, Ctrl V, Ctrl V

Below are few important points to note.

a) For N < 7, the output is N itself. b) Ctrl V can be used multiple times to print current buffer (See last two examples above). The idea is to
compute the optimal string length for N keystrokes by using a simple insight. The sequence of N keystrokes which produces an optimal string
length will end with a suffix of Ctrl-A, a Ctrl-C, followed by only Ctrl-V's (For N > 6).
The task is to find out the break=point after which we get the above suffix of keystrokes. Definition of a breakpoint is that instance after which we
need to only press Ctrl-A, Ctrl-C once and the only Ctrl-Vs afterwards to generate the optimal length. If we loop from N-3 to 1 and choose each
of these values for the break-point, and compute that optimal string they would produce. Once the loop ends, we will have the maximum of the
optimal lengths for various breakpoints, thereby giving us the optimal length for N keystrokes.
Below is C implementation based on above idea.
/* A recursive C program to print maximum number of A's using
following four keys */
#include<stdio.h>
// A recursive function that returns the optimal length string
// for N keystrokes
int findoptimal(int N)
{
// The optimal string length is N when N is smaller than 7
if (N <= 6)
return N;
// Initialize result
int max = 0;
// TRY ALL POSSIBLE BREAK-POINTS
// For any keystroke N, we need to loop from N-3 keystrokes
// back to 1 keystroke to find a breakpoint 'b' after which we
// will have Ctrl-A, Ctrl-C and then only Ctrl-V all the way.
int b;
for (b=N-3; b>=1; b--)
{
// If the breakpoint is s at b'th keystroke then
// the optimal string would have length
// (n-b-1)*screen[b-1];
int curr = (N-b-1)*findoptimal(b);
if (curr > max)

max = curr;
}
return max;
}
// Driver program
int main()
{
int N;
// for the rest of the array we will rely on the previous
// entries to compute new ones
for (N=1; N<=20; N++)
printf("Maximum Number of A's with %d keystrokes is %d\n",
N, findoptimal(N));
}

Output:
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum

Number
Number
Number
Number
Number
Number
Number
Number
Number
Number
Number
Number
Number
Number
Number
Number
Number
Number
Number
Number

of
of
of
of
of
of
of
of
of
of
of
of
of
of
of
of
of
of
of
of

A's
A's
A's
A's
A's
A's
A's
A's
A's
A's
A's
A's
A's
A's
A's
A's
A's
A's
A's
A's

with
with
with
with
with
with
with
with
with
with
with
with
with
with
with
with
with
with
with
with

1 keystrokes is 1
2 keystrokes is 2
3 keystrokes is 3
4 keystrokes is 4
5 keystrokes is 5
6 keystrokes is 6
7 keystrokes is 9
8 keystrokes is 12
9 keystrokes is 16
10 keystrokes is 20
11 keystrokes is 27
12 keystrokes is 36
13 keystrokes is 48
14 keystrokes is 64
15 keystrokes is 81
16 keystrokes is 108
17 keystrokes is 144
18 keystrokes is 192
19 keystrokes is 256
20 keystrokes is 324

The above function computes the same subproblems again and again. Recomputations of same subproblems can be avoided by storing the
solutions to subproblems and solving problems in bottom up manner.
Below is Dynamic Programming based C implementation where an auxiliary array screen[N] is used to store result of subproblems.
/* A Dynamic Programming based C program to find maximum number of A's
that can be printed using four keys */
#include<stdio.h>
// this function returns the optimal length string for N keystrokes
int findoptimal(int N)
{
// The optimal string length is N when N is smaller than 7
if (N <= 6)
return N;
// An array to store result of subproblems
int screen[N];
int b; // To pick a breakpoint
// Initializing the optimal lengths array for uptil 6 input
// strokes.
int n;
for (n=1; n<=6; n++)
screen[n-1] = n;
// Solve all subproblems in bottom manner
for (n=7; n<=N; n++)
{
// Initialize length of optimal string for n keystrokes
screen[n-1] = 0;
// For any keystroke n, we need to loop from n-3 keystrokes
// back to 1 keystroke to find a breakpoint 'b' after which we
// will have ctrl-a, ctrl-c and then only ctrl-v all the way.
for (b=n-3; b>=1; b--)
{
// if the breakpoint is at b'th keystroke then
// the optimal string would have length

// (n-b-1)*screen[b-1];
int curr = (n-b-1)*screen[b-1];
if (curr > screen[n-1])
screen[n-1] = curr;
}
}
return screen[N-1];
}
// Driver program
int main()
{
int N;
// for the rest of the array we will rely on the previous
// entries to compute new ones
for (N=1; N<=20; N++)
printf("Maximum Number of A's with %d keystrokes is %d\n",
N, findoptimal(N));
}

Output:
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum
Maximum

Number
Number
Number
Number
Number
Number
Number
Number
Number
Number
Number
Number
Number
Number
Number
Number
Number
Number
Number
Number

of
of
of
of
of
of
of
of
of
of
of
of
of
of
of
of
of
of
of
of

A's
A's
A's
A's
A's
A's
A's
A's
A's
A's
A's
A's
A's
A's
A's
A's
A's
A's
A's
A's

with
with
with
with
with
with
with
with
with
with
with
with
with
with
with
with
with
with
with
with

1 keystrokes is 1
2 keystrokes is 2
3 keystrokes is 3
4 keystrokes is 4
5 keystrokes is 5
6 keystrokes is 6
7 keystrokes is 9
8 keystrokes is 12
9 keystrokes is 16
10 keystrokes is 20
11 keystrokes is 27
12 keystrokes is 36
13 keystrokes is 48
14 keystrokes is 64
15 keystrokes is 81
16 keystrokes is 108
17 keystrokes is 144
18 keystrokes is 192
19 keystrokes is 256
20 keystrokes is 324

Thanks to Gaurav Saxena for providing the above approach to solve this problem.

Find the minimum cost to reach destination using a train

There are N stations on route of a train. The train goes from station 0 to N-1. The ticket cost for all pair of stations (i, j) is given where j is greater
than i. Find the minimum cost to reach the destination.
Consider the following example:
Input:
cost[N][N] = { {0, 15, 80, 90},
{INF, 0, 40, 50},
{INF, INF, 0, 70},
{INF, INF, INF, 0}
};
There are 4 stations and cost[i][j] indicates cost to reach j
from i. The entries where j < i are meaningless.
Output:
The minimum cost is 65
The minimum cost can be obtained by first going to station 1
from 0. Then from station 1 to station 3.

The minimum cost to reach N-1 from 0 can be recursively written as following:
minCost(0, N-1) = MIN { cost[0][n-1],
cost[0][1] + minCost(1, N-1),
minCost(0, 2) + minCost(2, N-1),
........,
minCost(0, N-2) + cost[N-2][n-1] }

The following is C++ implementation of above recursive formula.

// A naive recursive solution to find min cost path from station 0
// to station N-1
#include<iostream>
#include<climits>
using namespace std;
// infinite value
#define INF INT_MAX
// Number of stations
#define N 4
// A recursive function to find the shortest path from
// source 's' to destination 'd'.
int minCostRec(int cost[][N], int s, int d)
{
// If source is same as destination
// or destination is next to source
if (s == d || s+1 == d)
return cost[s][d];
// Initialize min cost as direct ticket from
// source 's' to destination 'd'.
int min = cost[s][d];
// Try every intermediate vertex to find minimum
for (int i = s+1; i<d; i++)
{
int c = minCostRec(cost, s, i) +
minCostRec(cost, i, d);
if (c < min)
min = c;
}
return min;
}
// This function returns the smallest possible cost to
// reach station N-1 from station 0. This function mainly
// uses minCostRec().
int minCost(int cost[][N])
{
return minCostRec(cost, 0, N-1);
}
// Driver program to test above function
int main()
{
int cost[N][N] = { {0, 15, 80, 90},

{INF, 0, 40, 50},

{INF, INF, 0, 70},
{INF, INF, INF, 0}
};
cout << "The Minimum cost to reach station "
<< N << " is " << minCost(cost);
return 0;
}

Output:
The Minimum cost to reach station 4 is 65

Time complexity of the above implementation is exponential as it tries every possible path from 0 to N-1. The above solution solves same
subrpoblems multiple times (it can be seen by drawing recursion tree for minCostPathRec(0, 5).
Since this problem has both properties of dynamic programming problems ((see this and this). Like other typical Dynamic Programming(DP)
problems, re-computations of same subproblems can be avoided by storing the solutions to subproblems and solving problems in bottom up
manner.
One dynamic programming solution is to create a 2D table and fill the table using above given recursive formula. The extra space required in this
solution would be O(N2) and time complexity would be O(N3)
We can solve this problem using O(N) extra space and O(N2) time. The idea is based on the fact that given input matrix is a Directed Acyclic
Graph (DAG). The shortest path in DAG can be calculated using the approach discussed in below post.
Shortest Path in Directed Acyclic Graph
We need to do less work here compared to above mentioned post as we know topological sorting of the graph. The topological sorting of vertices
here is 0, 1, ..., N-1. Following is the idea once topological sorting is known.
The idea in below code is to first calculate min cost for station 1, then for station 2, and so on. These costs are stored in an array dist[0...N-1].
1) The min cost for station 0 is 0, i.e., dist[0] = 0
2) The min cost for station 1 is cost[0][1], i.e., dist[1] = cost[0][1]
3) The min cost for station 2 is minimum of following two.
a) dist[0] + cost[0][2]
b) dist[1] + cost[1][2]
3) The min cost for station 3 is minimum of following three.
a) dist[0] + cost[0][3]
b) dist[1] + cost[1][3]
c) dist[2] + cost[2][3]
Similarly, dist[4], dist[5], ... dist[N-1] are calculated.
Below is C++ implementation of above idea.
// A Dynamic Programming based solution to find min cost
// to reach station N-1 from station 0.
#include<iostream>
#include<climits>
using namespace std;
#define INF INT_MAX
#define N 4
// This function returns the smallest possible cost to
// reach station N-1 from station 0.
int minCost(int cost[][N])
{
// dist[i] stores minimum cost to reach station i
// from station 0.
int dist[N];
for (int i=0; i<N; i++)
dist[i] = INF;
dist[0] = 0;
// Go through every station and check if using it
// as an intermediate station gives better path
for (int i=0; i<N; i++)
for (int j=i+1; j<N; j++)
if (dist[j] > dist[i] + cost[i][j])
dist[j] = dist[i] + cost[i][j];

return dist[N-1];
}
// Driver program to test above function
int main()
{
int cost[N][N] = { {0, 15, 80, 90},
{INF, 0, 40, 50},
{INF, INF, 0, 70},
{INF, INF, INF, 0}
};
cout << "The Minimum cost to reach station "
<< N << " is " << minCost(cost);
return 0;
}

Output:
The Minimum cost to reach station 4 is 65

Vertex Cover Problem | Set 2 (Dynamic Programming Solution for Tree)

A vertex cover of an undirected graph is a subset of its vertices such that for every edge (u, v) of the graph, either u or v is in vertex cover.
Although the name is Vertex Cover, the set covers all edges of the given graph.
The problem to find minimum size vertex cover of a graph is NP complete. But it can be solved in polynomial time for trees. In this post a solution
for Binary Tree is discussed. The same solution can be extended for n-ary trees.
For example, consider the following binary tree. The smallest vertex cover is {20, 50, 30} and size of the vertex cover is 3.

The idea is to consider following two possibilities for root and recursively for all nodes down the root.
1) Root is part of vertex cover: In this case root covers all children edges. We recursively calculate size of vertex covers for left and right
subtrees and add 1 to the result (for root).
2) Root is not part of vertex cover: In this case, both children of root must be included in vertex cover to cover all root to children edges. We
recursively calculate size of vertex covers of all grandchildren and number of children to the result (for two children of root).
Below is C implementation of above idea.
// A naive recursive C implementation for vertex cover problem for a tree
#include <stdio.h>
#include <stdlib.h>
// A utility function to find min of two integers
int min(int x, int y) { return (x < y)? x: y; }
/* A binary tree node has data, pointer to left child and a pointer to
right child */
struct node
{
int data;
struct node *left, *right;
};
// The function returns size of the minimum vertex cover
int vCover(struct node *root)
{
// The size of minimum vertex cover is zero if tree is empty or there
// is only one node
if (root == NULL)
return 0;
if (root->left == NULL && root->right == NULL)
return 0;
// Calculate size of vertex cover when root is part of it
int size_incl = 1 + vCover(root->left) + vCover(root->right);
// Calculate size of vertex cover when root is not part of it
int size_excl = 0;
if (root->left)
size_excl += 1 + vCover(root->left->left) + vCover(root->left->right);
if (root->right)
size_excl += 1 + vCover(root->right->left) + vCover(root->right->right);
// Return the minimum of two sizes
return min(size_incl, size_excl);
}
// A utility function to create a node
struct node* newNode( int data )

{
struct node* temp = (struct node *) malloc( sizeof(struct node) );
temp->data = data;
temp->left = temp->right = NULL;
return temp;
}
// Driver program to test above functions
int main()
{
// Let us construct the tree given in the above diagram
struct node *root
= newNode(20);
root->left
= newNode(8);
root->left->left
= newNode(4);
root->left->right
= newNode(12);
root->left->right->left = newNode(10);
root->left->right->right = newNode(14);
root->right
= newNode(22);
root->right->right
= newNode(25);
printf ("Size of the smallest vertex cover is %d ", vCover(root));
return 0;
}

Output:
Size of the smallest vertex cover is 3

Time complexity of the above naive recursive approach is exponential. It should be noted that the above function computes the same subproblems
again and again. For example, vCover of node with value 50 is evaluated twice as 50 is grandchild of 10 and child of 20.
Since same suproblems are called again, this problem has Overlapping Subprolems property. So Vertex Cover problem has both properties (see
this and this) of a dynamic programming problem. Like other typical Dynamic Programming(DP) problems, re-computations of same subproblems
can be avoided by storing the solutions to subproblems and solving problems in bottom up manner.
Following is C implementation of Dynamic Programming based solution. In the following solution, an additional field vc is added to tree nodes. The
initial value of vc is set as 0 for all nodes. The recursive function vCover() calculates vc for a node only if it is not already set.
/* Dynamic programming based program for Vertex Cover problem for
a Binary Tree */
#include <stdio.h>
#include <stdlib.h>
// A utility function to find min of two integers
int min(int x, int y) { return (x < y)? x: y; }
/* A binary tree node has data, pointer to left child and a pointer to
right child */
struct node
{
int data;
int vc;
struct node *left, *right;
};
// A memoization based function that returns size of the minimum vertex cover.
int vCover(struct node *root)
{
// The size of minimum vertex cover is zero if tree is empty or there
// is only one node
if (root == NULL)
return 0;
if (root->left == NULL && root->right == NULL)
return 0;
// If vertex cover for this node is already evaluated, then return it
// to save recomputation of same subproblem again.
if (root->vc != 0)
return root->vc;
// Calculate size of vertex cover when root is part of it
int size_incl = 1 + vCover(root->left) + vCover(root->right);
// Calculate size of vertex cover when root is not part of it
int size_excl = 0;
if (root->left)
size_excl += 1 + vCover(root->left->left) + vCover(root->left->right);
if (root->right)
size_excl += 1 + vCover(root->right->left) + vCover(root->right->right);

// Minimum of two values is vertex cover, store it before returning

root->vc = min(size_incl, size_excl);
return root->vc;
}
// A utility function to create a node
struct node* newNode( int data )
{
struct node* temp = (struct node *) malloc( sizeof(struct node) );
temp->data = data;
temp->left = temp->right = NULL;
temp->vc = 0; // Set the vertex cover as 0
return temp;
}
// Driver program to test above functions
int main()
{
// Let us construct the tree given in the above diagram
struct node *root
= newNode(20);
root->left
= newNode(8);
root->left->left
= newNode(4);
root->left->right
= newNode(12);
root->left->right->left = newNode(10);
root->left->right->right = newNode(14);
root->right
= newNode(22);
root->right->right
= newNode(25);
printf ("Size of the smallest vertex cover is %d ", vCover(root));
return 0;
}

Output:
Size of the smallest vertex cover is 3

References:
http://courses.csail.mit.edu/6.006/spring11/lectures/lec21.pdf
Exercise:
Extend the above solution for n-ary trees.

Count number of ways to reach a given score in a game

Consider a game where a player can score 3 or 5 or 10 points in a move. Given a total score n, find number of ways to reach the given score.
Examples:
Input: n = 20
Output: 4
There are following 4 ways to reach 20
(10, 10)
(5, 5, 10)
(5, 5, 5, 5)
(3, 3, 3, 3, 3, 5)
Input: n = 13
Output: 2
There are following 2 ways to reach 13
(3, 5, 5)
(3, 10)

We strongly recommend you to minimize the browser and try this yourself first.
This problem is a variation of coin change problem and can be solved in O(n) time and O(n) auxiliary space.
The idea is to create a table of size n+1 to store counts of all scores from 0 to n. For every possible move (3, 5 and 10), increment values in table.
// A C program to count number of possible ways to a given score
// can be reached in a game where a move can earn 3 or 5 or 10
#include <stdio.h>
// Returns number of ways to reach score n
int count(int n)
{
// table[i] will store count of solutions for
// value i.
int table[n+1], i;
// Initialize all table values as 0
memset(table, 0, sizeof(table));
// Base case (If given value is 0)
table[0] = 1;
// One by one consider given 3 moves and update the table[]
// values after the index greater than or equal to the
// value of the picked move
for (i=3; i<=n; i++)
table[i] += table[i-3];
for (i=5; i<=n; i++)
table[i] += table[i-5];
for (i=10; i<=n; i++)
table[i] += table[i-10];
return table[n];
}
// Driver program
int main(void)
{
int n = 20;
printf("Count for %d is %d\n", n, count(n));
n = 13;
printf("Count for %d is %d", n, count(n));
return 0;
}

Output:
Count for 20 is 4
Count for 13 is 2

Exercise: How to count score when (10, 5, 5), (5, 5, 10) and (5, 10, 5) are considered as different sequences of moves. Similarly, (5, 3, 3), (3,
5, 3) and (3, 3, 5) are considered different.

Weighted Job Scheduling

Given N jobs where every job is represented by following three elements of it.
1) Start Time
2) Finish Time.
3) Profit or Value Associated.
Find the maximum profit subset of jobs such that no two jobs in the subset overlap.
Example:
Input: Number of Jobs n = 4
Job Details {Start Time, Finish Time, Profit}
Job 1: {1, 2, 50}
Job 2: {3, 5, 20}
Job 3: {6, 19, 100}
Job 4: {2, 100, 200}
Output: The maximum profit is 250.
We can get the maximum profit by scheduling jobs 1 and 4.
Note that there is longer schedules possible Jobs 1, 2 and 3
but the profit with this schedule is 20+50+100 which is less than 250.

A simple version of this problem is discussed here where every job has same profit or value. The Greedy Strategy for activity selection doesnt
work here as the longer schedule may have smaller profit or value.
The above problem can be solved using following recursive solution.
1) First sort jobs according to finish time.
2) Now apply following recursive process.
// Here arr[] is array of n jobs
findMaximumProfit(arr[], n)
{
a) if (n == 1) return arr[0];
b) Return the maximum of following two profits.
(i) Maximum profit by excluding current job, i.e.,
findMaximumProfit(arr, n-1)
(ii) Maximum profit by including the current job
}
How to find the profit including current job?
The idea is to find the latest job before the current job (in
sorted array) that doesn't conflict with current job 'arr[n-1]'.
Once we find such a job, we recur for all jobs till that job and
add profit of current job to result.
In the above example, "job 1" is the latest non-conflicting
for "job 4" and "job 2" is the latest non-conflicting for "job 3".

The following is C++ implementation of above naive recursive method.

// C++ program for weighted job scheduling using Naive Recursive Method
#include <iostream>
#include <algorithm>
using namespace std;
// A job has start time, finish time and profit.
struct Job
{
int start, finish, profit;
};
// A utility function that is used for sorting events
// according to finish time
bool myfunction(Job s1, Job s2)
{
return (s1.finish < s2.finish);
}
// Find the latest job (in sorted array) that doesn't
// conflict with the job[i]. If there is no compatible job,
// then it returns -1.
int latestNonConflict(Job arr[], int i)
{
for (int j=i-1; j>=0; j--)
{
if (arr[j].finish <= arr[i-1].start)
return j;
}
return -1;

}
// A recursive function that returns the maximum possible
// profit from given array of jobs. The array of jobs must
// be sorted according to finish time.
int findMaxProfitRec(Job arr[], int n)
{
// Base case
if (n == 1) return arr[n-1].profit;
// Find profit when current job is inclueded
int inclProf = arr[n-1].profit;
int i = latestNonConflict(arr, n);
if (i != -1)
inclProf += findMaxProfitRec(arr, i+1);
// Find profit when current job is excluded
int exclProf = findMaxProfitRec(arr, n-1);
return max(inclProf, exclProf);
}
// The main function that returns the maximum possible
// profit from given array of jobs
int findMaxProfit(Job arr[], int n)
{
// Sort jobs according to finish time
sort(arr, arr+n, myfunction);
return findMaxProfitRec(arr, n);
}
// Driver program
int main()
{
Job arr[] = {{3, 10, 20}, {1, 2, 50}, {6, 19, 100}, {2, 100, 200}};
int n = sizeof(arr)/sizeof(arr[0]);
cout << "The optimal profit is " << findMaxProfit(arr, n);
return 0;
}

Output:
The optimal profit is 250

The above solution may contain many overlapping subproblems. For example if lastNonConflicting() always returns previous job, then
findMaxProfitRec(arr, n-1) is called twice and the time complexity becomes O(n*2n). As another example when lastNonConflicting() returns
previous to previous job, there are two recursive calls, for n-2 and n-1. In this example case, recursion becomes same as Fibonacci Numbers.
So this problem has both properties of Dynamic Programming, Optimal Substructure and Overlapping Subproblems.
Like other Dynamic Programming Problems, we can solve this problem by making a table that stores solution of subproblems.
Below is C++ implementation based on Dynamic Programming.
// C++ program for weighted job scheduling using Dynamic Programming.
#include <iostream>
#include <algorithm>
using namespace std;
// A job has start time, finish time and profit.
struct Job
{
int start, finish, profit;
};
// A utility function that is used for sorting events
// according to finish time
bool myfunction(Job s1, Job s2)
{
return (s1.finish < s2.finish);
}
// Find the latest job (in sorted array) that doesn't
// conflict with the job[i]
int latestNonConflict(Job arr[], int i)
{
for (int j=i-1; j>=0; j--)
{
if (arr[j].finish <= arr[i].start)
return j;

}
return -1;
}
// The main function that returns the maximum possible
// profit from given array of jobs
int findMaxProfit(Job arr[], int n)
{
// Sort jobs according to finish time
sort(arr, arr+n, myfunction);
// Create an array to store solutions of subproblems. table[i]
// stores the profit for jobs till arr[i] (including arr[i])
int *table = new int[n];
table[0] = arr[0].profit;
// Fill entries in M[] using recursive property
for (int i=1; i<n; i++)
{
// Find profit including the current job
int inclProf = arr[i].profit;
int l = latestNonConflict(arr, i);
if (l != -1)
inclProf += table[l];
// Store maximum of including and excluding
table[i] = max(inclProf, table[i-1]);
}
// Store result and free dynamic memory allocated for table[]
int result = table[n-1];
delete[] table;
return result;
}
// Driver program
int main()
{
Job arr[] = {{3, 10, 20}, {1, 2, 50}, {6, 19, 100}, {2, 100, 200}};
int n = sizeof(arr)/sizeof(arr[0]);
cout << "The optimal profit is " << findMaxProfit(arr, n);
return 0;
}

Output:
The optimal profit is 250

Time Complexity of the above Dynamic Programming Solution is O(n2). Note that the above solution can be optimized to O(nLogn) using Binary
Search in latestNonConflict() instead of linear search. Thanks to Garvit for suggesting this optimization.
References:
http://courses.cs.washington.edu/courses/cse521/13wi/slides/06dp-sched.pdf

Longest Even Length Substring such that Sum of First and Second Half is same
Given a string str of digits, find length of the longest substring of str, such that the length of the substring is 2k digits and sum of left k digits is equal
to the sum of right k digits.
Examples:
Input: str = "123123"
Output: 6
The complete string is of even length and sum of first and second
half digits is same
Input: str = "1538023"
Output: 4
The longest substring with same first and second half sum is "5380"

Simple Solution [ O(n3) ]

A Simple Solution is to check every substring of even length. The following is C based implementation of simple approach.
// A simple C based program to find length of longest even length
// substring with same sum of digits in left and right
#include<stdio.h>
#include<string.h>
int findLength(char *str)
{
int n = strlen(str);
int maxlen =0; // Initialize result
// Choose starting point of every substring
for (int i=0; i<n; i++)
{
// Choose ending point of even length substring
for (int j =i+1; j<n; j += 2)
{
int length = j-i+1;//Find length of current substr
// Calculate left & right sums for current substr
int leftsum = 0, rightsum =0;
for (int k =0; k<length/2; k++)
{
leftsum += (str[i+k]-'0');
rightsum += (str[i+k+length/2]-'0');
}
// Update result if needed
if (leftsum == rightsum && maxlen < length)
maxlen = length;
}
}
return maxlen;
}
// Driver program to test above function
int main(void)
{
char str[] = "1538023";
printf("Length of the substring is %d", findLength(str));
return 0;
}

Output:
Length of the substring is 4

Dynamic Programming [ O(n2) and O(n2) extra space]

The above solution can be optimized to work in O(n2) using Dynamic Programming. The idea is to build a 2D table that stores sums of
substrings. The following is C based implementation of Dynamic Programming approach.
// A C based program that uses Dynamic Programming to find length of the
// longest even substring with same sum of digits in left and right half
#include <stdio.h>
#include <string.h>
int findLength(char *str)

{
int n = strlen(str);
int maxlen = 0; // Initialize result
// A 2D table where sum[i][j] stores sum of digits
// from str[i] to str[j]. Only filled entries are
// the entries where j >= i
int sum[n][n];
// Fill the diagonal values for sunstrings of length 1
for (int i =0; i<n; i++)
sum[i][i] = str[i]-'0';
// Fill entries for substrings of length 2 to n
for (int len=2; len<=n; len++)
{
// Pick i and j for current substring
for (int i=0; i<n-len+1; i++)
{
int j = i+len-1;
int k = len/2;
// Calculate value of sum[i][j]
sum[i][j] = sum[i][j-k] + sum[j-k+1][j];
// Update result if 'len' is even, left and right
// sums are same and len is more than maxlen
if (len%2 == 0 && sum[i][j-k] == sum[(j-k+1)][j]
&& len > maxlen)
maxlen = len;
}
}
return maxlen;
}
// Driver program to test above function
int main(void)
{
char str[] = "153803";
printf("Length of the substring is %d", findLength(str));
return 0;
}

Output:
Length of the substring is 4

Time complexity of the above solution is O(n2), but it requires O(n2) extra space.

[A O(n2) and O(n) extra space solution]

The idea is to use a single dimensional array to store cumulative sum.
// A O(n^2) time and O(n) extra space solution
#include<bits/stdc++.h>
using namespace std;
int findLength(string str, int n)
{
int sum[n+1]; // To store cumulative sum from first digit to nth digit
sum[0] = 0;
/* Store cumulative sum of digits from first to last digit */
for (int i = 1; i <= n; i++)
sum[i] = (sum[i-1] + str[i-1] - '0'); /* convert chars to int */
int ans = 0; // initialize result
/* consider all even length substrings one by one */
for (int len = 2; len <= n; len += 2)
{
for (int i = 0; i <= n-len; i++)
{
int j = i + len - 1;
/* Sum of first and second half is same than update ans */
if (sum[i+len/2] - sum[i] == sum[i+len] - sum[i+len/2])
ans = max(ans, len);

}
}
return ans;
}
// Driver program to test above function
int main()
{
string str = "123123";
cout << "Length of the substring is " << findLength(str, str.length());
return 0;
}

Output:
Length of the substring is 6

Thanks to Gaurav Ahirwar for suggesting this method.

[A O(n2) time and O(1) extra space solution]

The idea is to consider all possible mid points (of even length substrings) and keep expanding on both sides to get and update optimal length as the
sum of two sides become equal.
Below is C++ implementation of the above idea.
// A O(n^2) time and O(1) extra space solution
#include<bits/stdc++.h>
using namespace std;
int findLength(string str, int n)
{
int ans = 0; // Initialize result
// Consider all possible midpoints one by one
for (int i = 0; i <= n-2; i++)
{
/* For current midpoint 'i', keep expanding substring on
both sides, if sum of both sides becomes equal update
ans */
int l = i, r = i + 1;
/* initialize left and right sum */
int lsum = 0, rsum = 0;
/* move on both sides till indexes go out of bounds */
while (r < n && l >= 0)
{
lsum += str[l] - '0';
rsum += str[r] - '0';
if (lsum == rsum)
ans = max(ans, r-l+1);
l--;
r++;
}
}
return ans;
}
// Driver program to test above function
int main()
{
string str = "123123";
cout << "Length of the substring is " << findLength(str, str.length());
return 0;
}

Output:
Length of the substring is 6

Thanks to Gaurav Ahirwar for suggesting this method.

Minimum Cost Polygon Triangulation

A triangulation of a convex polygon is formed by drawing diagonals between non-adjacent vertices (corners) such that the diagonals never
intersect. The problem is to find the cost of triangulation with the minimum cost. The cost of a triangulation is sum of the weights of its component
triangles. Weight of each triangle is its perimeter (sum of lengths of all sides)
See following example taken from this source.

Two triangulations of the same convex pentagon. The triangulation on the left has a cost of 8 + 2?2 + 2?5 (approximately 15.30), the
one on the right has a cost of 4 + 2?2 + 4?5 (approximately 15.77).
This problem has recursive substructure. The idea is to divide the polygon into three parts: a single triangle, the sub-polygon to the left, and the
sub-polygon to the right. We try all possible divisions like this and find the one that minimizes the cost of the triangle plus the cost of the
triangulation of the two sub-polygons.
Let Minimum Cost of triangulation of vertices from i to j be minCost(i, j)
If j <= i + 2 Then
minCost(i, j) = 0
Else
minCost(i, j) = Min { minCost(i, k) + minCost(k, j) + cost(i, k, j) }
Here k varies from 'i+1' to 'j-1'
Cost of a triangle formed by edges (i, j), (j, k) and (k, j) is
cost(i, j, k) = dist(i, j) + dist(j, k) + dist(k, j)

Following is C++ implementation of above naive recursive formula.

// Recursive implementation for minimum cost convex polygon triangulation
#include <iostream>
#include <cmath>
#define MAX 1000000.0
using namespace std;
// Structure of a point in 2D plane
struct Point
{
int x, y;
};
// Utility function to find minimum of two double values
double min(double x, double y)
{
return (x <= y)? x : y;
}
// A utility function
double dist(Point p1,
{
return sqrt((p1.x
(p1.y
}

to find distance between two points in a plane

Point p2)
- p2.x)*(p1.x - p2.x) +
- p2.y)*(p1.y - p2.y));

// A utility function to find cost of a triangle. The cost is considered

// as perimeter (sum of lengths of all edges) of the triangle
double cost(Point points[], int i, int j, int k)
{
Point p1 = points[i], p2 = points[j], p3 = points[k];
return dist(p1, p2) + dist(p2, p3) + dist(p3, p1);
}
// A recursive function to find minimum cost of polygon triangulation
// The polygon is represented by points[i..j].
double mTC(Point points[], int i, int j)
{
// There must be at least three points between i and j
// (including i and j)
if (j < i+2)

return 0;
// Initialize result as infinite
double res = MAX;
// Find minimum triangulation by considering all
for (int k=i+1; k<j; k++)
res = min(res, (mTC(points, i, k) + mTC(points, k, j) +
cost(points, i, k, j)));
return res;
}
// Driver program to test above functions
int main()
{
Point points[] = {{0, 0}, {1, 0}, {2, 1}, {1, 2}, {0, 2}};
int n = sizeof(points)/sizeof(points[0]);
cout << mTC(points, 0, n-1);
return 0;
}

Output:
15.3006

The above problem is similar to Matrix Chain Multiplication. The following is recursion tree for mTC(points[], 0, 4).

It can be easily seen in the above recursion tree that the problem has many overlapping subproblems. Since the problem has both properties:
Optimal Substructure and Overlapping Subproblems, it can be efficiently solved using dynamic programming.
Following is C++ implementation of dynamic programming solution.
// A Dynamic Programming based program to find minimum cost of convex
// polygon triangulation
#include <iostream>
#include <cmath>
#define MAX 1000000.0
using namespace std;
// Structure of a point in 2D plane
struct Point
{
int x, y;
};
// Utility function to find minimum of two double values
double min(double x, double y)
{
return (x <= y)? x : y;
}
// A utility function
double dist(Point p1,
{
return sqrt((p1.x
(p1.y
}

to find distance between two points in a plane

Point p2)
- p2.x)*(p1.x - p2.x) +
- p2.y)*(p1.y - p2.y));

// A utility function to find cost of a triangle. The cost is considered

// as perimeter (sum of lengths of all edges) of the triangle
double cost(Point points[], int i, int j, int k)
{

Point p1 = points[i], p2 = points[j], p3 = points[k];

return dist(p1, p2) + dist(p2, p3) + dist(p3, p1);
}
// A Dynamic programming based function to find minimum cost for convex
// polygon triangulation.
double mTCDP(Point points[], int n)
{
// There must be at least 3 points to form a triangle
if (n < 3)
return 0;
// table to store results of subproblems. table[i][j] stores cost of
// triangulation of points from i to j. The entry table[0][n-1] stores
// the final result.
double table[n][n];
// Fill table using above recursive formula. Note that the table
// is filled in diagonal fashion i.e., from diagonal elements to
// table[0][n-1] which is the result.
for (int gap = 0; gap < n; gap++)
{
for (int i = 0, j = gap; j < n; i++, j++)
{
if (j < i+2)
table[i][j] = 0.0;
else
{
table[i][j] = MAX;
for (int k = i+1; k < j; k++)
{
double val = table[i][k] + table[k][j] + cost(points,i,j,k);
if (table[i][j] > val)
table[i][j] = val;
}
}
}
}
return table[0][n-1];
}
// Driver program to test above functions
int main()
{
Point points[] = {{0, 0}, {1, 0}, {2, 1}, {1, 2}, {0, 2}};
int n = sizeof(points)/sizeof(points[0]);
cout << mTCDP(points, n);
return 0;
}

Output:
15.3006

Time complexity of the above dynamic programming solution is O(n3).

Please note that the above implementations assume that the points of covnvex polygon are given in order (either clockwise or anticlockwise)
Exercise:
Extend the above solution to print triangulation also. For the above example, the optimal triangulation is 0 3 4, 0 1 3, and 1 2 3.
Sources:
http://www.cs.utexas.edu/users/djimenez/utsa/cs3343/lecture12.html
http://www.cs.utoronto.ca/~heap/Courses/270F02/A4/chains/node2.html

Searching for Patterns | Set 1 (Naive Pattern Searching)

Given a text txt[0..n-1] and a pattern pat[0..m-1], write a function search(char pat[], char txt[]) that prints all occurrences of pat[] in txt[].
You may assume that n > m.
Examples:
1) Input:
txt[] = "THIS IS A TEST TEXT"
pat[] = "TEST"

Output:
Pattern found at index 10

2) Input:
txt[] = "AABAACAADAABAAABAA"
pat[] = "AABA"

Output:
Pattern found at index 0
Pattern found at index 9
Pattern found at index 13

Pattern searching is an important problem in computer science. When we do search for a string in notepad/word file or browser or database,
pattern searching algorithms are used to show the search results.
Naive Pattern Searching:
Slide the pattern over text one by one and check for a match. If a match is found, then slides by 1 again to check for subsequent matches.

C
// C program for Naive Pattern Searching algorithm
#include<stdio.h>
#include<string.h>
void search(char *pat, char *txt)
{
int M = strlen(pat);
int N = strlen(txt);
/* A loop to slide pat[] one by one */
for (int i = 0; i <= N - M; i++)
{
int j;
/* For current index i, check for pattern match */
for (j = 0; j < M; j++)
if (txt[i+j] != pat[j])
break;
if (j == M) // if pat[0...M-1] = txt[i, i+1, ...i+M-1]
printf("Pattern found at index %d \n", i);
}
}
/* Driver program to test above function */
int main()
{
char txt[] = "AABAACAADAABAAABAA";
char pat[] = "AABA";
search(pat, txt);
return 0;
}

Python
# Python program for Naive Pattern Searching
def search(pat, txt):
M = len(pat)
N = len(txt)
# A loop to slide pat[] one by one
for i in xrange(N-M+1):

# For current index i, check for pattern match

for j in xrange(M):
if txt[i+j] != pat[j]:
break
if j == M-1: # if pat[0...M-1] = txt[i, i+1, ...i+M-1]
print "Pattern found at index " + str(i)
# Driver program to test the above function
txt = "AABAACAADAABAAABAA"
pat = "AABA"
search (pat, txt)
# This code is contributed by Bhavya Jain

Pattern found at index 0

Pattern found at index 9
Pattern found at index 13

What is the best case?

The best case occurs when the first character of the pattern is not present in text at all.
txt[] = "AABCCAADDEE"
pat[] = "FAA"

The number of comparisons in best case is O(n).

What is the worst case ?
The worst case of Naive Pattern Searching occurs in following scenarios.
1) When all characters of the text and pattern are same.
txt[] = "AAAAAAAAAAAAAAAAAA"
pat[] = "AAAAA".

2) Worst case also occurs when only the last character is different.
txt[] = "AAAAAAAAAAAAAAAAAB"
pat[] = "AAAAB"

Number of comparisons in worst case is O(m*(n-m+1)). Although strings which have repeated characters are not likely to appear in English text,
they may well occur in other applications (for example, in binary texts). The KMP matching algorithm improves the worst case to O(n). We will be
covering KMP in the next post. Also, we will be writing more posts to cover all pattern searching algorithms and data structures.

Searching for Patterns | Set 2 (KMP Algorithm)

Given a text txt[0..n-1] and a pattern pat[0..m-1], write a function search(char pat[], char txt[]) that prints all occurrences of pat[] in txt[].
You may assume that n > m.
Examples:
1) Input:
txt[] = "THIS IS A TEST TEXT"
pat[] = "TEST"

Output:
Pattern found at index 10

2) Input:
txt[] = "AABAACAADAABAAABAA"
pat[] = "AABA"

Output:
Pattern found at index 0
Pattern found at index 9
Pattern found at index 13

Pattern searching is an important problem in computer science. When we do search for a string in notepad/word file or browser or database,
pattern searching algorithms are used to show the search results.
We have discussed Naive pattern searching algorithm in the previous post. The worst case complexity of Naive algorithm is O(m(n-m+1)). Time
complexity of KMP algorithm is O(n) in worst case.
KMP (Knuth Morris Pratt) Pattern Searching
The Naive pattern searching algorithm doesnt work well in cases where we see many matching characters followed by a mismatching character.
Following are some examples.
txt[] = "AAAAAAAAAAAAAAAAAB"
pat[] = "AAAAB"
txt[] = "ABABABCABABABCABABABC"
pat[] = "ABABAC" (not a worst case, but a bad case for Naive)

The KMP matching algorithm uses degenerating property (pattern having same sub-patterns appearing more than once in the pattern) of the
pattern and improves the worst case complexity to O(n). The basic idea behind KMPs algorithm is: whenever we detect a mismatch (after some
matches), we already know some of the characters in the text (since they matched the pattern characters prior to the mismatch). We take
advantage of this information to avoid matching the characters that we know will anyway match.
KMP algorithm does some preprocessing over the pattern pat[] and constructs an auxiliary array lps[] of size m (same as size of pattern). Here
name lps indicates longest proper prefix which is also suffix.. For each sub-pattern pat[0i] where i = 0 to m-1, lps[i] stores length of the
maximum matching proper prefix which is also a suffix of the sub-pattern pat[0..i].
lps[i] = the longest proper prefix of pat[0..i]
which is also a suffix of pat[0..i].

Examples:
For the pattern AABAACAABAA, lps[] is [0, 1, 0, 1, 2, 0, 1, 2, 3, 4, 5]
For the pattern ABCDE, lps[] is [0, 0, 0, 0, 0]
For the pattern AAAAA, lps[] is [0, 1, 2, 3, 4]
For the pattern AAABAAA, lps[] is [0, 1, 2, 0, 1, 2, 3]
For the pattern AAACAAAAAC, lps[] is [0, 1, 2, 0, 1, 2, 3, 3, 3, 4]
Searching Algorithm:
Unlike the Naive algo where we slide the pattern by one, we use a value from lps[] to decide the next sliding position. Let us see how we do that.
When we compare pat[j] with txt[i] and see a mismatch, we know that characters pat[0..j-1] match with txt[i-j+1i-1], and we also know that
lps[j-1] characters of pat[0j-1] are both proper prefix and suffix which means we do not need to match these lps[j-1] characters with txt[i-ji-1]
because we know that these characters will anyway match. See KMPSearch() in the below code for details.
Preprocessing Algorithm:
In the preprocessing part, we calculate values in lps[]. To do that, we keep track of the length of the longest prefix suffix value (we use len variable
for this purpose) for the previous index. We initialize lps[0] and len as 0. If pat[len] and pat[i] match, we increment len by 1 and assign the
incremented value to lps[i]. If pat[i] and pat[len] do not match and len is not 0, we update len to lps[len-1]. See computeLPSArray () in the below
code for details.

C
// C program for implementation of KMP pattern searching
// algorithm
#include<stdio.h>
#include<string.h>
#include<stdlib.h>
void computeLPSArray(char *pat, int M, int *lps);
void KMPSearch(char *pat, char *txt)
{
int M = strlen(pat);
int N = strlen(txt);
// create lps[] that will hold the longest prefix suffix
// values for pattern
int *lps = (int *)malloc(sizeof(int)*M);
int j = 0; // index for pat[]
// Preprocess the pattern (calculate lps[] array)
computeLPSArray(pat, M, lps);
int i = 0; // index for txt[]
while (i < N)
{
if (pat[j] == txt[i])
{
j++;
i++;
}
if (j == M)
{
printf("Found pattern at index %d \n", i-j);
j = lps[j-1];
}
// mismatch after j matches
else if (i < N && pat[j] != txt[i])
{
// Do not match lps[0..lps[j-1]] characters,
// they will match anyway
if (j != 0)
j = lps[j-1];
else
i = i+1;
}
}
free(lps); // to avoid memory leak
}
void computeLPSArray(char *pat, int M, int *lps)
{
int len = 0; // length of the previous longest prefix suffix
int i;
lps[0] = 0; // lps[0] is always 0
i = 1;
// the loop calculates lps[i] for i = 1 to M-1
while (i < M)
{
if (pat[i] == pat[len])
{
len++;
lps[i] = len;
i++;
}
else // (pat[i] != pat[len])
{
if (len != 0)
{
// This is tricky. Consider the example
// AAACAAAA and i = 7.
len = lps[len-1];
// Also, note that we do not increment i here
}
else // if (len == 0)

{
lps[i] = 0;
i++;
}
}
}
}
// Driver program to test above function
int main()
{
char *txt = "ABABDABACDABABCABAB";
char *pat = "ABABCABAB";
KMPSearch(pat, txt);
return 0;
}

Python
# Python program for KMP Algorithm
def KMPSearch(pat, txt):
M = len(pat)
N = len(txt)
# create lps[] that will hold the longest prefix suffix
# values for pattern
lps = [0]*M
j = 0 # index for pat[]
# Preprocess the pattern (calculate lps[] array)
computeLPSArray(pat, M, lps)
i = 0 # index for txt[]
while i < N:
if pat[j] == txt[i]:
i+=1
j+=1
if j==M:
print "Found pattern at index " + str(i-j)
j = lps[j-1]
# mismatch after j matches
elif i < N and pat[j] != txt[i]:
# Do not match lps[0..lps[j-1]] characters,
# they will match anyway
if j != 0:
j = lps[j-1]
else:
i+=1
def computeLPSArray(pat, M, lps):
len = 0 # length of the previous longest prefix suffix
lps[0] # lps[0] is always 0
i = 1
# the loop calculates lps[i] for i = 1 to M-1
while i < M:
if pat[i]==pat[len]:
len+=1
lps[i] = len
i+=1
else:
if len!=0:
# This is tricky. Consier the example AAACAAAA
# and i = 7
len = lps[len-1]
# Also, note that we do not increment i here
else:
lps[i] = 0
i+=1
txt = "ABABDABACDABABCABAB"
pat = "ABABCABAB"
KMPSearch(pat, txt)
# This code is contributed by Bhavya Jain

Found pattern at index 10

Searching for Patterns | Set 3 (Rabin-Karp Algorithm)

Given a text txt[0..n-1] and a pattern pat[0..m-1], write a function search(char pat[], char txt[]) that prints all occurrences of pat[] in txt[].
You may assume that n > m.
Examples:
1) Input:
txt[] = "THIS IS A TEST TEXT"
pat[] = "TEST"

Output:
Pattern found at index 10
2) Input:
txt[] = "AABAACAADAABAAABAA"
pat[] = "AABA"

Output:
Pattern found at index 0
Pattern found at index 9
Pattern found at index 13

The Naive String Matching algorithm slides the pattern one by one. After each slide, it one by one checks characters at the current shift and if all
characters match then prints the match.
Like the Naive Algorithm, Rabin-Karp algorithm also slides the pattern one by one. But unlike the Naive algorithm, Rabin Karp algorithm matches
the hash value of the pattern with the hash value of current substring of text, and if the hash values match then only it starts matching individual
characters. So Rabin Karp algorithm needs to calculate hash values for following strings.
1) Pattern itself.
2) All the substrings of text of length m.
Since we need to efficiently calculate hash values for all the substrings of size m of text, we must have a hash function which has following property.
Hash at the next shift must be efficiently computable from the current hash value and next character in text or we can say hash(txt[s+1 .. s+m])
must be efficiently computable from hash(txt[s .. s+m-1]) and txt[s+m] i.e., hash(txt[s+1 .. s+m])= rehash(txt[s+m], hash(txt[s .. s+m1]) and rehash must be O(1) operation.
The hash function suggested by Rabin and Karp calculates an integer value. The integer value for a string is numeric value of a string. For example,
if all possible characters are from 1 to 10, the numeric value of 122 will be 122. The number of possible characters is higher than 10 (256 in
general) and pattern length can be large. So the numeric values cannot be practically stored as an integer. Therefore, the numeric value is calculated
using modular arithmetic to make sure that the hash values can be stored in an integer variable (can fit in memory words). To do rehashing, we
need to take off the most significant digit and add the new least significant digit for in hash value. Rehashing is done using the following formula.
hash( txt[s+1 .. s+m] ) = d ( hash( txt[s .. s+m-1]) txt[s]*h ) + txt[s + m] ) mod q
hash( txt[s .. s+m-1] ) : Hash value at shift s.
hash( txt[s+1 .. s+m] ) : Hash value at next shift (or shift s+1)
d: Number of characters in the alphabet
q: A prime number
h: d^(m-1)

C/C++
/* Following program is a C implementation of Rabin Karp
Algorithm given in the CLRS book */
#include<stdio.h>
#include<string.h>
// d is the number of characters in input alphabet
#define d 256
/* pat -> pattern
txt -> text
q -> A prime number
*/
void search(char pat[], char txt[], int q)
{
int M = strlen(pat);
int N = strlen(txt);

int
int
int
int

i, j;
p = 0; // hash value for pattern
t = 0; // hash value for txt
h = 1;

// The value of h would be "pow(d, M-1)%q"

for (i = 0; i < M-1; i++)
h = (h*d)%q;
// Calculate the hash value of pattern and first
// window of text
for (i = 0; i < M; i++)
{
p = (d*p + pat[i])%q;
t = (d*t + txt[i])%q;
}
// Slide the pattern over text one by one
for (i = 0; i <= N - M; i++)
{
//
//
//
if
{

Check the hash values of current window of text

and pattern. If the hash values match then only
check for characters on by one
( p == t )
/* Check for characters one by one */
for (j = 0; j < M; j++)
{
if (txt[i+j] != pat[j])
break;
}
// if p == t and pat[0...M-1] = txt[i, i+1, ...i+M-1]
if (j == M)
printf("Pattern found at index %d \n", i);

}
// Calculate hash value for next window of text: Remove
// leading digit, add trailing digit
if ( i < N-M )
{
t = (d*(t - txt[i]*h) + txt[i+M])%q;
// We might get negative value of t, converting it
// to positive
if (t < 0)
t = (t + q);
}
}
}
/* Driver program to test above function */
int main()
{
char txt[] = "GEEKS FOR GEEKS";
char pat[] = "GEEK";
int q = 101; // A prime number
search(pat, txt, q);
return 0;
}

Python
# Following program is the python implementation of
# Rabin Karp Algorithm given in CLRS book
# d is the number of characters in input alphabet
d = 256
# pat -> pattern
# txt -> text
# q
-> A prime number
def search(pat, txt, q):
M = len(pat)
N = len(txt)
i = 0
j = 0
p = 0
# hash value for pattern

t = 0
h = 1

# hash value for txt

# The value of h would be "pow(d, M-1)%q"

for i in xrange(M-1):
h = (h*d)%q
# Calculate the hash value of pattern and first window
# of text
for i in xrange(M):
p = (d*p + ord(pat[i]))%q
t = (d*t + ord(txt[i]))%q
# Slide the pattern over text one by one
for i in xrange(N-M+1):
# Check the hash values of current window of text and
# pattern if the hash values match then only check
# for characters on by one
if p==t:
# Check for characters one by one
for j in xrange(M):
if txt[i+j] != pat[j]:
break
j+=1
# if p == t and pat[0...M-1] = txt[i, i+1, ...i+M-1]
if j==M:
print "Pattern found at index " + str(i)
# Calculate hash value for next window of text: Remove
# leading digit, add trailing digit
if i < N-M:
t = (d*(t-ord(txt[i])*h) + ord(txt[i+M]))%q
# We might get negative values of t, converting it to
# positive
if t < 0:
t = t+q
# Driver program to test the above function
txt = "GEEKS FOR GEEKS"
pat = "GEEK"
q = 101 # A prime number
search(pat,txt,q)
# This code is contributed by Bhavya Jain

Pattern found at index 0

Pattern found at index 10

The average and best case running time of the Rabin-Karp algorithm is O(n+m), but its worst-case time is O(nm). Worst case of Rabin-Karp
algorithm occurs when all characters of pattern and text are same as the hash values of all the substrings of txt[] match with hash value of pat[]. For
example pat[] = AAA and txt[] = AAAAAAA.
References:
http://net.pku.edu.cn/~course/cs101/2007/resource/Intro2Algorithm/book6/chap34.htm
http://www.cs.princeton.edu/courses/archive/fall04/cos226/lectures/string.4up.pdf
http://en.wikipedia.org/wiki/Rabin-Karp_string_search_algorithm
Related Posts:
Searching for Patterns | Set 1 (Naive Pattern Searching)
Searching for Patterns | Set 2 (KMP Algorithm)

Searching for Patterns | Set 4 (A Naive Pattern Searching Question)

Question: We have discussed Naive String matching algorithm here. Consider a situation where all characters of pattern are different. Can we
modify the original Naive String Matching algorithm so that it works better for these types of patterns. If we can, then what are the changes to
original algorithm?
Solution: In the original Naive String matching algorithm , we always slide the pattern by 1. When all characters of pattern are different, we can
slide the pattern by more than 1. Let us see how can we do this. When a mismatch occurs after j matches, we know that the first character of
pattern will not match the j matched characters because all characters of pattern are different. So we can always slide the pattern by j without
missing any valid shifts. Following is the modified code that is optimized for the special patterns.

C
/* C program for A modified Naive Pattern Searching
algorithm that is optimized for the cases when all
characters of pattern are different */
#include<stdio.h>
#include<string.h>
/* A modified Naive Pettern Searching algorithn that is optimized
for the cases when all characters of pattern are different */
void search(char pat[], char txt[])
{
int M = strlen(pat);
int N = strlen(txt);
int i = 0;
while (i <= N - M)
{
int j;
/* For current index i, check for pattern match */
for (j = 0; j < M; j++)
if (txt[i+j] != pat[j])
break;
if (j == M) // if pat[0...M-1] = txt[i, i+1, ...i+M-1]
{
printf("Pattern found at index %d \n", i);
i = i + M;
}
else if (j == 0)
i = i + 1;
else
i = i + j; // slide the pattern by j
}
}
/* Driver program to test above function */
int main()
{
char txt[] = "ABCEABCDABCEABCD";
char pat[] = "ABCD";
search(pat, txt);
return 0;
}

Python
# Python program for A modified Naive Pattern Searching
# algorithm that is optimized for the cases when all
# characters of pattern are different
def search(pat, txt):
M = len(pat)
N = len(txt)
i = 0
while i <= N-M:
# For current index i, check for pattern match
for j in xrange(M):
if txt[i+j] != pat[j]:
break
j += 1
if j==M:
# if pat[0...M-1] = txt[i,i+1,...i+M-1]
print "Pattern found at index " + str(i)
i = i + M

elif j==0:
i = i + 1
else:
i = i+ j

# slide the pattern by j

# Driver program to test the above function

txt = "ABCEABCDABCEABCD"
pat = "ABCD"
search(pat, txt)
# This code is contributed by Bhavya Jain

Pattern found at index 4

Pattern found at index 12

Searching for Patterns | Set 5 (Finite Automata)

Given a text txt[0..n-1] and a pattern pat[0..m-1], write a function search(char pat[], char txt[]) that prints all occurrences of pat[] in txt[].
You may assume that n > m.
Examples:
1) Input:
txt[] = "THIS IS A TEST TEXT"
pat[] = "TEST"

Output:
Pattern found at index 10
2) Input:
txt[] = "AABAACAADAABAAABAA"
pat[] = "AABA"

Output:
Pattern found at index 0
Pattern found at index 9
Pattern found at index 13

Pattern searching is an important problem in computer science. When we do search for a string in notepad/word file or browser or database,
pattern searching algorithms are used to show the search results.
We have discussed the following algorithms in the previous posts:
Naive Algorithm
KMP Algorithm
Rabin Karp Algorithm
In this post, we will discuss Finite Automata (FA) based pattern searching algorithm. In FA based algorithm, we preprocess the pattern and build a
2D array that represents a Finite Automata. Construction of the FA is the main tricky part of this algorithm. Once the FA is built, the searching is
simple. In search, we simply need to start from the first state of the automata and first character of the text. At every step, we consider next
character of text, look for the next state in the built FA and move to new state. If we reach final state, then pattern is found in text. Time complexity
of the search prcess is O(n).
Before we discuss FA construction, let us take a look at the following FA for pattern ACACAGA.

The abvoe diagrams represent graphical and tabular representations of pattern ACACAGA.
Number of states in FA will be M+1 where M is length of the pattern. The main thing to construct FA is to get the next state from the current state
for every possible character. Given a character x and a state k, we can get the next state by considering the string pat[0..k-1]x which is basically
concatenation of pattern characters pat[0], pat[1] pat[k-1] and the character x. The idea is to get length of the longest prefix of the given pattern
such that the prefix is also suffix of pat[0..k-1]x. The value of length gives us the next state. For example, let us see how to get the next state from
current state 5 and character C in the above diagram. We need to consider the string, pat[0..5]C which is ACACAC. The lenght of the longest
prefix of the pattern such that the prefix is suffix of ACACACis 4 (ACAC). So the next state (from state 5) is 4 for character C.

In the following code, computeTF() constructs the FA. The time complexity of the computeTF() is O(m^3*NO_OF_CHARS) where m is length
of the pattern and NO_OF_CHARS is size of alphabet (total number of possible characters in pattern and text). The implementation tries all
possible prefixes starting from the longest possible that can be a suffix of pat[0..k-1]x. There are better implementations to construct FA in
O(m*NO_OF_CHARS) (Hint: we can use something like lps array construction in KMP algorithm). We have covered the better implementation
in our next post on pattern searching.
#include<stdio.h>
#include<string.h>
#define NO_OF_CHARS 256
int getNextState(char *pat, int M, int state, int x)
{
// If the character c is same as next character in pattern,
// then simply increment state
if (state < M && x == pat[state])
return state+1;
int ns, i; // ns stores the result which is next state
// ns finally contains the longest prefix which is also suffix
// in "pat[0..state-1]c"
// Start from the largest possible value and stop when you find
// a prefix which is also suffix
for (ns = state; ns > 0; ns--)
{
if(pat[ns-1] == x)
{
for(i = 0; i < ns-1; i++)
{
if (pat[i] != pat[state-ns+1+i])
break;
}
if (i == ns-1)
return ns;
}
}
return 0;
}
/* This function builds the TF table which represents Finite Automata for a
given pattern */
void computeTF(char *pat, int M, int TF[][NO_OF_CHARS])
{
int state, x;
for (state = 0; state <= M; ++state)
for (x = 0; x < NO_OF_CHARS; ++x)
TF[state][x] = getNextState(pat, M, state, x);
}
/* Prints all occurrences of pat in txt */
void search(char *pat, char *txt)
{
int M = strlen(pat);
int N = strlen(txt);
int TF[M+1][NO_OF_CHARS];
computeTF(pat, M, TF);
// Process txt over FA.
int i, state=0;
for (i = 0; i < N; i++)
{
state = TF[state][txt[i]];
if (state == M)
{
printf ("\n patterb found at index %d", i-M+1);
}
}
}
// Driver program to test above function
int main()
{
char *txt = "AABAACAADAABAAABAA";
char *pat = "AABA";
search(pat, txt);
return 0;

Output:
Pattern found at index 0
Pattern found at index 9
Pattern found at index 13

References:
Introduction to Algorithms by Thomas H. Cormen, Charles E. Leiserson, Ronald L. Rivest, Clifford Stein

Pattern Searching | Set 6 (Efficient Construction of Finite Automata)

In the previous post, we discussed Finite Automata based pattern searching algorithm. The FA (Finite Automata) construction method discussed in
previous post takes O((m^3)*NO_OF_CHARS) time. FA can be constructed in O(m*NO_OF_CHARS) time. In this post, we will discuss the
O(m*NO_OF_CHARS) algorithm for FA construction. The idea is similar to lps (longest prefix suffix) array construction discussed in the KMP
algorithm. We use previously filled rows to fill a new row.

The abvoe diagrams represent graphical and tabular representations of pattern ACACAGA.
Algorithm:
1) Fill the first row. All entries in first row are always 0 except the entry for pat[0] character. For pat[0] character, we always need to go to state
1.
2) Initialize lps as 0. lps for the first index is always 0.
3) Do following for rows at index i = 1 to M. (M is the length of the pattern)
..a) Copy the entries from the row at index equal to lps.
..b) Update the entry for pat[i] character to i+1.
..c) Update lps lps = TF[lps][pat[i]] where TF is the 2D array which is being constructed.
Implementation
Following is C implementation for the above algorithm.
#include<stdio.h>
#include<string.h>
#define NO_OF_CHARS 256
/* This function builds the TF table which represents Finite Automata for a
given pattern */
void computeTransFun(char *pat, int M, int TF[][NO_OF_CHARS])
{
int i, lps = 0, x;
// Fill entries in first row
for (x =0; x < NO_OF_CHARS; x++)
TF[0][x] = 0;
TF[0][pat[0]] = 1;
// Fill entries in other rows
for (i = 1; i<= M; i++)
{
// Copy values from row at index lps
for (x = 0; x < NO_OF_CHARS; x++)
TF[i][x] = TF[lps][x];
// Update the entry corresponding to this character
TF[i][pat[i]] = i + 1;
// Update lps for next row to be filled
if (i < M)
lps = TF[lps][pat[i]];
}
}
/* Prints all occurrences of pat in txt */

void search(char *pat, char *txt)

{
int M = strlen(pat);
int N = strlen(txt);
int TF[M+1][NO_OF_CHARS];
computeTransFun(pat, M, TF);
// process text over FA.
int i, j=0;
for (i = 0; i < N; i++)
{
j = TF[j][txt[i]];
if (j == M)
{
printf ("\n pattern found at index %d", i-M+1);
}
}
}
/* Driver program to test above function */
int main()
{
char *txt = "GEEKS FOR GEEKS";
char *pat = "GEEKS";
search(pat, txt);
getchar();
return 0;
}

Output:
pattern found at index 0
pattern found at index 10

Time Complexity for FA construction is O(M*NO_OF_CHARS). The code for search is same as the previous post and time complexity for it is
O(n).

Pattern Searching | Set 7 (Boyer Moore Algorithm Bad Character Heuristic)

Given a text txt[0..n-1] and a pattern pat[0..m-1], write a function search(char pat[], char txt[]) that prints all occurrences of pat[] in txt[]. You
may assume that n > m.
Examples:
1) Input:
txt[] = "THIS IS A TEST TEXT"
pat[] = "TEST"
Output:
Pattern found at index 10
2) Input:
txt[] = "AABAACAADAABAAABAA"
pat[] = "AABA"
Output:
Pattern found at index 0
Pattern found at index 9
Pattern found at index 13

Pattern searching is an important problem in computer science. When we do search for a string in notepad/word file or browser or database,
pattern searching algorithms are used to show the search results.
We have discussed the following algorithms in the previous posts:
Naive Algorithm
KMP Algorithm
Rabin Karp Algorithm
Finite Automata based Algorithm
In this post, we will discuss Boyer Moore pattern searching algorithm. Like KMP and Finite Automata algorithms, Boyer Moore algorithm also
preprocesses the pattern.
Boyer Moore is a combination of following two approaches.
1) Bad Character Heuristic
2) Good Suffix Heuristic
Both of the above heuristics can also be used independently to search a pattern in a text. Let us first understand how two independent approaches
work together in the Boyer Moore algorithm. If we take a look at the Naive algorithm, it slides the pattern over the text one by one. KMP
algorithm does preprocessing over the pattern so that the pattern can be shifted by more than one. The Boyer Moore algorithm does
preprocessing for the same reason. It preporcesses the pattern and creates different arrays for both heuristics. At every step, it slides the pattern
by max of the slides suggested by the two heuristics. So it uses best of the two heuristics at every step. Unlike the previous pattern searching
algorithms, Boyer Moore algorithm starts matching from the last character of the pattern.
In this post, we will discuss bad character heuristic, and discuss Good Suffix heuristic in the next post.
The idea of bad character heuristic is simple. The character of the text which doesnt match with the current character of pattern is called the Bad
Character. Whenever a character doesnt match, we slide the pattern in such a way that aligns the bad character with the last occurrence of it in
pattern. We preprocess the pattern and store the last occurrence of every possible character in an array of size equal to alphabet size. If the
character is not present at all, then it may result in a shift by m (length of pattern). Therefore, the bad character heuristic takes O(n/m) time in the
best case.
/* Program for Bad Character Heuristic of Boyer Moore String Matching Algorithm */
# include <limits.h>
# include <string.h>
# include <stdio.h>
# define NO_OF_CHARS 256
// A utility function to get maximum of two integers
int max (int a, int b) { return (a > b)? a: b; }
// The preprocessing function for Boyer Moore's bad character heuristic
void badCharHeuristic( char *str, int size, int badchar[NO_OF_CHARS])
{
int i;
// Initialize all occurrences as -1

for (i = 0; i < NO_OF_CHARS; i++)

badchar[i] = -1;
// Fill the actual value of last occurrence of a character
for (i = 0; i < size; i++)
badchar[(int) str[i]] = i;
}
/* A pattern searching function that uses Bad Character Heuristic of
Boyer Moore Algorithm */
void search( char *txt, char *pat)
{
int m = strlen(pat);
int n = strlen(txt);
int badchar[NO_OF_CHARS];
/* Fill the bad character array by calling the preprocessing
function badCharHeuristic() for given pattern */
badCharHeuristic(pat, m, badchar);
int s = 0; // s is shift of the pattern with respect to text
while(s <= (n - m))
{
int j = m-1;
/* Keep reducing index j of pattern while characters of
pattern and text are matching at this shift s */
while(j >= 0 && pat[j] == txt[s+j])
j--;
/* If the pattern is present at current shift, then index j
will become -1 after the above loop */
if (j < 0)
{
printf("\n pattern occurs at shift = %d", s);
/* Shift the pattern so that the next character in text
aligns with the last occurrence of it in pattern.
The condition s+m < n is necessary for the case when
pattern occurs at the end of text */
s += (s+m < n)? m-badchar[txt[s+m]] : 1;
}
else
/* Shift the pattern so that the bad character in text
aligns with the last occurrence of it in pattern. The
max function is used to make sure that we get a positive
shift. We may get a negative shift if the last occurrence
of bad character in pattern is on the right side of the
current character. */
s += max(1, j - badchar[txt[s+j]]);
}
}
/* Driver program to test above funtion */
int main()
{
char txt[] = "ABAAABCD";
char pat[] = "ABC";
search(txt, pat);
return 0;
}

Output:
pattern occurs at shift = 4

The Bad Character Heuristic may take O(mn) time in worst case. The worst case occurs when all characters of the text and pattern are same. For
example, txt[] = AAAAAAAAAAAAAAAAAA and pat[] = AAAAA.

Suffix Array | Set 1 (Introduction)

We strongly recommend to read following post on suffix trees as a pre-requisite for this post.
Pattern Searching | Set 8 (Suffix Tree Introduction)
A suffix array is a sorted array of all suffixes of a given string. The definition is similar to Suffix Tree which is compressed trie of all suffixes
of the given text. Any suffix tree based algorithm can be replaced with an algorithm that uses a suffix array enhanced with additional information
and solves the same problem in the same time complexity (Source Wiki).
A suffix array can be constructed from Suffix tree by doing a DFS traversal of the suffix tree. In fact Suffix array and suffix tree both can be
constructed from each other in linear time.
Advantages of suffix arrays over suffix trees include improved space requirements, simpler linear time construction algorithms (e.g., compared to
Ukkonens algorithm) and improved cache locality (Source: Wiki)
Example:
Let the given string be "banana".
0
1
2
3
4
5

banana
anana
nana
ana
na
a

Sort the Suffixes

---------------->
alphabetically

5
3
1
0
4
2

a
ana
anana
banana
na
nana

So the suffix array for "banana" is {5, 3, 1, 0, 4, 2}

Naive method to build Suffix Array

A simple method to construct suffix array is to make an array of all suffixes and then sort the array. Following is implementation of simple method.
// Naive algorithm for building suffix array of a given text
#include <iostream>
#include <cstring>
#include <algorithm>
using namespace std;
// Structure to store information of a suffix
struct suffix
{
int index;
char *suff;
};
// A comparison function used by sort() to compare two suffixes
int cmp(struct suffix a, struct suffix b)
{
return strcmp(a.suff, b.suff) < 0? 1 : 0;
}
// This is the main function that takes a string 'txt' of size n as an
// argument, builds and return the suffix array for the given string
int *buildSuffixArray(char *txt, int n)
{
// A structure to store suffixes and their indexes
struct suffix suffixes[n];
// Store suffixes and their indexes in an array of structures.
// The structure is needed to sort the suffixes alphabatically
// and maintain their old indexes while sorting
for (int i = 0; i < n; i++)
{
suffixes[i].index = i;
suffixes[i].suff = (txt+i);
}
// Sort the suffixes using the comparison function
// defined above.
sort(suffixes, suffixes+n, cmp);
// Store indexes of all sorted suffixes in the suffix array
int *suffixArr = new int[n];
for (int i = 0; i < n; i++)
suffixArr[i] = suffixes[i].index;
// Return the suffix array
return suffixArr;
}

// A utility function to print an array of given size

void printArr(int arr[], int n)
{
for(int i = 0; i < n; i++)
cout << arr[i] << " ";
cout << endl;
}
// Driver program to test above functions
int main()
{
char txt[] = "banana";
int n = strlen(txt);
int *suffixArr = buildSuffixArray(txt, n);
cout << "Following is suffix array for " << txt << endl;
printArr(suffixArr, n);
return 0;
}

Output:
Following is suffix array for banana
5 3 1 0 4 2

The time complexity of above method to build suffix array is O(n2Logn) if we consider a O(nLogn) algorithm used for sorting. The sorting step
itself takes O(n2Logn) time as every comparison is a comparison of two strings and the comparison takes O(n) time.
There are many efficient algorithms to build suffix array. We will soon be covering them as separate posts.
Search a pattern using the built Suffix Array
To search a pattern in a text, we preprocess the text and build a suffix array of the text. Since we have a sorted array of all suffixes, Binary Search
can be used to search. Following is the search function. Note that the function doesnt report all occurrences of pattern, it only report one of them.
// This code only contains search() and main. To make it a complete running
// above code or see http://ideone.com/1Io9eN
// A suffix array based search function to search a given pattern
// 'pat' in given text 'txt' using suffix array suffArr[]
void search(char *pat, char *txt, int *suffArr, int n)
{
int m = strlen(pat); // get length of pattern, needed for strncmp()
// Do simple binary search for the pat in txt using the
// built suffix array
int l = 0, r = n-1; // Initilize left and right indexes
while (l <= r)
{
// See if 'pat' is prefix of middle suffix in suffix array
int mid = l + (r - l)/2;
int res = strncmp(pat, txt+suffArr[mid], m);
// If match found at the middle, print it and return
if (res == 0)
{
cout << "Pattern found at index " << suffArr[mid];
return;
}
// Move to left half if pattern is alphabtically less than
// the mid suffix
if (res < 0) r = mid - 1;
// Otherwise move to right half
else l = mid + 1;
}
// We reach here if return statement in loop is not executed
cout << "Pattern not found";
}
// Driver program to test above function
int main()
{
char txt[] = "banana"; // text
char pat[] = "nan"; // pattern to be searched in text
// Build suffix array
int n = strlen(txt);
int *suffArr = buildSuffixArray(txt, n);

// search pat in txt using the built suffix array

search(pat, txt, suffArr, n);
return 0;
}

Output:
Pattern found at index 2

The time complexity of the above search function is O(mLogn). There are more efficient algorithms to search pattern once the suffix array is built.
In fact there is a O(m) suffix array based algorithm to search a pattern. We will soon be discussing efficient algorithm for search.
Applications of Suffix Array
Suffix array is an extremely useful data structure, it can be used for a wide range of problems. Following are some famous problems where Suffix
array can be used.
1) Pattern Searching
2) Finding the longest repeated substring
3) Finding the longest common substring
4) Finding the longest palindrome in a string
See this for more problems where Suffix arrays can be used.
This post is a simple introduction. There is a lot to cover in Suffix arrays. We have discussed a O(nLogn) algorithm for Suffix Array construction
here. We will soon be discussing more efficient suffix array algorithms.
References:
http://www.stanford.edu/class/cs97si/suffix-array.pdf
http://en.wikipedia.org/wiki/Suffix_array

Anagram Substring Search (Or Search for all permutations)

Given a text txt[0..n-1] and a pattern pat[0..m-1], write a function search(char pat[], char txt[]) that prints all occurrences of pat[] and its
permutations (or anagrams) in txt[]. You may assume that n > m.
Expected time complexity is O(n)
Examples:
1) Input: txt[] =
Output: Found
Found
Found
2) Input: txt[] =
Output: Found
Found
Found

"BACDGABCDA" pat[] = "ABCD"

at Index 0
at Index 5
at Index 6
"AAABABAA" pat[] = "AABA"
at Index 0
at Index 1
at Index 4

This problem is slightly different from standard pattern searching problem, here we need to search for anagrams as well. Therefore, we cannot
directly apply standard pattern searching algorithms like KMP, Rabin Karp, Boyer Moore, etc.
A simple idea is to modify Rabin Karp Algorithm. For example we can keep the hash value as sum of ASCII values of all characters under modulo
of a big prime number. For every character of text, we can add the current character to hash value and subtract the first character of previous
window. This solution looks good, but like standard Rabin Karp, the worst case time complexity of this solution is O(mn). The worst case occurs
when all hash values match and we one by one match all characters.
We can achieve O(n) time complexity under the assumption that alphabet size is fixed which is typically true as we have maximum 256 possible
characters in ASCII. The idea is to use two count arrays:
1) The first count array store frequencies of characters in pattern.
2) The second count array stores frequencies of characters in current window of text.
The important thing to note is, time complexity to compare two count arrays is O(1) as the number of elements in them are fixed (independent of
pattern and text sizes). Following are steps of this algorithm.
1) Store counts of frequencies of pattern in first count array countP[]. Also store counts of frequencies of characters in first window of text in
array countTW[].
2) Now run a loop from i = M to N-1. Do following in loop.
..a) If the two count arrays are identical, we found an occurrence.
..b) Increment count of current character of text in countTW[]
..c) Decrement count of first character in previous window in countWT[]
3) The last window is not checked by above loop, so explicitly check it.
Following is C++ implementation of above algorithm.
// C++ program to search all anagrams of a pattern in a text
#include<iostream>
#include<cstring>
#define MAX 256
using namespace std;
// This function returns true if contents of arr1[] and arr2[]
// are same, otherwise false.
bool compare(char arr1[], char arr2[])
{
for (int i=0; i<MAX; i++)
if (arr1[i] != arr2[i])
return false;
return true;
}
// This function search for all permutations of pat[] in txt[]
void search(char *pat, char *txt)
{
int M = strlen(pat), N = strlen(txt);
// countP[]: Store count of all characters of pattern
// countTW[]: Store count of current window of text
char countP[MAX] = {0}, countTW[MAX] = {0};
for (int i = 0; i < M; i++)
{
(countP[pat[i]])++;
(countTW[txt[i]])++;
}

// Traverse through remaining characters of pattern

for (int i = M; i < N; i++)
{
// Compare counts of current window of text with
// counts of pattern[]
if (compare(countP, countTW))
cout << "Found at Index " << (i - M) << endl;
// Add current character to current window
(countTW[txt[i]])++;
// Remove the first character of previous window
countTW[txt[i-M]]--;
}
// Check for the last window in text
if (compare(countP, countTW))
cout << "Found at Index " << (N - M) << endl;
}
/* Driver program to test above function */
int main()
{
char txt[] = "BACDGABCDA";
char pat[] = "ABCD";
search(pat, txt);
return 0;
}

Output:
Found at Index 0
Found at Index 5
Found at Index 6

Pattern Searching using a Trie of all Suffixes

Problem Statement: Given a text txt[0..n-1] and a pattern pat[0..m-1], write a function search(char pat[], char txt[]) that prints all occurrences of
pat[] in txt[]. You may assume that n > m.
As discussed in the previous post, we discussed that there are two ways efficiently solve the above problem.
1) Preprocess Pattern: KMP Algorithm, Rabin Karp Algorithm, Finite Automata, Boyer Moore Algorithm.
2) Preoprocess Text: Suffix Tree
The best possible time complexity achieved by first (preprocssing pattern) is O(n) and by second (preprocessing text) is O(m) where m and n are
lengths of pattern and text respectively.
Note that the second way does the searching only in O(m) time and it is preferred when text doesnt doesnt change very frequently and there are
many search queries. We have discussed Suffix Tree (A compressed Trie of all suffixes of Text) .
Implementation of Suffix Tree may be time consuming for problems to be coded in a technical interview or programming contexts. In this post
simple implementation of a Standard Trie of all Suffixes is discussed. The implementation is close to suffix tree, the only thing is, its a simple Trie
instead of compressed Trie.
As discussed in Suffix Tree post, the idea is, every pattern that is present in text (or we can say every substring of text) must be a prefix of one of
all possible suffixes. So if we build a Trie of all suffixes, we can find the pattern in O(m) time where m is pattern length.
Building a Trie of Suffixes
1) Generate all suffixes of given text.
2) Consider all suffixes as individual words and build a trie.
Let us consider an example text banana\0? where \0? is string termination character. Following are all suffixes of banana\0?
banana\0
anana\0
nana\0
ana\0
na\0
a\0
\0

If we consider all of the above suffixes as individual words and build a Trie, we get following.

How to search a pattern in the built Trie?

Following are steps to search a pattern in the built Trie.
1) Starting from the first character of the pattern and root of the Trie, do following for every character.
..a) For the current character of pattern, if there is an edge from the current node, follow the edge.
..b) If there is no edge, print pattern doesnt exist in text and return.
2) If all characters of pattern have been processed, i.e., there is a path from root for characters of the given pattern, then print print all indexes
where pattern is present. To store indexes, we use a list with every node that stores indexes of suffixes starting at the node.
Following is C++ implementation of the above idea.
// A simple C++ implementation of substring search using trie of suffixes
#include<iostream>
#include<list>
#define MAX_CHAR 256
using namespace std;

// A Suffix Trie (A Trie of all suffixes) Node

class SuffixTrieNode
{
private:
SuffixTrieNode *children[MAX_CHAR];
list<int> *indexes;
public:
SuffixTrieNode() // Constructor
{
// Create an empty linked list for indexes of
// suffixes starting from this node
indexes = new list<int>;
// Initialize all child pointers as NULL
for (int i = 0; i < MAX_CHAR; i++)
children[i] = NULL;
}
// A recursive function to insert a suffix of the txt
// in subtree rooted with this node
void insertSuffix(string suffix, int index);
// A function to search a pattern in subtree rooted
// with this node.The function returns pointer to a linked
// list containing all indexes where pattern is present.
// The returned indexes are indexes of last characters
// of matched text.
list<int>* search(string pat);
};
// A Trie of all suffixes
class SuffixTrie
{
private:
SuffixTrieNode root;
public:
// Constructor (Builds a trie of suffies of the given text)
SuffixTrie(string txt)
{
// Consider all suffixes of given string and insert
// them into the Suffix Trie using recursive function
// insertSuffix() in SuffixTrieNode class
for (int i = 0; i < txt.length(); i++)
root.insertSuffix(txt.substr(i), i);
}
// Function to searches a pattern in this suffix trie.
void search(string pat);
};
// A recursive function to insert a suffix of the txt in
// subtree rooted with this node
void SuffixTrieNode::insertSuffix(string s, int index)
{
// Store index in linked list
indexes->push_front(index);
// If string has more characters
if (s.length() > 0)
{
// Find the first character
char cIndex = s.at(0);
// If there is no edge for this character, add a new edge
if (children[cIndex] == NULL)
children[cIndex] = new SuffixTrieNode();
// Recur for next suffix
children[cIndex]->insertSuffix(s.substr(1), index+1);
}
}
// A recursive function to search a pattern in subtree rooted with
// this node
list<int>* SuffixTrieNode::search(string s)
{
// If all characters of pattern have been processed,
if (s.length() == 0)
return indexes;
// if there is an edge from the current node of suffix trie,

// follow the edge.

if (children[s.at(0)] != NULL)
return (children[s.at(0)])->search(s.substr(1));
// If there is no edge, pattern doesnt exist in text
else return NULL;
}
/* Prints all occurrences of pat in the Suffix Trie S (built for text)*/
void SuffixTrie::search(string pat)
{
// Let us call recursive search function for root of Trie.
// We get a list of all indexes (where pat is present in text) in
// variable 'result'
list<int> *result = root.search(pat);
// Check if the list of indexes is empty or not
if (result == NULL)
cout << "Pattern not found" << endl;
else
{
list<int>::iterator i;
int patLen = pat.length();
for (i = result->begin(); i != result->end(); ++i)
cout << "Pattern found at position " << *i - patLen<< endl;
}
}
// driver program to test above functions
int main()
{
// Let us build a suffix trie for text "geeksforgeeks.org"
string txt = "geeksforgeeks.org";
SuffixTrie S(txt);
cout << "Search for 'ee'" << endl;
S.search("ee");
cout << "\nSearch for 'geek'" << endl;
S.search("geek");
cout << "\nSearch for 'quiz'" << endl;
S.search("quiz");
cout << "\nSearch for 'forgeeks'" << endl;
S.search("forgeeks");
return 0;
}

Output:
Search for 'ee'
Pattern found at position 9
Pattern found at position 1
Search for 'geek'
Pattern found at position 8
Pattern found at position 0
Search for 'quiz'
Pattern not found
Search for 'forgeeks'
Pattern found at position 5

Time Complexity of the above search function is O(m+k) where m is length of the pattern and k is the number of occurrences of pattern in text.

Manachers Algorithm Linear Time Longest Palindromic Substring Part 1

Given a string, find the longest substring which is palindrome.
if the given string is forgeeksskeegfor, the output should be geeksskeeg
if the given string is abaaba, the output should be abaaba
if the given string is abababa, the output should be abababa
if the given string is abcbabcbabcba, the output should be abcbabcba
We have already discussed Nave [O(n3)] and quadratic [O(n2)] approaches at Set 1 and Set 2.
In this article, we will talk about Manachers algorithm which finds Longest Palindromic Substring in linear time.
One way (Set 2) to find a palindrome is to start from the center of the string and compare characters in both directions one by one. If
corresponding characters on both sides (left and right of the center) match, then they will make a palindrome.
Lets consider string abababa.
Here center of the string is 4th character (with index 3) b. If we match characters in left and right of the center, all characters match and so string
abababa is a palindrome.

Here center position is not only the actual string character position but it could be the position between two characters also.
Consider string abaaba of even length. This string is palindrome around the position between 3rd and 4th characters a and a respectively.

To find Longest Palindromic Substring of a string of length N, one way is take each possible 2*N + 1 centers (the N character positions, N-1
between two character positions and 2 positions at left and right ends), do the character match in both left and right directions at each 2*N+ 1
centers and keep track of LPS. This approach takes O(N^2) time and thats what we are doing in Set 2.
Lets consider two strings abababa and abaaba as shown below:

In these two strings, left and right side of the center positions (position 7 in 1st string and position 6 in 2nd string) are symmetric. Why? Because
the whole string is palindrome around the center position.
If we need to calculate Longest Palindromic Substring at each 2*N+1 positions from left to right, then palindromes symmetric property could help
to avoid some of the unnecessary computations (i.e. character comparison). If there is a palindrome of some length L cantered at any position P,
then we may not need to compare all characters in left and right side at position P+1. We already calculated LPS at positions before P and they
can help to avoid some of the comparisons after position P.
This use of information from previous positions at a later point of time makes the Manachers algorithm linear. In Set 2, there is no reuse of previous
information and so that is quadratic.
Manachers algorithm is probably considered complex to understand, so here we will discuss it in as detailed way as we can. Some of its portions
may require multiple reading to understand it properly.
Lets look at string abababa. In 3rd figure above, 15 center positions are shown. We need to calculate length of longest palindromic string at each
of these positions.
At position 0, there is no LPS at all (no character on left side to compare), so length of LPS will be 0.
At position 1, LPS is a, so length of LPS will be 1.
At position 2, there is no LPS at all (left and right characters a and b dont match), so length of LPS will be 0.
At position 3, LPS is aba, so length of LPS will be 3.
At position 4, there is no LPS at all (left and right characters b and a dont match), so length of LPS will be 0.
At position 5, LPS is ababa, so length of LPS will be 5.

and so on
We store all these palindromic lengths in an array, say L. Then string S and LPS Length L look like below:

Similarly, LPS Length L of string abaaba will look like:

In LPS Array L:
LPS length value at odd positions (the actual character positions) will be odd and greater than or equal to 1 (1 will come from the center
character itself if nothing else matches in left and right side of it)
LPS length value at even positions (the positions between two characters, extreme left and right positions) will be even and greater than or
equal to 0 (0 will come when there is no match in left and right side)
Position and index for the string are two different things here. For a given string S of length N, indexes will be from 0 to N-1 (total N
indexes) and positions will be from 0 to 2*N (total 2*N+1 positions).
LPS length value can be interpreted in two ways, one in terms of index and second in terms of position. LPS value d at position I (L[i] = d) tells
that:
Substring from position i-d to i+d is a palindrome of length d (in terms of position)
Substring from index (i-d)/2 to [(i+d)/2 1] is a palindrome of length d (in terms of index)
e.g. in string abaaba, L[3] = 3 means substring from position 0 (3-3) to 6 (3+3) is a palindrome which is aba of length 3, it also means that
substring from index 0 [(3-3)/2] to 2 [(3+3)/2 1] is a palindrome which is aba of length 3.

Now the main task is to compute LPS array efficiently. Once this array is computed, LPS of string S will be centered at position with maximum
LPS length value.
We will see it in Part 2.

Manachers Algorithm Linear Time Longest Palindromic Substring Part 2

In Manachers Algorithm Part 1, we gone through some of the basics and LPS length array.
Here we will see how to calculate LPS length array efficiently.
To calculate LPS array efficiently, we need to understand how LPS length for any position may relate to LPS length value of any previous already
calculated position.
For string abaaba, we see following:

If we look around position 3:

LPS length value at position 2 and position 4 are same
LPS length value at position 1 and position 5 are same
We calculate LPS length values from left to right starting from position 0, so we can see if we already know LPS length values at positions 1, 2 and
3 already then we may not need to calculate LPS length at positions 4 and 5 because they are equal to LPS length values at corresponding
positions on left side of position 3.
If we look around position 6:
LPS length value at position 5 and position 7 are same
LPS length value at position 4 and position 8 are same
. and so on.
If we already know LPS length values at positions 1, 2, 3, 4, 5 and 6 already then we may not need to calculate LPS length at positions 7, 8, 9, 10
and 11 because they are equal to LPS length values at corresponding positions on left side of position 6.
For string abababa, we see following:

If we already know LPS length values at positions 1, 2, 3, 4, 5, 6 and 7 already then we may not need to calculate LPS length at positions 8, 9,
10, 11, 12 and 13 because they are equal to LPS length values at corresponding positions on left side of position 7.
Can you see why LPS length values are symmetric around positions 3, 6, 9 in string abaaba? Thats because there is a palindromic substring around
these positions. Same is the case in string abababa around position 7.
Is it always true that LPS length values around at palindromic center position are always symmetric (same)?
Answer is NO.
Look at positions 3 and 11 in string abababa. Both positions have LPS length 3. Immediate left and right positions are symmetric (with value 0),
but not the next one. Positions 1 and 5 (around position 3) are not symmetric. Similarly, positions 9 and 13 (around position 11) are not
symmetric.
At this point, we can see that if there is a palindrome in a string centered at some position, then LPS length values around the center position may
or may not be symmetric depending on some situation. If we can identify the situation when left and right positions WILL BE SYMMETRIC
around the center position, we NEED NOT calculate LPS length of the right position because it will be exactly same as LPS value of
corresponding position on the left side which is already known. And this fact where we are avoiding LPS length computation at few positions
makes Manachers Algorithm linear.
In situations when left and right positions WILL NOT BE SYMMETRIC around the center position, we compare characters in left and right side
to find palindrome, but here also algorithm tries to avoid certain no of comparisons. We will see all these scenarios soon.
Lets introduce few terms to proceed further:
(click to see it clearly)

centerPosition This is the position for which LPS length is calculated and lets say LPS length at centerPosition is d (i.e. L[centerPosition] =
d)
centerRightPosition This is the position which is right to the centerPosition and d position away from centerPosition (i.e.
centerRightPosition = centerPosition + d)
centerLeftPosition This is the position which is left to the centerPosition and d position away from centerPosition (i.e. centerLeftPosition
= centerPosition d)
currentRightPosition This is the position which is right of the centerPosition for which LPS length is not yet known and has to be
calculated
currentLeftPosition This is the position on the left side of centerPosition which corresponds to the currentRightPosition
centerPosition currentLeftPosition = currentRightPosition centerPosition
currentLeftPosition = 2* centerPosition currentRightPosition
i-left palindrome The palindrome i positions left of centerPosition, i.e. at currentLeftPosition
i-right palindrome The palindrome i positions right of centerPosition, i.e. at currentRightPosition
center palindrome The palindrome at centerPosition
When we are at centerPosition for which LPS length is known, then we also know LPS length of all positions smaller than centerPosition. Lets say
LPS length at centerPosition is d, i.e.
L[centerPosition] = d
It means that substring between positions centerPosition-d to centerPosition+d is a palindrom.
Now we proceed further to calculate LPS length of positions greater than centerPosition.
Lets say we are at currentRightPosition ( > centerPosition) where we need to find LPS length.
For this we look at LPS length of currentLeftPosition which is already calculated.
If LPS length of currentLeftPosition is less than centerRightPosition currentRightPosition, then LPS length of currentRightPosition will be equal to
LPS length of currentLeftPosition. So
L[currentRightPosition] = L[currentLeftPosition] if L[currentLeftPosition] < centerRightPosition currentRightPosition. This is Case 1.
Lets consider below scenario for string abababa:
(click to see it clearly)

We have calculated LPS length up-to position 7 where L[7] = 7, if we consider position 7 as centerPosition, then centerLeftPosition will be 0 and
centerRightPosition will be 14.
Now we need to calculate LPS length of other positions on the right of centerPosition.
For currentRightPosition = 8, currentLeftPosition is 6 and L[currentLeftPosition] = 0
Also centerRightPosition currentRightPosition = 14 8 = 6
Case 1 applies here and so L[currentRightPosition] = L[8] = 0
Case 1 applies to positions 10 and 12, so,
L[10] = L[4] = 0
L[12] = L[2] = 0
If we look at position 9, then:
currentRightPosition = 9
currentLeftPosition = 2* centerPosition currentRightPosition = 2*7 9 = 5
centerRightPosition currentRightPosition = 14 9 = 5
Here L[currentLeftPosition] = centerRightPosition currentRightPosition, so Case 1 doesnt apply here. Also note that centerRightPosition is the
extreme end position of the string. That means center palindrome is suffix of input string. In that case, L[currentRightPosition] =
L[currentLeftPosition]. This is Case 2.
Case 2 applies to positions 9, 11, 13 and 14, so:
L[9] = L[5] = 5
L[11] = L[3] = 3
L[13] = L[1] = 1
L[14] = L[0] = 0

What is really happening in Case 1 and Case 2? This is just utilizing the palindromic symmetric property and without any character match, it is
finding LPS length of new positions.
When a bigger length palindrome contains a smaller length palindrome centered at left side of its own center, then based on symmetric property,
there will be another same smaller palindrome centered on the right of bigger palindrome center. If left side smaller palindrome is not prefix of
bigger palindrome, then Case 1 applies and if it is a prefix AND bigger palindrome is suffix of the input string itself, then Case 2 applies.
The longest palindrome i places to the right of the current center (the i-right palindrome) is as long as the longest palindrome i places to
the left of the current center (the i-left palindrome) if the i-left palindrome is completely contained in the longest palindrome around the
current center (the center palindrome) and the i-left palindrome is not a prefix of the center palindrome (Case 1) or (i.e. when i-left
palindrome is a prefix of center palindrome) if the center palindrome is a suffix of the entire string (Case 2).
In Case 1 and Case 2, i-right palindrome cant expand more than corresponding i-left palindrome (can you visualize why it cant expand more?),
and so LPS length of i-right palindrome is exactly same as LPS length of i-left palindrome.
Here both i-left and i-right palindromes are completely contained in center palindrome (i.e. L[currentLeftPosition] <= centerRightPosition
currentRightPosition)
Now if i-left palindrome is not a prefix of center palindrome (L[currentLeftPosition] < centerRightPosition currentRightPosition), that means that ileft palindrome was not able to expand up-to position centerLeftPosition.
If we look at following with centerPosition = 11, then
(click to see it clearly)

centerLeftPosition would be 11 9 = 2, and centerRightPosition would be 11 + 9 = 20

If we take currentRightPosition = 15, its currentLeftPosition is 7. Case 1 applies here and so L[15] = 3. i-left palindrome at position 7 is bab
which is completely contained in center palindrome at position 11 (which is dbabcbabd). We can see that i-right palindrome (at position 15) cant
expand more than i-left palindrome (at position 7).
If there was a possibility of expansion, i-left palindrome could have expanded itself more already. But there is no such possibility as i-left
palindrome is prefix of center palindrome. So due to symmetry property, i-right palindrome will be exactly same as i-left palindrome and it cant
expand more. This makes L[currentRightPosition] = L[currentLeftPosition] in Case 1.
Now if we consider centerPosition = 19, then centerLeftPosition = 12 and centerRightPosition = 26
If we take currentRightPosition = 23, its currentLeftPosition is 15. Case 2 applies here and so L[23] = 3. i-left palindrome at position 15 is bab
which is completely contained in center palindrome at position 19 (which is babdbab). In Case 2, where i-left palindrome is prefix of center
palindrome, i-right palindrome cant expand more than length of i-left palindrome because center palindrome is suffix of input string so there are no
more character left to compare and expand. This makes L[currentRightPosition] = L[currentLeftPosition] in Case 2.
Case 1: L[currentRightPosition] = L[currentLeftPosition] applies when:
i-left palindrome is completely contained in center palindrome
i-left palindrome is NOT a prefix of center palindrome
Both above conditions are satisfied when
L[currentLeftPosition] < centerRightPosition currentRightPosition
Case 2: L[currentRightPosition] = L[currentLeftPosition] applies when:
i-left palindrome is prefix of center palindrome (means completely contained also)
center palindrome is suffix of input string
Above conditions are satisfied when
L[currentLeftPosition] = centerRightPosition currentRightPosition (For 1st condition) AND
centerRightPosition = 2*N where N is input string length N (For 2nd condition).
Case 3: L[currentRightPosition] > = L[currentLeftPosition] applies when:
i-left palindrome is prefix of center palindrome (and so i-left palindrome is completely contained in center palindrome)
center palindrome is NOT suffix of input string
Above conditions are satisfied when
L[currentLeftPosition] = centerRightPosition currentRightPosition (For 1st condition) AND
centerRightPosition < 2*N where N is input string length N (For 2nd condition).
In this case, there is a possibility of i-right palindrome expansion and so length of i-right palindrome is at least as long as length of i-left palindrome.
Case 4: L[currentRightPosition] > = centerRightPosition currentRightPosition applies when:

i-left palindrome is NOT completely contained in center palindrome

Above condition is satisfied when
L[currentLeftPosition] > centerRightPosition currentRightPosition
In this case, length of i-right palindrome is at least as long (centerRightPosition currentRightPosition) and there is a possibility of i-right palindrome
expansion.
In following figure,
(click to see it clearly)

If we take center position 7, then Case 3 applies at currentRightPosition 11 because i-left palindrome at currentLeftPosition 3 is a prefix of center
palindrome and i-right palindrome is not suffix of input string, so here L[11] = 9, which is greater than i-left palindrome length L[3] = 3. In the case,
it is guaranteed that L[11] will be at least 3, and so in implementation, we 1st set L[11] = 3 and then we try to expand it by comparing characters
in left and right side starting from distance 4 (As up-to distance 3, it is already known that characters will match).
If we take center position 11, then Case 4 applies at currentRightPosition 15 because L[currentLeftPosition] = L[7] = 7 > centerRightPosition
currentRightPosition = 20 15 = 5. In the case, it is guaranteed that L[15] will be at least 5, and so in implementation, we 1st set L[15] = 5 and then
we try to expand it by comparing characters in left and right side starting from distance 5 (As up-to distance 5, it is already known that characters
will match).
Now one point left to discuss is, when we work at one center position and compute LPS lengths for different rightPositions, how to know that
what would be next center position. We change centerPosition to currentRightPosition if palindrome centered at currentRightPosition expands
beyond centerRightPosition.
Here we have seen four different cases on how LPS length of a position will depend on a previous positions LPS length.
In Part 3, we have discussed code implementation of it and also we have looked at these four cases in a different way and implement that too.

Manachers Algorithm Linear Time Longest Palindromic Substring Part 3

In Manachers Algorithm Part 1 and Part 2, we gone through some of the basics, understood LPS length array and how to calculate it efficiently
based on four cases. Here we will implement the same.
We have seen that there are no new character comparison needed in case 1 and case 2. In case 3 and case 4, necessary new comparison are
needed.
In following figure,
(click to see it clearly)

If at all we need a comparison, we will only compare actual characters, which are at odd positions like 1, 3, 5, 7, etc.
Even positions do not represent a character in string, so no comparison will be preformed for even positions.
If two characters at different odd positions match, then they will increase LPS length by 2.
There are many ways to implement this depending on how even and odd positions are handled. One way would be to create a new string 1st
where we insert some unique character (say #, $ etc) in all even positions and then run algorithm on that (to avoid different way of even and odd
position handling). Other way could be to work on given string itself but here even and odd positions should be handled appropriately.
Here we will start with given string itself. When there is a need of expansion and character comparison required, we will expand in left and right
positions one by one. When odd position is found, comparison will be done and LPS Length will be incremented by ONE. When even position is
found, no comparison done and LPS Length will be incremented by ONE (So overall, one odd and one even positions on both left and right side
will increase LPS Length by TWO).

C/C++
// A C program to implement Manachers Algorithm
#include <stdio.h>
#include <string.h>
char text[100];
void findLongestPalindromicString()
{
int N = strlen(text);
if(N == 0)
return;
N = 2*N + 1; //Position count
int L[N]; //LPS Length Array
L[0] = 0;
L[1] = 1;
int C = 1; //centerPosition
int R = 2; //centerRightPosition
int i = 0; //currentRightPosition
int iMirror; //currentLeftPosition
int expand = -1;
int diff = -1;
int maxLPSLength = 0;
int maxLPSCenterPosition = 0;
int start = -1;
int end = -1;
//Uncomment it to print LPS Length array
//printf("%d %d ", L[0], L[1]);
for (i = 2; i < N; i++)
{
//get currentLeftPosition iMirror for currentRightPosition i
iMirror = 2*C-i;
//Reset expand - means no expansion required
expand = 0;
diff = R - i;
//If currentRightPosition i is within centerRightPosition R
if(diff > 0)
{
if(L[iMirror] < diff) // Case 1
L[i] = L[iMirror];
else if(L[iMirror] == diff && i == N-1) // Case 2
L[i] = L[iMirror];
else if(L[iMirror] == diff && i < N-1) // Case 3
{
L[i] = L[iMirror];
expand = 1; // expansion required
}
else if(L[iMirror] > diff) // Case 4
{

L[i] = diff;
expand = 1; // expansion required
}
}
else
{
L[i] = 0;
expand = 1; // expansion required
}
if (expand == 1)
{
//Attempt to expand palindrome centered at currentRightPosition i
//Here for odd positions, we compare characters and
//if match then increment LPS Length by ONE
//If even position, we just increment LPS by ONE without
//any character comparison
while (((i + L[i]) < N && (i - L[i]) > 0) &&
( ((i + L[i] + 1) % 2 == 0) ||
(text[(i + L[i] + 1)/2] == text[(i-L[i]-1)/2] )))
{
L[i]++;
}
}
if(L[i] > maxLPSLength) // Track maxLPSLength
{
maxLPSLength = L[i];
maxLPSCenterPosition = i;
}
// If palindrome centered at currentRightPosition i
// expand beyond centerRightPosition R,
// adjust centerPosition C based on expanded palindrome.
if (i + L[i] > R)
{
C = i;
R = i + L[i];
}
//Uncomment it to print LPS Length array
//printf("%d ", L[i]);
}
//printf("\n");
start = (maxLPSCenterPosition - maxLPSLength)/2;
end = start + maxLPSLength - 1;
//printf("start: %d end: %d\n", start, end);
printf("LPS of string is %s : ", text);
for(i=start; i<=end; i++)
printf("%c", text[i]);
printf("\n");
}
int main(int argc, char *argv[])
{
strcpy(text, "babcbabcbaccba");
findLongestPalindromicString();
strcpy(text, "abaaba");
findLongestPalindromicString();
strcpy(text, "abababa");
findLongestPalindromicString();
strcpy(text, "abcbabcbabcba");
findLongestPalindromicString();
strcpy(text, "forgeeksskeegfor");
findLongestPalindromicString();
strcpy(text, "caba");
findLongestPalindromicString();
strcpy(text, "abacdfgdcaba");
findLongestPalindromicString();
strcpy(text, "abacdfgdcabba");
findLongestPalindromicString();
strcpy(text, "abacdedcaba");

findLongestPalindromicString();
return 0;
}

Python
# Python program to implement Manacher's Algorithm
def findLongestPalindromicString(text):
N = len(text)
if N == 0:
return
N = 2*N+1
# Position count
L = [0] * N
L[0] = 0
L[1] = 1
C = 1
# centerPosition
R = 2
# centerRightPosition
i = 0
# currentRightPosition
iMirror = 0
# currentLeftPosition
maxLPSLength = 0
maxLPSCenterPosition = 0
start = -1
end = -1
diff = -1
# Uncomment it to print LPS Length array
# printf("%d %d ", L[0], L[1]);
for i in xrange(2,N):
# get currentLeftPosition iMirror for currentRightPosition i
iMirror = 2*C-i
L[i] = 0
diff = R - i
# If currentRightPosition i is within centerRightPosition R
if diff > 0:
L[i] = min(L[iMirror], diff)
# Attempt to expand palindrome centered at currentRightPosition i
# Here for odd positions, we compare characters and
# if match then increment LPS Length by ONE
# If even position, we just increment LPS by ONE without
# any character comparison
try:
while ((i+L[i]) < N and (i-L[i]) > 0) and \
(((i+L[i]+1) % 2 == 0) or \
(text[(i+L[i]+1)/2] == text[(i-L[i]-1)/2])):
L[i]+=1
except Exception as e:
pass
if L[i] > maxLPSLength:
maxLPSLength = L[i]
maxLPSCenterPosition = i

# Track maxLPSLength

# If palindrome centered at currentRightPosition i

# expand beyond centerRightPosition R,
# adjust centerPosition C based on expanded palindrome.
if i + L[i] > R:
C = i
R = i + L[i]
# Uncomment it to print LPS Length array
# printf("%d ", L[i]);
start = (maxLPSCenterPosition - maxLPSLength) / 2
end = start + maxLPSLength - 1
print "LPS of string is " + text + " : ",
print text[start:end+1],
print "\n",
# Driver program
text1 = "babcbabcbaccba"
findLongestPalindromicString(text1)
text2 = "abaaba"
findLongestPalindromicString(text2)
text3 = "abababa"
findLongestPalindromicString(text3)

text4 = "abcbabcbabcba"
findLongestPalindromicString(text4)
text5 = "forgeeksskeegfor"
findLongestPalindromicString(text5)
text6 = "caba"
findLongestPalindromicString(text6)
text7 = "abacdfgdcaba"
findLongestPalindromicString(text7)
text8 = "abacdfgdcabba"
findLongestPalindromicString(text8)
text9 = "abacdedcaba"
findLongestPalindromicString(text9)
# This code is contributed by BHAVYA JAIN

Output:
LPS
LPS
LPS
LPS
LPS
LPS
LPS
LPS
LPS

of
of
of
of
of
of
of
of
of

string
string
string
string
string
string
string
string
string

is
is
is
is
is
is
is
is
is

babcbabcbaccba : abcbabcba
abaaba : abaaba
abababa : abababa
abcbabcbabcba : abcbabcbabcba
forgeeksskeegfor : geeksskeeg
caba : aba
abacdfgdcaba : aba
abacdfgdcabba : abba
abacdedcaba : abacdedcaba

This is the implementation based on the four cases discussed in Part 2. In Part 4, we have discussed a different way to look at these four cases and
few other approaches.

Manachers Algorithm Linear Time Longest Palindromic Substring Part 4

In Manachers Algorithm Part 1 and Part 2, we gone through some of the basics, understood LPS length array and how to calculate it efficiently
based on four cases. In Part 3, we implemented the same.
Here we will review the four cases again and try to see it differently and implement the same.
All four cases depends on LPS length value at currentLeftPosition (L[iMirror]) and value of (centerRightPosition currentRightPosition), i.e. (R i).
These two information are know before which helps us to reuse previous available information and avoid unnecessary character comparison.
(click to see it clearly)

If we look at all four cases, we will see that we 1st set minimum of L[iMirror] and R-i to L[i] and then we try to expand the palindrome
in whichever case it can expand.
Above observation may look more intuitive, easier to understand and implement, given that one understands LPS length array, position, index,
symmetry property etc.

C/C++
// A C program to implement Manachers Algorithm
#include <stdio.h>
#include <string.h>
char text[100];
int min(int a, int b)
{
int res = a;
if(b < a)
res = b;
return res;
}
void findLongestPalindromicString()
{
int N = strlen(text);
if(N == 0)
return;
N = 2*N + 1; //Position count
int L[N]; //LPS Length Array
L[0] = 0;
L[1] = 1;
int C = 1; //centerPosition
int R = 2; //centerRightPosition
int i = 0; //currentRightPosition
int iMirror; //currentLeftPosition
int maxLPSLength = 0;
int maxLPSCenterPosition = 0;
int start = -1;
int end = -1;
int diff = -1;
//Uncomment it to print LPS Length array
//printf("%d %d ", L[0], L[1]);
for (i = 2; i < N; i++)
{
//get currentLeftPosition iMirror for currentRightPosition i
iMirror = 2*C-i;
L[i] = 0;
diff = R - i;
//If currentRightPosition i is within centerRightPosition R
if(diff > 0)
L[i] = min(L[iMirror], diff);
//Attempt to expand palindrome centered at currentRightPosition i
//Here for odd positions, we compare characters and
//if match then increment LPS Length by ONE
//If even position, we just increment LPS by ONE without
//any character comparison
while ( ((i + L[i]) < N && (i - L[i]) > 0) &&
( ((i + L[i] + 1) % 2 == 0) ||
(text[(i + L[i] + 1)/2] == text[(i - L[i] - 1)/2] )))
{
L[i]++;

}
if(L[i] > maxLPSLength) // Track maxLPSLength
{
maxLPSLength = L[i];
maxLPSCenterPosition = i;
}
//If palindrome centered at currentRightPosition i
//expand beyond centerRightPosition R,
//adjust centerPosition C based on expanded palindrome.
if (i + L[i] > R)
{
C = i;
R = i + L[i];
}
//Uncomment it to print LPS Length array
//printf("%d ", L[i]);
}
//printf("\n");
start = (maxLPSCenterPosition - maxLPSLength)/2;
end = start + maxLPSLength - 1;
printf("LPS of string is %s : ", text);
for(i=start; i<=end; i++)
printf("%c", text[i]);
printf("\n");
}
int main(int argc, char *argv[])
{
strcpy(text, "babcbabcbaccba");
findLongestPalindromicString();
strcpy(text, "abaaba");
findLongestPalindromicString();
strcpy(text, "abababa");
findLongestPalindromicString();
strcpy(text, "abcbabcbabcba");
findLongestPalindromicString();
strcpy(text, "forgeeksskeegfor");
findLongestPalindromicString();
strcpy(text, "caba");
findLongestPalindromicString();
strcpy(text, "abacdfgdcaba");
findLongestPalindromicString();
strcpy(text, "abacdfgdcabba");
findLongestPalindromicString();
strcpy(text, "abacdedcaba");
findLongestPalindromicString();
return 0;
}

Python
# Python program to implement Manacher's Algorithm
def findLongestPalindromicString(text):
N = len(text)
if N == 0:
return
N = 2*N+1
# Position count
L = [0] * N
L[0] = 0
L[1] = 1
C = 1
# centerPosition
R = 2
# centerRightPosition
i = 0
# currentRightPosition
iMirror = 0
# currentLeftPosition
maxLPSLength = 0
maxLPSCenterPosition = 0
start = -1

end = -1
diff = -1
# Uncomment it to print LPS Length array
# printf("%d %d ", L[0], L[1]);
for i in xrange(2,N):
# get currentLeftPosition iMirror for currentRightPosition i
iMirror = 2*C-i
L[i] = 0
diff = R - i
# If currentRightPosition i is within centerRightPosition R
if diff > 0:
L[i] = min(L[iMirror], diff)
# Attempt to expand palindrome centered at currentRightPosition i
# Here for odd positions, we compare characters and
# if match then increment LPS Length by ONE
# If even position, we just increment LPS by ONE without
# any character comparison
try:
while ((i + L[i]) < N and (i - L[i]) > 0) and \
(((i + L[i] + 1) % 2 == 0) or \
(text[(i + L[i] + 1) / 2] == text[(i - L[i] - 1) / 2])):
L[i]+=1
except Exception as e:
pass
if L[i] > maxLPSLength:
maxLPSLength = L[i]
maxLPSCenterPosition = i

# Track maxLPSLength

# If palindrome centered at currentRightPosition i

# expand beyond centerRightPosition R,
# adjust centerPosition C based on expanded palindrome.
if i + L[i] > R:
C = i
R = i + L[i]
# Uncomment it to print LPS Length array
# printf("%d ", L[i]);
start = (maxLPSCenterPosition - maxLPSLength) / 2
end = start + maxLPSLength - 1
print "LPS of string is " + text + " : ",
print text[start:end+1],
print "\n",
# Driver program
text1 = "babcbabcbaccba"
findLongestPalindromicString(text1)
text2 = "abaaba"
findLongestPalindromicString(text2)
text3 = "abababa"
findLongestPalindromicString(text3)
text4 = "abcbabcbabcba"
findLongestPalindromicString(text4)
text5 = "forgeeksskeegfor"
findLongestPalindromicString(text5)
text6 = "caba"
findLongestPalindromicString(text6)
text7 = "abacdfgdcaba"
findLongestPalindromicString(text7)
text8 = "abacdfgdcabba"
findLongestPalindromicString(text8)
text9 = "abacdedcaba"
findLongestPalindromicString(text9)
# This code is contributed by BHAVYA JAIN

Output:
LPS of string is babcbabcbaccba : abcbabcba
LPS of string is abaaba : abaaba

LPS
LPS
LPS
LPS
LPS
LPS
LPS

of
of
of
of
of
of
of

string
string
string
string
string
string
string

is
is
is
is
is
is
is

abababa : abababa
abcbabcbabcba : abcbabcbabcba
forgeeksskeegfor : geeksskeeg
caba : aba
abacdfgdcaba : aba
abacdfgdcabba : abba
abacdedcaba : abacdedcaba

Other Approaches
We have discussed two approaches here. One in Part 3 and other in current article. In both approaches, we worked on given string. Here we had
to handle even and odd positions differently while comparing characters for expansion (because even positions do not represent any character in
string).
To avoid this different handling of even and odd positions, we need to make even positions also to represent some character (actually all even
positions should represent SAME character because they MUST match while character comparison). One way to do this is to set some character
at all even positions by modifying given string or create a new copy of given string. For example, if input string is abcb, new string should be
#a#b#c#b# if we add # as unique character at even positions.
The two approaches discussed already can be modified a bit to work on modified string where different handling of even and odd positions will not
be needed.
We may also add two DIFFERENT characters (not yet used anywhere in string at even and odd positions) at start and end of string as sentinels to
avoid bound check. With these changes string abcb will look like ^#a#b#c#b#$ where ^ and $ are sentinels.
This implementation may look cleaner with the cost of more memory.
We are not implementing these here as its a simple change in given implementations.
Implementation of approach discussed in current article on a modified string can be found at Longest Palindromic Substring Part II and a Java
Translation of the same by Princeton.

Longest Even Length Substring such that Sum of First and Second Half is same
Given a string str of digits, find length of the longest substring of str, such that the length of the substring is 2k digits and sum of left k digits is equal
to the sum of right k digits.
Examples:
Input: str = "123123"
Output: 6
The complete string is of even length and sum of first and second
half digits is same
Input: str = "1538023"
Output: 4
The longest substring with same first and second half sum is "5380"

Simple Solution [ O(n3) ]

A Simple Solution is to check every substring of even length. The following is C based implementation of simple approach.
// A simple C based program to find length of longest even length
// substring with same sum of digits in left and right
#include<stdio.h>
#include<string.h>
int findLength(char *str)
{
int n = strlen(str);
int maxlen =0; // Initialize result
// Choose starting point of every substring
for (int i=0; i<n; i++)
{
// Choose ending point of even length substring
for (int j =i+1; j<n; j += 2)
{
int length = j-i+1;//Find length of current substr
// Calculate left & right sums for current substr
int leftsum = 0, rightsum =0;
for (int k =0; k<length/2; k++)
{
leftsum += (str[i+k]-'0');
rightsum += (str[i+k+length/2]-'0');
}
// Update result if needed
if (leftsum == rightsum && maxlen < length)
maxlen = length;
}
}
return maxlen;
}
// Driver program to test above function
int main(void)
{
char str[] = "1538023";
printf("Length of the substring is %d", findLength(str));
return 0;
}

Output:
Length of the substring is 4

Dynamic Programming [ O(n2) and O(n2) extra space]

The above solution can be optimized to work in O(n2) using Dynamic Programming. The idea is to build a 2D table that stores sums of
substrings. The following is C based implementation of Dynamic Programming approach.
// A C based program that uses Dynamic Programming to find length of the
// longest even substring with same sum of digits in left and right half
#include <stdio.h>
#include <string.h>
int findLength(char *str)

{
int n = strlen(str);
int maxlen = 0; // Initialize result
// A 2D table where sum[i][j] stores sum of digits
// from str[i] to str[j]. Only filled entries are
// the entries where j >= i
int sum[n][n];
// Fill the diagonal values for sunstrings of length 1
for (int i =0; i<n; i++)
sum[i][i] = str[i]-'0';
// Fill entries for substrings of length 2 to n
for (int len=2; len<=n; len++)
{
// Pick i and j for current substring
for (int i=0; i<n-len+1; i++)
{
int j = i+len-1;
int k = len/2;
// Calculate value of sum[i][j]
sum[i][j] = sum[i][j-k] + sum[j-k+1][j];
// Update result if 'len' is even, left and right
// sums are same and len is more than maxlen
if (len%2 == 0 && sum[i][j-k] == sum[(j-k+1)][j]
&& len > maxlen)
maxlen = len;
}
}
return maxlen;
}
// Driver program to test above function
int main(void)
{
char str[] = "153803";
printf("Length of the substring is %d", findLength(str));
return 0;
}

Output:
Length of the substring is 4

Time complexity of the above solution is O(n2), but it requires O(n2) extra space.

[A O(n2) and O(n) extra space solution]

The idea is to use a single dimensional array to store cumulative sum.
// A O(n^2) time and O(n) extra space solution
#include<bits/stdc++.h>
using namespace std;
int findLength(string str, int n)
{
int sum[n+1]; // To store cumulative sum from first digit to nth digit
sum[0] = 0;
/* Store cumulative sum of digits from first to last digit */
for (int i = 1; i <= n; i++)
sum[i] = (sum[i-1] + str[i-1] - '0'); /* convert chars to int */
int ans = 0; // initialize result
/* consider all even length substrings one by one */
for (int len = 2; len <= n; len += 2)
{
for (int i = 0; i <= n-len; i++)
{
int j = i + len - 1;
/* Sum of first and second half is same than update ans */
if (sum[i+len/2] - sum[i] == sum[i+len] - sum[i+len/2])
ans = max(ans, len);

}
}
return ans;
}
// Driver program to test above function
int main()
{
string str = "123123";
cout << "Length of the substring is " << findLength(str, str.length());
return 0;
}

Output:
Length of the substring is 6

Thanks to Gaurav Ahirwar for suggesting this method.

[A O(n2) time and O(1) extra space solution]

The idea is to consider all possible mid points (of even length substrings) and keep expanding on both sides to get and update optimal length as the
sum of two sides become equal.
Below is C++ implementation of the above idea.
// A O(n^2) time and O(1) extra space solution
#include<bits/stdc++.h>
using namespace std;
int findLength(string str, int n)
{
int ans = 0; // Initialize result
// Consider all possible midpoints one by one
for (int i = 0; i <= n-2; i++)
{
/* For current midpoint 'i', keep expanding substring on
both sides, if sum of both sides becomes equal update
ans */
int l = i, r = i + 1;
/* initialize left and right sum */
int lsum = 0, rsum = 0;
/* move on both sides till indexes go out of bounds */
while (r < n && l >= 0)
{
lsum += str[l] - '0';
rsum += str[r] - '0';
if (lsum == rsum)
ans = max(ans, r-l+1);
l--;
r++;
}
}
return ans;
}
// Driver program to test above function
int main()
{
string str = "123123";
cout << "Length of the substring is " << findLength(str, str.length());
return 0;
}

Output:
Length of the substring is 6

Thanks to Gaurav Ahirwar for suggesting this method.

Print all possible strings that can be made by placing spaces

Given a string you need to print all possible strings that can be made by placing spaces (zero or one) in between them.
Input: str[] = "ABC"
Output: ABC
AB C
A BC
A B C

Source: Amazon Interview Experience | Set 158, Round 1 ,Q 1.

The idea is to use recursion and create a buffer that one by one contains all output strings having spaces. We keep updating buffer in every
recursive call. If the length of given string is n our updated string can have maximum length of n + (n-1) i.e. 2n-1. So we create buffer size of 2n
(one extra character for string termination).
We leave 1st character as it is, starting from the 2nd character, we can either fill a space or a character. Thus one can write a recursive function
like below.

C/C++
// C++ program to print permutations of a given string with spaces.
#include <iostream>
#include <cstring>
using namespace std;
/* Function recursively prints the strings having space pattern.
i and j are indices in 'str[]' and 'buff[]' respectively */
void printPatternUtil(char str[], char buff[], int i, int j, int n)
{
if (i==n)
{
buff[j] = '\0';
cout << buff << endl;
return;
}
// Either put the character
buff[j] = str[i];
printPatternUtil(str, buff, i+1, j+1, n);
// Or put a space followed by next character
buff[j] = ' ';
buff[j+1] = str[i];
printPatternUtil(str, buff, i+1, j+2, n);
}
// This function creates buf[] to store individual output string and uses
// printPatternUtil() to print all permutations.
void printPattern(char *str)
{
int n = strlen(str);
// Buffer to hold the string containing spaces
char buf[2*n]; // 2n-1 characters and 1 string terminator
// Copy the first character as it is, since it will be always
// at first position
buf[0] = str[0];
printPatternUtil(str, buf, 1, 1, n);
}
// Driver program to test above functions
int main()
{
char *str = "ABCD";
printPattern(str);
return 0;
}

Python
# Python program to print permutations of a given string with
# spaces.
# Utility function

def toString(List):
s = ""
for x in List:
if x == '\0':
break
s += x
return s
# Function recursively prints the strings having space pattern.
# i and j are indices in 'str[]' and 'buff[]' respectively
def printPatternUtil(string, buff, i, j, n):
if i == n:
buff[j] = '\0'
print toString(buff)
return
# Either put the character
buff[j] = string[i]
printPatternUtil(string, buff, i+1, j+1, n)
# Or put a space followed by next character
buff[j] = ' '
buff[j+1] = string[i]
printPatternUtil(string, buff, i+1, j+2, n)
# This function creates buf[] to store individual output string
# and uses printPatternUtil() to print all permutations.
def printPattern(string):
n = len(string)
# Buffer to hold the string containing spaces
buff = [0] * (2*n) # 2n-1 characters and 1 string terminator
# Copy the first character as it is, since it will be always
# at first position
buff[0] = string[0]
printPatternUtil(string, buff, 1, 1, n)
# Driver program
string = "ABCD"
printPattern(string)
# This code is contributed by BHAVYA JAIN

ABCD
ABC D
AB CD
AB C D
A BCD
A BC D
A B CD
A B C D

Time Complexity: Since number of Gaps are n-1, there are total 2^(n-1) patters each having length ranging from n to 2n-1. Thus overall
complexity would be O(n*(2^n)).

Write a program to print all permutations of a given string

A permutation, also called an arrangement number or order, is a rearrangement of the elements of an ordered list S into a one-to-one
correspondence with S itself. A string of length n has n! permutation.
Source: Mathword(http://mathworld.wolfram.com/Permutation.html)
Below are the permutations of string ABC.
ABC, ACB, BAC, BCA, CAB, CBA
Here is a solution using backtracking.

C/C++
// C program to print all permutations with duplicates allowed
#include <stdio.h>
#include <string.h>
/* Function to swap values at two pointers */
void swap(char *x, char *y)
{
char temp;
temp = *x;
*x = *y;
*y = temp;
}
/* Function to print permutations of string
This function takes three parameters:
1. String
2. Starting index of the string
3. Ending index of the string. */
void permute(char *a, int l, int r)
{
int i;
if (l == r)
printf("%s\n", a);
else
{
for (i = l; i <= r; i++)
{
swap((a+l), (a+i));
permute(a, l+1, r);
swap((a+l), (a+i)); //backtrack
}
}
}
/* Driver program to test above functions */
int main()
{
char str[] = "ABC";
int n = strlen(str);
permute(str, 0, n-1);
return 0;
}

Python
# Python program to print all permutations with
# duplicates allowed

# Function to swap values

def swap(a,l,r):
t = a[l]
a[l] = a[r]
a[r] = t
return a
def toList(string):
List = []
for x in string:
List.append(x)
return List
def toString(List):
return ''.join(List)
# Function to print permutations of string
# This function takes three parameters:
# 1. String
# 2. Starting index of the string
# 3. Ending index of the string.
def permute(a, l, r):
if l==r:
print toString(a)
else:
for i in xrange(l,r+1):
a = swap(a,l,i)
permute(a, l+1, r)
a = swap(a,l,i) # backtrack
# Driver program to test the above function
string = "ABC"
n = len(string)
a = toList(string)
permute(a, 0, n-1)
# This code is contributed by Bhavya Jain

ABC
ACB
BAC
BCA
CBA
CAB

Algorithm Paradigm: Backtracking

Time Complexity: O(n*n!)

Backtracking | Set 1 (The Knights tour problem)

Backtracking is an algorithmic paradigm that tries different solutions until finds a solution that works. Problems which are typically solved using
backtracking technique have following property in common. These problems can only be solved by trying every possible configuration and each
configuration is tried only once. A Naive solution for these problems is to try all configurations and output a configuration that follows given
problem constraints. Backtracking works in incremental way and is an optimization over the Naive solution where all possible configurations are
generated and tried.
For example, consider the following Knights Tour problem.
The knight is placed on the first block of an empty board and, moving according to the rules of chess, must visit each square exactly
once.
Let us first discuss the Naive algorithm for this problem and then the Backtracking algorithm.
Naive Algorithm for Knights tour
The Naive Algorithm is to generate all tours one by one and check if the generated tour satisfies the constraints.
while there are untried tours
{
generate the next tour
if this tour covers all squares
{
print this path;
}
}

Backtracking works in an incremental way to attack problems. Typically, we start from an empty solution vector and one by one add items
(Meaning of item varies from problem to problem. In context of Knights tour problem, an item is a Knights move). When we add an item, we
check if adding the current item violates the problem constraint, if it does then we remove the item and try other alternatives. If none of the
alternatives work out then we go to previous stage and remove the item added in the previous stage. If we reach the initial stage back then we say
that no solution exists. If adding an item doesnt violate constraints then we recursively add items one by one. If the solution vector becomes
complete then we print the solution.
Backtracking Algorithm for Knights tour
Following is the Backtracking algorithm for Knights tour problem.
If all squares are visited
print the solution
Else
a) Add one of the next moves to solution vector and recursively
check if this move leads to a solution. (A Knight can make maximum
eight moves. We choose one of the 8 moves in this step).
b) If the move chosen in the above step doesn't lead to a solution
then remove this move from the solution vector and try other
alternative moves.
c) If none of the alternatives work then return false (Returning false
will remove the previously added item in recursion and if false is
returned by the initial call of recursion then "no solution exists" )

Following are implementations for Knights tour problem. It prints one of the possible solutions in 2D matrix form. Basically, the output is a 2D 8*8
matrix with numbers from 0 to 63 and these numbers show steps made by Knight.

C
// C program for Knight Tour problem
#include<stdio.h>
#define N 8
int solveKTUtil(int x, int y, int movei, int sol[N][N],
int xMove[], int yMove[]);
/* A utility function to check if i,j are valid indexes
for N*N chessboard */
bool isSafe(int x, int y, int sol[N][N])
{
return ( x >= 0 && x < N && y >= 0 &&
y < N && sol[x][y] == -1);
}
/* A utility function to print solution matrix sol[N][N] */
void printSolution(int sol[N][N])
{
for (int x = 0; x < N; x++)
{

for (int y = 0; y < N; y++)

printf(" %2d ", sol[x][y]);
printf("\n");
}
}
/* This function solves the Knight Tour problem using
Backtracking. This function mainly uses solveKTUtil()
to solve the problem. It returns false if no complete
tour is possible, otherwise return true and prints the
tour.
Please note that there may be more than one solutions,
this function prints one of the feasible solutions. */
bool solveKT()
{
int sol[N][N];
/* Initialization of solution matrix */
for (int x = 0; x < N; x++)
for (int y = 0; y < N; y++)
sol[x][y] = -1;
/* xMove[] and yMove[] define next move of Knight.
xMove[] is for next value of x coordinate
yMove[] is for next value of y coordinate */
int xMove[8] = { 2, 1, -1, -2, -2, -1, 1, 2 };
int yMove[8] = { 1, 2, 2, 1, -1, -2, -2, -1 };
// Since the Knight is initially at the first block
sol[0][0] = 0;
/* Start from 0,0 and explore all tours using
solveKTUtil() */
if (solveKTUtil(0, 0, 1, sol, xMove, yMove) == false)
{
printf("Solution does not exist");
return false;
}
else
printSolution(sol);
return true;
}
/* A recursive utility function to solve Knight Tour
problem */
int solveKTUtil(int x, int y, int movei, int sol[N][N],
int xMove[N], int yMove[N])
{
int k, next_x, next_y;
if (movei == N*N)
return true;
/* Try all next moves from the current coordinate x, y */
for (k = 0; k < 8; k++)
{
next_x = x + xMove[k];
next_y = y + yMove[k];
if (isSafe(next_x, next_y, sol))
{
sol[next_x][next_y] = movei;
if (solveKTUtil(next_x, next_y, movei+1, sol,
xMove, yMove) == true)
return true;
else
sol[next_x][next_y] = -1;// backtracking
}
}
return false;
}
/* Driver program to test above functions */
int main()
{
solveKT();
return 0;
}

Java

// Java program for Knight Tour problem

class KnightTour {
static int N = 8;
/* A utility function to check if i,j are
valid indexes for N*N chessboard */
static boolean isSafe(int x, int y, int sol[][]) {
return (x >= 0 && x < N && y >= 0 &&
y < N && sol[x][y] == -1);
}
/* A utility function to print solution
matrix sol[N][N] */
static void printSolution(int sol[][]) {
for (int x = 0; x < N; x++) {
for (int y = 0; y < N; y++)
System.out.print(sol[x][y] + " ");
System.out.println();
}
}
/* This function solves the Knight Tour problem
using Backtracking. This function mainly
uses solveKTUtil() to solve the problem. It
returns false if no complete tour is possible,
otherwise return true and prints the tour.
Please note that there may be more than one
solutions, this function prints one of the
feasible solutions. */
static boolean solveKT() {
int sol[][] = new int[8][8];
/* Initialization of solution matrix */
for (int x = 0; x < N; x++)
for (int y = 0; y < N; y++)
sol[x][y] = -1;
/* xMove[] and yMove[] define next move of Knight.
xMove[] is for next value of x coordinate
yMove[] is for next value of y coordinate */
int xMove[] = {2, 1, -1, -2, -2, -1, 1, 2};
int yMove[] = {1, 2, 2, 1, -1, -2, -2, -1};
// Since the Knight is initially at the first block
sol[0][0] = 0;
/* Start from 0,0 and explore all tours using
solveKTUtil() */
if (!solveKTUtil(0, 0, 1, sol, xMove, yMove)) {
System.out.println("Solution does not exist");
return false;
} else
printSolution(sol);
return true;
}
/* A recursive utility function to solve Knight
Tour problem */
static boolean solveKTUtil(int x, int y, int movei,
int sol[][], int xMove[],
int yMove[]) {
int k, next_x, next_y;
if (movei == N * N)
return true;
/* Try all next moves from the current coordinate
x, y */
for (k = 0; k < 8; k++) {
next_x = x + xMove[k];
next_y = y + yMove[k];
if (isSafe(next_x, next_y, sol)) {
sol[next_x][next_y] = movei;
if (solveKTUtil(next_x, next_y, movei + 1,
sol, xMove, yMove))
return true;
else
sol[next_x][next_y] = -1;// backtracking
}
}

return false;
}
/* Driver program to test above functions */
public static void main(String args[]) {
solveKT();
}
}
// This code is contributed by Abhishek Shankhadhar

0
37
58
35
42
47
56
51

59
34
1
48
57
50
43
46

38
31
36
41
2
45
52
55

33
60
39
26
49
54
3
44

30
9
32
61
40
25
22
53

17
62
27
10
23
20
13
4

8
29
18
15
6
11
24
21

63
16
7
28
19
14
5
12

Note that Backtracking is not the best solution for the Knights tour problem. See this for other better solutions. The purpose of this post is to
explain Backtracking with an example.
References:
http://see.stanford.edu/materials/icspacs106b/H19-RecBacktrackExamples.pdf
http://www.cis.upenn.edu/~matuszek/cit594-2009/Lectures/35-backtracking.ppt
http://mathworld.wolfram.com/KnightsTour.html
http://en.wikipedia.org/wiki/Knight%27s_tour

Backtracking | Set 2 (Rat in a Maze)

We have discussed Backtracking and Knights tour problem in Set 1. Let us discuss Rat in a Maze as another example problem that can be solved
using Backtracking.
A Maze is given as N*N binary matrix of blocks where source block is the upper left most block i.e., maze[0][0] and destination block is lower
rightmost block i.e., maze[N-1][N-1]. A rat starts from source and has to reach destination. The rat can move only in two directions: forward and
down.
In the maze matrix, 0 means the block is dead end and 1 means the block can be used in the path from source to destination. Note that this is a
simple version of the typical Maze problem. For example, a more complex version can be that the rat can move in 4 directions and a more
complex version can be with limited number of moves.
Following is an example maze.
Gray blocks are dead ends (value = 0).

Following is binary matrix representation of the above maze.

{1,
{1,
{0,
{1,

0,
1,
1,
1,

0,
0,
0,
1,

0}
1}
0}
1}

Following is maze with highlighted solution path.

Following is the solution matrix (output of program) for the above input matrx.
{1, 0, 0,
{1, 1, 0,
{0, 1, 0,
{0, 1, 1,
All enteries in solution

0}
0}
0}
1}
path are marked as 1.

Naive Algorithm
The Naive Algorithm is to generate all paths from source to destination and one by one check if the generated path satisfies the constraints.
while there are untried paths
{
generate the next path
if this path has all blocks as 1
{
print this path;
}
}

Backtrackng Algorithm
If destination is reached
print the solution matrix
Else
a) Mark current cell in solution matrix as 1.
b) Move forward in horizontal direction and recursively check if this
move leads to a solution.
c) If the move chosen in the above step doesn't lead to a solution
then move down and check if this move leads to a solution.

d) If none of the above solutions work then unmark this cell as 0

(BACKTRACK) and return false.

Implementation of Backtracking solution

C/C++
/* C/C++ program to solve Rat in a Maze problem using
backtracking */
#include<stdio.h>
// Maze size
#define N 4
bool solveMazeUtil(int maze[N][N], int x, int y, int sol[N][N]);
/* A utility function to print solution matrix sol[N][N] */
void printSolution(int sol[N][N])
{
for (int i = 0; i < N; i++)
{
for (int j = 0; j < N; j++)
printf(" %d ", sol[i][j]);
printf("\n");
}
}
/* A utility function to check if x,y is valid index for N*N maze */
bool isSafe(int maze[N][N], int x, int y)
{
// if (x,y outside maze) return false
if(x >= 0 && x < N && y >= 0 && y < N && maze[x][y] == 1)
return true;
return false;
}
/* This function solves the Maze problem using Backtracking. It mainly
uses solveMazeUtil() to solve the problem. It returns false if no
path is possible, otherwise return true and prints the path in the
form of 1s. Please note that there may be more than one solutions,
this function prints one of the feasible solutions.*/
bool solveMaze(int maze[N][N])
{
int sol[N][N] = { {0, 0, 0, 0},
{0, 0, 0, 0},
{0, 0, 0, 0},
{0, 0, 0, 0}
};
if(solveMazeUtil(maze, 0, 0, sol) == false)
{
printf("Solution doesn't exist");
return false;
}
printSolution(sol);
return true;
}
/* A recursive utility function to solve Maze problem */
bool solveMazeUtil(int maze[N][N], int x, int y, int sol[N][N])
{
// if (x,y is goal) return true
if(x == N-1 && y == N-1)
{
sol[x][y] = 1;
return true;
}
// Check if maze[x][y] is valid
if(isSafe(maze, x, y) == true)
{
// mark x,y as part of solution path
sol[x][y] = 1;
/* Move forward in x direction */
if (solveMazeUtil(maze, x+1, y, sol) == true)
return true;

/* If moving in x direction doesn't give solution then

Move down in y direction */
if (solveMazeUtil(maze, x, y+1, sol) == true)
return true;
/* If none of the above movements work then BACKTRACK:
unmark x,y as part of solution path */
sol[x][y] = 0;
return false;
}
return false;
}
// driver program to test above function
int main()
{
int maze[N][N] = { {1, 0, 0, 0},
{1, 1, 0, 1},
{0, 1, 0, 0},
{1, 1, 1, 1}
};
solveMaze(maze);
return 0;
}

Java
/* Java program to solve Rat in a Maze problem using
backtracking */
public class RatMaze
{
final int N = 4;
/* A utility function to print solution matrix
sol[N][N] */
void printSolution(int sol[][])
{
for (int i = 0; i < N; i++)
{
for (int j = 0; j < N; j++)
System.out.print(" " + sol[i][j] +
" ");
System.out.println();
}
}
/* A utility function to check if x,y is valid
index for N*N maze */
boolean isSafe(int maze[][], int x, int y)
{
// if (x,y outside maze) return false
return (x >= 0 && x < N && y >= 0 &&
y < N && maze[x][y] == 1);
}
/* This function solves the Maze problem using
Backtracking. It mainly uses solveMazeUtil()
to solve the problem. It returns false if no
path is possible, otherwise return true and
prints the path in the form of 1s. Please note
that there may be more than one solutions, this
function prints one of the feasible solutions.*/
boolean solveMaze(int maze[][])
{
int sol[][] = {{0, 0, 0, 0},
{0, 0, 0, 0},
{0, 0, 0, 0},
{0, 0, 0, 0}
};
if (solveMazeUtil(maze, 0, 0, sol) == false)
{
System.out.print("Solution doesn't exist");
return false;
}
printSolution(sol);

return true;
}
/* A recursive utility function to solve Maze
problem */
boolean solveMazeUtil(int maze[][], int x, int y,
int sol[][])
{
// if (x,y is goal) return true
if (x == N - 1 && y == N - 1)
{
sol[x][y] = 1;
return true;
}
// Check if maze[x][y] is valid
if (isSafe(maze, x, y) == true)
{
// mark x,y as part of solution path
sol[x][y] = 1;
/* Move forward in x direction */
if (solveMazeUtil(maze, x + 1, y, sol))
return true;
/* If moving in x direction doesn't give
solution then Move down in y direction */
if (solveMazeUtil(maze, x, y + 1, sol))
return true;
/* If none of the above movements work then
BACKTRACK: unmark x,y as part of solution
path */
sol[x][y] = 0;
return false;
}
return false;
}
public static void main(String args[])
{
RatMaze rat = new RatMaze();
int maze[][] = {{1, 0, 0, 0},
{1, 1, 0, 1},
{0, 1, 0, 0},
{1, 1, 1, 1}
};
rat.solveMaze(maze);
}
}
// This code is contributed by Abhishek Shankhadhar

1
1
0
0

0
1
1
1

0
0
0
1

0
0
0
1

Backtracking | Set 3 (N Queen Problem)

We have discussed Knights tour and Rat in a Maze problems in Set 1 and Set 2 respectively. Let us discuss N Queen as another example
problem that can be solved using Backtracking.
The N Queen is the problem of placing N chess queens on an NN chessboard so that no two queens attack each other. For example, following is
a solution for 4 Queen problem.

The expected output is a binary matrix which has 1s for the blocks where queens are placed. For example following is the output matrix for above
4 queen solution.
{
{
{
{

0,
0,
1,
0,

1,
0,
0,
0,

0,
0,
0,
1,

0}
1}
0}
0}

Naive Algorithm
Generate all possible configurations of queens on board and print a configuration that satisfies the given constraints.
while there are untried conflagrations
{
generate the next configuration
if queens don't attack in this configuration then
{
print this configuration;
}
}

Backtracking Algorithm
The idea is to place queens one by one in different columns, starting from the leftmost column. When we place a queen in a column, we check for
clashes with already placed queens. In the current column, if we find a row for which there is no clash, we mark this row and column as part of the
solution. If we do not find such a row due to clashes then we backtrack and return false.
1) Start in the leftmost column
2) If all queens are placed
return true
3) Try all rows in the current column. Do following for every tried row.
a) If the queen can be placed safely in this row then mark this [row,
column] as part of the solution and recursively check if placing
queen here leads to a solution.
b) If placing queen in [row, column] leads to a solution then return
true.
c) If placing queen doesn't lead to a solution then umark this [row,
column] (Backtrack) and go to step (a) to try other rows.
3) If all rows have been tried and nothing worked, return false to trigger
backtracking.

Implementation of Backtracking solution

C/C++
/* C/C++ program to solve N Queen Problem using
backtracking */
#define N 4
#include<stdio.h>
/* A utility function to print solution */
void printSolution(int board[N][N])
{
for (int i = 0; i < N; i++)
{
for (int j = 0; j < N; j++)
printf(" %d ", board[i][j]);
printf("\n");

}
}
/* A utility function to check if a queen can
be placed on board[row][col]. Note that this
function is called when "col" queens are
already placed in columns from 0 to col -1.
So we need to check only left side for
attacking queens */
bool isSafe(int board[N][N], int row, int col)
{
int i, j;
/* Check this row on left side */
for (i = 0; i < col; i++)
if (board[row][i])
return false;
/* Check upper diagonal on left side */
for (i=row, j=col; i>=0 && j>=0; i--, j--)
if (board[i][j])
return false;
/* Check lower diagonal on left side */
for (i=row, j=col; j>=0 && i<N; i++, j--)
if (board[i][j])
return false;
return true;
}
/* A recursive utility function to solve N
Queen problem */
bool solveNQUtil(int board[N][N], int col)
{
/* base case: If all queens are placed
then return true */
if (col >= N)
return true;
/* Consider this column and try placing
this queen in all rows one by one */
for (int i = 0; i < N; i++)
{
/* Check if queen can be placed on
board[i][col] */
if ( isSafe(board, i, col) )
{
/* Place this queen in board[i][col] */
board[i][col] = 1;
/* recur to place rest of the queens */
if ( solveNQUtil(board, col + 1) )
return true;
/* If placing queen in board[i][col]
doesn't lead to a solution, then
remove queen from board[i][col] */
board[i][col] = 0; // BACKTRACK
}
}
/* If queen can not be place in any row in
this colum col then return false */
return false;
}
/* This function solves the N Queen problem using
Backtracking. It mainly uses solveNQUtil() to
solve the problem. It returns false if queens
cannot be placed, otherwise return true and
prints placement of queens in the form of 1s.
Please note that there may be more than one
solutions, this function prints one of the
feasible solutions.*/
bool solveNQ()
{
int board[N][N] = { {0, 0, 0, 0},
{0, 0, 0, 0},
{0, 0, 0, 0},
{0, 0, 0, 0}

};
if ( solveNQUtil(board, 0) == false )
{
printf("Solution does not exist");
return false;
}
printSolution(board);
return true;
}
// driver program to test above function
int main()
{
solveNQ();
return 0;
}

Java
/* Java program to solve N Queen Problem using
backtracking */
public class NQueenProblem
{
final int N = 4;
/* A utility function to print solution */
void printSolution(int board[][])
{
for (int i = 0; i < N; i++)
{
for (int j = 0; j < N; j++)
System.out.print(" " + board[i][j]
+ " ");
System.out.println();
}
}
/* A utility function to check if a queen can
be placed on board[row][col]. Note that this
function is called when "col" queens are already
placeed in columns from 0 to col -1. So we need
to check only left side for attacking queens */
boolean isSafe(int board[][], int row, int col)
{
int i, j;
/* Check this row on left side */
for (i = 0; i < col; i++)
if (board[row][i] == 1)
return false;
/* Check upper diagonal on left side */
for (i=row, j=col; i>=0 && j>=0; i--, j--)
if (board[i][j] == 1)
return false;
/* Check lower diagonal on left side */
for (i=row, j=col; j>=0 && i<N; i++, j--)
if (board[i][j] == 1)
return false;
return true;
}
/* A recursive utility function to solve N
Queen problem */
boolean solveNQUtil(int board[][], int col)
{
/* base case: If all queens are placed
then return true */
if (col >= N)
return true;
/* Consider this column and try placing
this queen in all rows one by one */
for (int i = 0; i < N; i++)
{
/* Check if queen can be placed on

board[i][col] */
if (isSafe(board, i, col))
{
/* Place this queen in board[i][col] */
board[i][col] = 1;
/* recur to place rest of the queens */
if (solveNQUtil(board, col + 1) == true)
return true;
/* If placing queen in board[i][col]
doesn't lead to a solution then
remove queen from board[i][col] */
board[i][col] = 0; // BACKTRACK
}
}
/* If queen can not be place in any row in
this colum col, then return false */
return false;
}
/* This function solves the N Queen problem using
Backtracking. It mainly uses solveNQUtil() to
solve the problem. It returns false if queens
cannot be placed, otherwise return true and
prints placement of queens in the form of 1s.
Please note that there may be more than one
solutions, this function prints one of the
feasible solutions.*/
boolean solveNQ()
{
int board[][] = {{0, 0, 0, 0},
{0, 0, 0, 0},
{0, 0, 0, 0},
{0, 0, 0, 0}
};
if (solveNQUtil(board, 0) == false)
{
System.out.print("Solution does not exist");
return false;
}
printSolution(board);
return true;
}
// driver program to test above function
public static void main(String args[])
{
NQueenProblem Queen = new NQueenProblem();
Queen.solveNQ();
}
}
// This code is contributed by Abhishek Shankhadhar

0
1
0
0

0
0
0
1

1
0
0
0

0
0
1
0

Sources:
http://see.stanford.edu/materials/icspacs106b/H19-RecBacktrackExamples.pdf
http://en.literateprograms.org/Eight_queens_puzzle_%28C%29
http://en.wikipedia.org/wiki/Eight_queens_puzzle

Backtracking | Set 4 (Subset Sum)

Subset sum problem is to find subset of elements that are selected from a given setwhose sum adds up to a given number K. We are considering
the set contains non-negative values. It is assumed that the input set is unique (no duplicates are presented).
ExhaustiveSearch Algorithm for Subset Sum
One way to find subsets that sum to K is to consider all possible subsets. A power set contains all those subsets generated from a given set. The
size of such a power set is 2N.
Backtracking Algorithm for Subset Sum
Using exhaustive search we consider all subsets irrespective of whether they satisfy givenconstraintsor not. Backtracking can be used to make a
systematic consideration of the elements to be selected.
Assume given set of 4 elements, say w[1] w[4]. Tree diagrams can be used to design backtracking algorithms. The following tree diagramdepicts
approach of generating variable sized tuple.

In the above tree, a node represents function call and a branch represents candidate element. The root node contains 4 children. Inother words,
root considers every element of the set as different branch. The next levelsub-treescorrespondto the subsets that includes the parent node. The
branches at each level represent tuple element to be considered. For example, if we are at level 1, tuple_vector[1] can take any value of four
branches generated. If we are at level 2 of left most node,tuple_vector[2] can take any value of three branches generated, and so on
For example the left most child of root generates all those subsets that include w[1]. Similarly the second child of root generates all those subsets
that includes w[2] and excludes w[1].
As we go down along depth of tree we add elements so far, and if the added sum issatisfyingexplicit constraints, we will continue to generate child
nodes further. Whenever the constraints are not met, we stop further generation of sub-treesof that node, and backtrack to previous node to
explore the nodes not yet explored. In many scenarios, it saves considerableamountof processing time.
The tree should trigger a clue to implement the backtracking algorithm (try yourself). It prints all those subsets whose sum add up to given
number.We need to explore the nodes along the breadth and depth of the tree. Generating nodes along breadth is controlled by loop and nodes
along the depth are generated using recursion (post order traversal). Pseudo code given below,
if(subset is satisfying the constraint)
print the subset
exclude the current element and consider next element
else
generate the nodes of present level along breadth of tree and
recur for next levels

Following is C implementation of subset sum using variable size tuple vector. Note that the following program explores all possibilities similar to
exhaustive search. It is to demonstrate how backtracking can be used. See next code to verify, how we can optimize the backtracking solution.
#include <stdio.h>
#include <stdlib.h>
#define ARRAYSIZE(a) (sizeof(a))/(sizeof(a[0]))
static int total_nodes;
// prints subset found

void printSubset(int A[], int size)

{
for(int i = 0; i < size; i++)
{
printf("%*d", 5, A[i]);
}
printf("\n");
}
// inputs
// s
- set vector
// t
- tuplet vector
// s_size
- set size
// t_size
- tuplet size so far
// sum
- sum so far
// ite
- nodes count
// target_sum - sum to be found
void subset_sum(int s[], int t[],
int s_size, int t_size,
int sum, int ite,
int const target_sum)
{
total_nodes++;
if( target_sum == sum )
{
// We found subset
printSubset(t, t_size);
// Exclude previously added item and consider next candidate
subset_sum(s, t, s_size, t_size-1, sum - s[ite], ite + 1, target_sum);
return;
}
else
{
// generate nodes along the breadth
for( int i = ite; i < s_size; i++ )
{
t[t_size] = s[i];
// consider next level node (along depth)
subset_sum(s, t, s_size, t_size + 1, sum + s[i], i + 1, target_sum);
}
}
}
// Wrapper to print subsets that sum to target_sum
// input is weights vector and target_sum
void generateSubsets(int s[], int size, int target_sum)
{
int *tuplet_vector = (int *)malloc(size * sizeof(int));
subset_sum(s, tuplet_vector, size, 0, 0, 0, target_sum);
free(tuplet_vector);
}
int main()
{
int weights[] = {10, 7, 5, 18, 12, 20, 15};
int size = ARRAYSIZE(weights);
generateSubsets(weights, size, 35);
printf("Nodes generated %d\n", total_nodes);
return 0;
}

The power of backtracking appears when we combine explicit and implicit constraints, and we stop generating nodes when these checks fail. We
can improve the above algorithm by strengthening the constraint checks and presorting the data. By sorting the initial array, we need not to
consider rest of the array, once the sum so far is greater than target number. We can backtrack and check other possibilities.
Similarly, assume the array is presorted and we found one subset. We can generate next node excluding the present node only when inclusion of
next nodesatisfiesthe constraints. Given below is optimized implementation (it prunes the subtree if it is not satisfying contraints).
#include <stdio.h>
#include <stdlib.h>
#define ARRAYSIZE(a) (sizeof(a))/(sizeof(a[0]))
static int total_nodes;
// prints subset found

void printSubset(int A[], int size)

{
for(int i = 0; i < size; i++)
{
printf("%*d", 5, A[i]);
}
printf("\n");
}
// qsort compare function
int comparator(const void *pLhs, const void *pRhs)
{
int *lhs = (int *)pLhs;
int *rhs = (int *)pRhs;
return *lhs > *rhs;
}
// inputs
// s
- set vector
// t
- tuplet vector
// s_size
- set size
// t_size
- tuplet size so far
// sum
- sum so far
// ite
- nodes count
// target_sum - sum to be found
void subset_sum(int s[], int t[],
int s_size, int t_size,
int sum, int ite,
int const target_sum)
{
total_nodes++;
if( target_sum == sum )
{
// We found sum
printSubset(t, t_size);
// constraint check
if( ite + 1 < s_size && sum - s[ite] + s[ite+1] <= target_sum )
{
// Exclude previous added item and consider next candidate
subset_sum(s, t, s_size, t_size-1, sum - s[ite], ite + 1, target_sum);
}
return;
}
else
{
// constraint check
if( ite < s_size && sum + s[ite] <= target_sum )
{
// generate nodes along the breadth
for( int i = ite; i < s_size; i++ )
{
t[t_size] = s[i];
if( sum + s[i] <= target_sum )
{
// consider next level node (along depth)
subset_sum(s, t, s_size, t_size + 1, sum + s[i], i + 1, target_sum);
}
}
}
}
}
// Wrapper that prints subsets that sum to target_sum
void generateSubsets(int s[], int size, int target_sum)
{
int *tuplet_vector = (int *)malloc(size * sizeof(int));
int total = 0;
// sort the set
qsort(s, size, sizeof(int), &comparator);
for( int i = 0; i < size; i++ )
{
total += s[i];
}

if( s[0] <= target_sum && total >= target_sum )

{
subset_sum(s, tuplet_vector, size, 0, 0, 0, target_sum);
}
free(tuplet_vector);
}
int main()
{
int weights[] = {15, 22, 14, 26, 32, 9, 16, 8};
int target = 53;
int size = ARRAYSIZE(weights);
generateSubsets(weights, size, target);
printf("Nodes generated %d\n", total_nodes);
return 0;
}

As another approach, we can generate the tree in fixed size tupleanalogsto binary pattern. We will kill thesub-treeswhen the constraints are not
satisfied.

Backtracking | Set 5 (m Coloring Problem)

Given an undirected graph and a number m, determine if the graph can be colored with at most m colors such that no two adjacent vertices of the
graph are colored with same color. Here coloring of a graph means assignment of colors to all vertices.
Input:
1) A 2D array graph[V][V] where V is the number of vertices in graph and graph[V][V] is adjacency matrix representation of the graph. A value
graph[i][j] is 1 if there is a direct edge from i to j, otherwise graph[i][j] is 0.
2) An integer m which is maximum number of colors that can be used.
Output:
An array color[V] that should have numbers from 1 to m. color[i] should represent the color assigned to the ith vertex. The code should also return
false if the graph cannot be colored with m colors.
Following is an example graph (from Wiki page ) that can be colored with 3 colors.

Naive Algorithm
Generate all possible configurations of colors and print a configuration that satisfies the given constraints.
while there are untried conflagrations
{
generate the next configuration
if no adjacent vertices are colored with same color
{
print this configuration;
}
}

There will be V^m configurations of colors.

Backtracking Algorithm
The idea is to assign colors one by one to different vertices, starting from the vertex 0. Before assigning a color, we check for safety by considering
already assigned colors to the adjacent vertices. If we find a color assignment which is safe, we mark the color assignment as part of solution. If we
do not a find color due to clashes then we backtrack and return false.
Implementation of Backtracking solution

C/C++
#include<stdio.h>
// Number of vertices in the graph
#define V 4
void printSolution(int color[]);
/* A utility function to check if the current color assignment
is safe for vertex v */
bool isSafe (int v, bool graph[V][V], int color[], int c)
{
for (int i = 0; i < V; i++)
if (graph[v][i] && c == color[i])
return false;
return true;
}
/* A recursive utility function to solve m coloring problem */
bool graphColoringUtil(bool graph[V][V], int m, int color[], int v)
{
/* base case: If all vertices are assigned a color then
return true */
if (v == V)

return true;
/* Consider this vertex v and try different colors */
for (int c = 1; c <= m; c++)
{
/* Check if assignment of color c to v is fine*/
if (isSafe(v, graph, color, c))
{
color[v] = c;
/* recur to assign colors to rest of the vertices */
if (graphColoringUtil (graph, m, color, v+1) == true)
return true;
/* If assigning color c doesn't lead to a solution
then remove it */
color[v] = 0;
}
}
/* If no color can be assigned to this vertex then return false */
return false;
}
/* This function solves the m Coloring problem using Backtracking.
It mainly uses graphColoringUtil() to solve the problem. It returns
false if the m colors cannot be assigned, otherwise return true and
prints assignments of colors to all vertices. Please note that there
may be more than one solutions, this function prints one of the
feasible solutions.*/
bool graphColoring(bool graph[V][V], int m)
{
// Initialize all color values as 0. This initialization is needed
// correct functioning of isSafe()
int *color = new int[V];
for (int i = 0; i < V; i++)
color[i] = 0;
// Call graphColoringUtil() for vertex 0
if (graphColoringUtil(graph, m, color, 0) == false)
{
printf("Solution does not exist");
return false;
}
// Print the solution
printSolution(color);
return true;
}
/* A utility function to print solution */
void printSolution(int color[])
{
printf("Solution Exists:"
" Following are the assigned colors \n");
for (int i = 0; i < V; i++)
printf(" %d ", color[i]);
printf("\n");
}
// driver program to test above function
int main()
{
/* Create following graph and test whether it is 3 colorable
(3)---(2)
| / |
| / |
| / |
(0)---(1)
*/
bool graph[V][V] = {{0, 1, 1, 1},
{1, 0, 1, 0},
{1, 1, 0, 1},
{1, 0, 1, 0},
};
int m = 3; // Number of colors
graphColoring (graph, m);
return 0;
}

Java
/* Java program for solution of M Coloring problem
using backtracking */
public class mColoringProblem {
final int V = 4;
int color[];
/* A utility function to check if the current
color assignment is safe for vertex v */
boolean isSafe(int v, int graph[][], int color[],
int c)
{
for (int i = 0; i < V; i++)
if (graph[v][i] == 1 && c == color[i])
return false;
return true;
}
/* A recursive utility function to solve m
coloring problem */
boolean graphColoringUtil(int graph[][], int m,
int color[], int v)
{
/* base case: If all vertices are assigned
a color then return true */
if (v == V)
return true;
/* Consider this vertex v and try different
colors */
for (int c = 1; c <= m; c++)
{
/* Check if assignment of color c to v
is fine*/
if (isSafe(v, graph, color, c))
{
color[v] = c;
/* recur to assign colors to rest
of the vertices */
if (graphColoringUtil(graph, m,
color, v + 1))
return true;
/* If assigning color c doesn't lead
to a solution then remove it */
color[v] = 0;
}
}
/* If no color can be assigned to this vertex
then return false */
return false;
}
/* This function solves the m Coloring problem using
Backtracking. It mainly uses graphColoringUtil()
to solve the problem. It returns false if the m
colors cannot be assigned, otherwise return true
and prints assignments of colors to all vertices.
Please note that there may be more than one
solutions, this function prints one of the
feasible solutions.*/
boolean graphColoring(int graph[][], int m)
{
// Initialize all color values as 0. This
// initialization is needed correct functioning
// of isSafe()
color = new int[V];
for (int i = 0; i < V; i++)
color[i] = 0;
// Call graphColoringUtil() for vertex 0
if (!graphColoringUtil(graph, m, color, 0))
{
System.out.println("Solution does not exist");
return false;
}

// Print the solution

printSolution(color);
return true;
}
/* A utility function to print solution */
void printSolution(int color[])
{
System.out.println("Solution Exists: Following" +
" are the assigned colors");
for (int i = 0; i < V; i++)
System.out.print(" " + color[i] + " ");
System.out.println();
}
// driver program to test above function
public static void main(String args[])
{
mColoringProblem Coloring = new mColoringProblem();
/* Create following graph and test whether it is
3 colorable
(3)---(2)
| / |
| / |
| / |
(0)---(1)
*/
int graph[][] = {{0, 1, 1, 1},
{1, 0, 1, 0},
{1, 1, 0, 1},
{1, 0, 1, 0},
};
int m = 3; // Number of colors
Coloring.graphColoring(graph, m);
}
}
// This code is contributed by Abhishek Shankhadhar

Solution Exists: Following are the assigned colors

1 2 3 2

References:
http://en.wikipedia.org/wiki/Graph_coloring

Backtracking | Set 6 (Hamiltonian Cycle)

Hamiltonian Path in an undirected graph is a path that visits each vertex exactly once. A Hamiltonian cycle (or Hamiltonian circuit) is a Hamiltonian
Path such that there is an edge (in graph) from the last vertex to the first vertex of the Hamiltonian Path. Determine whether a given graph contains
Hamiltonian Cycle or not. If it contains, then print the path. Following are the input and output of the required function.
Input:
A 2D array graph[V][V] where V is the number of vertices in graph and graph[V][V] is adjacency matrix representation of the graph. A value
graph[i][j] is 1 if there is a direct edge from i to j, otherwise graph[i][j] is 0.
Output:
An array path[V] that should contain the Hamiltonian Path. path[i] should represent the ith vertex in the Hamiltonian Path. The code should also
return false if there is no Hamiltonian Cycle in the graph.
For example, a Hamiltonian Cycle in the following graph is {0, 1, 2, 4, 3, 0}. There are more Hamiltonian Cycles in the graph like {0, 3, 4, 2, 1,
0}
(0)--(1)--(2)
| / \ |
| / \ |
| /
\ |
(3)-------(4)

And the following graph doesnt contain any Hamiltonian Cycle.

(0)--(1)--(2)
| / \ |
| / \ |
| /
\ |
(3)
(4)

Naive Algorithm
Generate all possible configurations of vertices and print a configuration that satisfies the given constraints. There will be n! (n factorial)
configurations.
while there are untried conflagrations
{
generate the next configuration
if ( there are edges between two consecutive vertices of this
configuration and there is an edge from the last vertex to
the first ).
{
print this configuration;
break;
}
}

Backtracking Algorithm
Create an empty path array and add vertex 0 to it. Add other vertices, starting from the vertex 1. Before adding a vertex, check for whether it is
adjacent to the previously added vertex and not already added. If we find such a vertex, we add the vertex as part of the solution. If we do not
find a vertex then we return false.
Implementation of Backtracking solution
Following are implementations of the Backtracking solution.

C/C++
/* C/C++ program for solution of Hamiltonian Cycle problem
using backtracking */
#include<stdio.h>
// Number of vertices in the graph
#define V 5
void printSolution(int path[]);
/* A utility function to check if the vertex v can be added at
index 'pos' in the Hamiltonian Cycle constructed so far (stored
in 'path[]') */
bool isSafe(int v, bool graph[V][V], int path[], int pos)
{
/* Check if this vertex is an adjacent vertex of the previously
added vertex. */
if (graph [ path[pos-1] ][ v ] == 0)

return false;
/* Check if the vertex has already been included.
This step can be optimized by creating an array of size V */
for (int i = 0; i < pos; i++)
if (path[i] == v)
return false;
return true;
}
/* A recursive utility function to solve hamiltonian cycle problem */
bool hamCycleUtil(bool graph[V][V], int path[], int pos)
{
/* base case: If all vertices are included in Hamiltonian Cycle */
if (pos == V)
{
// And if there is an edge from the last included vertex to the
// first vertex
if ( graph[ path[pos-1] ][ path[0] ] == 1 )
return true;
else
return false;
}
// Try different vertices as a next candidate in Hamiltonian Cycle.
// We don't try for 0 as we included 0 as starting point in in hamCycle()
for (int v = 1; v < V; v++)
{
/* Check if this vertex can be added to Hamiltonian Cycle */
if (isSafe(v, graph, path, pos))
{
path[pos] = v;
/* recur to construct rest of the path */
if (hamCycleUtil (graph, path, pos+1) == true)
return true;
/* If adding vertex v doesn't lead to a solution,
then remove it */
path[pos] = -1;
}
}
/* If no vertex can be added to Hamiltonian Cycle constructed so far,
then return false */
return false;
}
/* This function solves the Hamiltonian Cycle problem using Backtracking.
It mainly uses hamCycleUtil() to solve the problem. It returns false
if there is no Hamiltonian Cycle possible, otherwise return true and
prints the path. Please note that there may be more than one solutions,
this function prints one of the feasible solutions. */
bool hamCycle(bool graph[V][V])
{
int *path = new int[V];
for (int i = 0; i < V; i++)
path[i] = -1;
/* Let us put vertex 0 as the first vertex in the path. If there is
a Hamiltonian Cycle, then the path can be started from any point
of the cycle as the graph is undirected */
path[0] = 0;
if ( hamCycleUtil(graph, path, 1) == false )
{
printf("\nSolution does not exist");
return false;
}
printSolution(path);
return true;
}
/* A utility function to print solution */
void printSolution(int path[])
{
printf ("Solution Exists:"
" Following is one Hamiltonian Cycle \n");
for (int i = 0; i < V; i++)
printf(" %d ", path[i]);

// Let us print the first vertex again to show the complete cycle
printf(" %d ", path[0]);
printf("\n");
}
// driver program to test above function
int main()
{
/* Let us create the following graph
(0)--(1)--(2)
| / \ |
| / \ |
| /
\ |
(3)-------(4)
*/
bool graph1[V][V] = {{0, 1, 0, 1, 0},
{1, 0, 1, 1, 1},
{0, 1, 0, 0, 1},
{1, 1, 0, 0, 1},
{0, 1, 1, 1, 0},
};
// Print the solution
hamCycle(graph1);
/* Let us create the following graph
(0)--(1)--(2)
| / \ |
| / \ |
| /
\ |
(3)
(4)
*/
bool graph2[V][V] = {{0, 1, 0, 1, 0},
{1, 0, 1, 1, 1},
{0, 1, 0, 0, 1},
{1, 1, 0, 0, 0},
{0, 1, 1, 0, 0},
};
// Print the solution
hamCycle(graph2);
return 0;
}

Java
/* Java program for solution of Hamiltonian Cycle problem
using backtracking */
class HamiltonianCycle
{
final int V = 5;
int path[];
/* A utility function to check if the vertex v can be
added at index 'pos'in the Hamiltonian Cycle
constructed so far (stored in 'path[]') */
boolean isSafe(int v, int graph[][], int path[], int pos)
{
/* Check if this vertex is an adjacent vertex of
the previously added vertex. */
if (graph[path[pos - 1]][v] == 0)
return false;
/* Check if the vertex has already been included.
This step can be optimized by creating an array
of size V */
for (int i = 0; i < pos; i++)
if (path[i] == v)
return false;
return true;
}
/* A recursive utility function to solve hamiltonian
cycle problem */
boolean hamCycleUtil(int graph[][], int path[], int pos)
{
/* base case: If all vertices are included in
Hamiltonian Cycle */
if (pos == V)

{
// And if there is an edge from the last included
// vertex to the first vertex
if (graph[path[pos - 1]][path[0]] == 1)
return true;
else
return false;
}
// Try different vertices as a next candidate in
// Hamiltonian Cycle. We don't try for 0 as we
// included 0 as starting point in in hamCycle()
for (int v = 1; v < V; v++)
{
/* Check if this vertex can be added to Hamiltonian
Cycle */
if (isSafe(v, graph, path, pos))
{
path[pos] = v;
/* recur to construct rest of the path */
if (hamCycleUtil(graph, path, pos + 1) == true)
return true;
/* If adding vertex v doesn't lead to a solution,
then remove it */
path[pos] = -1;
}
}
/* If no vertex can be added to Hamiltonian Cycle
constructed so far, then return false */
return false;
}
/* This function solves the Hamiltonian Cycle problem using
Backtracking. It mainly uses hamCycleUtil() to solve the
problem. It returns false if there is no Hamiltonian Cycle
possible, otherwise return true and prints the path.
Please note that there may be more than one solutions,
this function prints one of the feasible solutions. */
int hamCycle(int graph[][])
{
path = new int[V];
for (int i = 0; i < V; i++)
path[i] = -1;
/* Let us put vertex 0 as the first vertex in the path.
If there is a Hamiltonian Cycle, then the path can be
started from any point of the cycle as the graph is
undirected */
path[0] = 0;
if (hamCycleUtil(graph, path, 1) == false)
{
System.out.println("\nSolution does not exist");
return 0;
}
printSolution(path);
return 1;
}
/* A utility function to print solution */
void printSolution(int path[])
{
System.out.println("Solution Exists: Following" +
" is one Hamiltonian Cycle");
for (int i = 0; i < V; i++)
System.out.print(" " + path[i] + " ");
// Let us print the first vertex again to show the
// complete cycle
System.out.println(" " + path[0] + " ");
}
// driver program to test above function
public static void main(String args[])
{
HamiltonianCycle hamiltonian =
new HamiltonianCycle();
/* Let us create the following graph

(0)--(1)--(2)
| / \ |
| / \ |
| /
\ |
(3)-------(4)
int graph1[][] =
{1, 0, 1, 1,
{0, 1, 0, 0,
{1, 1, 0, 0,
{0, 1, 1, 1,
};

*/
{{0, 1, 0, 1, 0},
1},
1},
1},
0},

// Print the solution

hamiltonian.hamCycle(graph1);
/* Let us create
(0)--(1)--(2)
| / \ |
| / \ |
| /
\ |
(3)
(4)
int graph2[][] =
{1, 0, 1, 1,
{0, 1, 0, 0,
{1, 1, 0, 0,
{0, 1, 1, 0,
};

the following graph

*/
{{0, 1, 0, 1, 0},
1},
1},
0},
0},

// Print the solution

hamiltonian.hamCycle(graph2);
}
}
// This code is contributed by Abhishek Shankhadhar

Solution Exists: Following is one Hamiltonian Cycle

0 1 2 4 3 0
Solution does not exist

Backtracking | Set 7 (Sudoku)

Given a partially filled 99 2D array grid[9][9], the goal is to assign digits (from 1 to 9) to the empty cells so that every row, column, and subgrid of
size 33 contains exactly one instance of the digits from 1 to 9.

Naive Algorithm
The Naive Algorithm is to generate all possible configurations of numbers from 1 to 9 to fill the empty cells. Try every configuration one by one
until the correct configuration is found.
Backtracking Algorithm
Like all other Backtracking problems, we can solve Sudoku by one by one assigning numbers to empty cells. Before assigning a number, we
check whether it is safe to assign. We basically check that the same number is not present in current row, current column and current 3X3 subgrid.
After checking for safety, we assign the number, and recursively check whether this assignment leads to a solution or not. If the assignment doesnt
lead to a solution, then we try next number for current empty cell. And if none of number (1 to 9) lead to solution, we return false.
Find row, col of an unassigned cell
If there is none, return true
For digits from 1 to 9
a) If there is no conflict for digit at row,col
assign digit to row,col and recursively try fill in rest of grid
b) If recursion successful, return true
c) Else, remove digit and try another
If all digits have been tried and nothing worked, return false

Following is C++ implementation for Sudoku problem. It prints the completely filled grid as output.
// A Backtracking program in C++ to solve Sudoku problem
#include <stdio.h>
// UNASSIGNED is used for empty cells in sudoku grid
#define UNASSIGNED 0
// N is used for size of Sudoku grid. Size will be NxN
#define N 9
// This function finds an entry in grid that is still unassigned
bool FindUnassignedLocation(int grid[N][N], int &row, int &col);
// Checks whether it will be legal to assign num to the given row,col
bool isSafe(int grid[N][N], int row, int col, int num);
/* Takes a partially filled-in grid and attempts to assign values to
all unassigned locations in such a way to meet the requirements
for Sudoku solution (non-duplication across rows, columns, and boxes) */
bool SolveSudoku(int grid[N][N])
{
int row, col;
// If there is no unassigned location, we are done
if (!FindUnassignedLocation(grid, row, col))
return true; // success!
// consider digits 1 to 9
for (int num = 1; num <= 9; num++)
{
// if looks promising
if (isSafe(grid, row, col, num))
{
// make tentative assignment
grid[row][col] = num;
// return, if success, yay!

if (SolveSudoku(grid))
return true;
// failure, unmake & try again
grid[row][col] = UNASSIGNED;
}
}
return false; // this triggers backtracking
}
/* Searches the grid to find an entry that is still unassigned. If
found, the reference parameters row, col will be set the location
that is unassigned, and true is returned. If no unassigned entries
remain, false is returned. */
bool FindUnassignedLocation(int grid[N][N], int &row, int &col)
{
for (row = 0; row < N; row++)
for (col = 0; col < N; col++)
if (grid[row][col] == UNASSIGNED)
return true;
return false;
}
/* Returns a boolean which indicates whether any assigned entry
in the specified row matches the given number. */
bool UsedInRow(int grid[N][N], int row, int num)
{
for (int col = 0; col < N; col++)
if (grid[row][col] == num)
return true;
return false;
}
/* Returns a boolean which indicates whether any assigned entry
in the specified column matches the given number. */
bool UsedInCol(int grid[N][N], int col, int num)
{
for (int row = 0; row < N; row++)
if (grid[row][col] == num)
return true;
return false;
}
/* Returns a boolean which indicates whether any assigned entry
within the specified 3x3 box matches the given number. */
bool UsedInBox(int grid[N][N], int boxStartRow, int boxStartCol, int num)
{
for (int row = 0; row < 3; row++)
for (int col = 0; col < 3; col++)
if (grid[row+boxStartRow][col+boxStartCol] == num)
return true;
return false;
}
/* Returns a boolean which indicates whether it will be legal to assign
num to the given row,col location. */
bool isSafe(int grid[N][N], int row, int col, int num)
{
/* Check if 'num' is not already placed in current row,
current column and current 3x3 box */
return !UsedInRow(grid, row, num) &&
!UsedInCol(grid, col, num) &&
!UsedInBox(grid, row - row%3 , col - col%3, num);
}
/* A utility function to print grid */
void printGrid(int grid[N][N])
{
for (int row = 0; row < N; row++)
{
for (int col = 0; col < N; col++)
printf("%2d", grid[row][col]);
printf("\n");
}
}
/* Driver Program to test above functions */
int main()
{
// 0 means unassigned cells
int grid[N][N] = {{3, 0, 6, 5, 0, 8, 4, 0, 0},

{5, 2,
{0, 8,
{0, 0,
{9, 0,
{0, 5,
{1, 3,
{0, 0,
{0, 0,
if (SolveSudoku(grid) ==
printGrid(grid);
else
printf("No solution

0, 0,
7, 0,
3, 0,
0, 8,
0, 0,
0, 0,
0, 0,
5, 2,
true)

0,
0,
1,
6,
9,
0,
0,
0,

0,
0,
0,
3,
0,
0,
0,
6,

0,
0,
0,
0,
6,
2,
0,
3,

0,
3,
8,
0,
0,
5,
7,
0,

0},
1},
0},
5},
0},
0},
4},
0}};

exists");

return 0;
}

Output:
3
5
4
2
9
8
1
6
7

1
2
8
6
7
5
3
9
4

6
9
7
3
4
1
8
2
5

5
1
6
4
8
7
9
3
2

7
3
2
1
6
9
4
5
8

8
4
9
5
3
2
7
1
6

4
7
5
9
1
6
2
8
3

9
6
3
8
2
4
5
7
1

2
8
1
7
5
3
6
4
9

References:
http://see.stanford.edu/materials/icspacs106b/H19-RecBacktrackExamples.pdf

Tug of War
Given a set of n integers, divide the set in two subsets of n/2 sizes each such that the difference of the sum of two subsets is as minimum as
possible. If n is even, then sizes of two subsets must be strictly n/2 and if n is odd, then size of one subset must be (n-1)/2 and size of other subset
must be (n+1)/2.
For example, let given set be {3, 4, 5, -3, 100, 1, 89, 54, 23, 20}, the size of set is 10. Output for this set should be {4, 100, 1, 23, 20} and {3,
5, -3, 89, 54}. Both output subsets are of size 5 and sum of elements in both subsets is same (148 and 148).
Let us consider another example where n is odd. Let given set be {23, 45, -34, 12, 0, 98, -99, 4, 189, -1, 4}. The output subsets should be {45,
-34, 12, 98, -1} and {23, 0, -99, 4, 189, 4}. The sums of elements in two subsets are 120 and 121 respectively.
The following solution tries every possible subset of half size. If one subset of half size is formed, the remaining elements form the other subset. We
initialize current set as empty and one by one build it. There are two possibilities for every element, either it is part of current set, or it is part of the
remaining elements (other subset). We consider both possibilities for every element. When the size of current set becomes n/2, we check whether
this solutions is better than the best solution available so far. If it is, then we update the best solution.
Following is C++ implementation for Tug of War problem. It prints the required arrays.
#include <iostream>
#include <stdlib.h>
#include <limits.h>
using namespace std;
// function that tries every possible solution by calling itself recursively
void TOWUtil(int* arr, int n, bool* curr_elements, int no_of_selected_elements,
bool* soln, int* min_diff, int sum, int curr_sum, int curr_position)
{
// checks whether the it is going out of bound
if (curr_position == n)
return;
// checks that the numbers of elements left are not less than the
// number of elements required to form the solution
if ((n/2 - no_of_selected_elements) > (n - curr_position))
return;
// consider the cases when current element is not included in the solution
TOWUtil(arr, n, curr_elements, no_of_selected_elements,
soln, min_diff, sum, curr_sum, curr_position+1);
// add the current element to the solution
no_of_selected_elements++;
curr_sum = curr_sum + arr[curr_position];
curr_elements[curr_position] = true;
// checks if a solution is formed
if (no_of_selected_elements == n/2)
{
// checks if the solution formed is better than the best solution so far
if (abs(sum/2 - curr_sum) < *min_diff)
{
*min_diff = abs(sum/2 - curr_sum);
for (int i = 0; i<n; i++)
soln[i] = curr_elements[i];
}
}
else
{
// consider the cases where current element is included in the solution
TOWUtil(arr, n, curr_elements, no_of_selected_elements, soln,
min_diff, sum, curr_sum, curr_position+1);
}
// removes current element before returning to the caller of this function
curr_elements[curr_position] = false;
}
// main function that generate an arr
void tugOfWar(int *arr, int n)
{
// the boolen array that contains the inclusion and exclusion of an element
// in current set. The number excluded automatically form the other set
bool* curr_elements = new bool[n];
// The inclusion/exclusion array for final solution
bool* soln = new bool[n];

int min_diff = INT_MAX;

int sum = 0;
for (int i=0; i<n; i++)
{
sum += arr[i];
curr_elements[i] = soln[i] = false;
}
// Find the solution using recursive function TOWUtil()
TOWUtil(arr, n, curr_elements, 0, soln, &min_diff, sum, 0, 0);
// Print the solution
cout << "The first subset is: ";
for (int i=0; i<n; i++)
{
if (soln[i] == true)
cout << arr[i] << " ";
}
cout << "\nThe second subset is: ";
for (int i=0; i<n; i++)
{
if (soln[i] == false)
cout << arr[i] << " ";
}
}
// Driver program to test above functions
int main()
{
int arr[] = {23, 45, -34, 12, 0, 98, -99, 4, 189, -1, 4};
int n = sizeof(arr)/sizeof(arr[0]);
tugOfWar(arr, n);
return 0;
}

Output:
The first subset is: 45 -34 12 98 -1
The second subset is: 23 0 -99 4 189 4

Backtracking | Set 8 (Solving Cryptarithmetic Puzzles)

Newspapers and magazines often have crypt-arithmetic puzzles of the form:
SEND
+ MORE
-------MONEY
--------

The goal here is to assign each letter a digit from 0 to 9 so that the arithmetic works out correctly. The rules are that all occurrences of a letter must
be assigned the same digit, and no digit can be assigned to more than one letter.
First, create a list of all the characters that need assigning to pass to Solve
If all characters are assigned, return true if puzzle is solved, false otherwise
Otherwise, consider the first unassigned character
for (every possible choice among the digits not in use)
make that choice and then recursively try to assign the rest of the characters
if recursion sucessful, return true
if !successful, unmake assignment and try another digit
If all digits have been tried and nothing worked, return false to trigger backtracking
/* ExhaustiveSolve
* --------------* This is the "not-very-smart" version of cryptarithmetic solver. It takes
* the puzzle itself (with the 3 strings for the two addends and sum) and a
* string of letters as yet unassigned. If no more letters to assign
* then we've hit a base-case, if the current letter-to-digit mapping solves
* the puzzle, we're done, otherwise we return false to trigger backtracking
* If we have letters to assign, we take the first letter from that list, and
* try assigning it the digits from 0 to 9 and then recursively working
* through solving puzzle from here. If we manage to make a good assignment
* that works, we've succeeded, else we need to unassign that choice and try
* another digit. This version is easy to write, since it uses a simple
* approach (quite similar to permutations if you think about it) but it is
* not so smart because it doesn't take into account the structure of the
* puzzle constraints (for example, once the two digits for the addends have
* been assigned, there is no reason to try anything other than the correct
* digit for the sum) yet it tries a lot of useless combos regardless
*/
bool ExhaustiveSolve(puzzleT puzzle, string lettersToAssign)
{
if (lettersToAssign.empty()) // no more choices to make
return PuzzleSolved(puzzle); // checks arithmetic to see if works
for (int digit = 0; digit <= 9; digit++) // try all digits
{
if (AssignLetterToDigit(lettersToAssign[0], digit))
{
if (ExhaustiveSolve(puzzle, lettersToAssign.substr(1)))
return true;
UnassignLetterFromDigit(lettersToAssign[0], digit);
}
}
return false; // nothing worked, need to backtrack
}

The algorithm above actually has a lot in common with the permutations algorithm, it pretty much just creates all arrangements of the mapping from
characters to digits and tries each until one works or all have been successfully tried. For a large puzzle, this could take a while.
A smarter algorithm could take into account the structure of the puzzle and avoid going down dead-end paths. For example, if we assign the
characters starting from the ones place and moving to the left, at each stage, we can verify the correctness of what we have so far before we
continue onwards. This definitely complicates the code but leads to a tremendous improvement in efficiency, making it much more feasible to solve
large puzzles.
Below pseudocode in this case has more special cases, but the same general design
Start by examining the rightmost digit of the topmost row, with a carry of 0
If we are beyond the leftmost digit of the puzzle, return true if no carry, false otherwise
If we are currently trying to assign a char in one of the addends
If char already assigned, just recur on row beneath this one, adding value into sum
If not assigned, then
for (every possible choice among the digits not in use)

make that choice and then on row beneath this one, if successful, return true
if !successful, unmake assignment and try another digit
return false if no assignment worked to trigger backtracking
Else if trying to assign a char in the sum
If char assigned & matches correct,
recur on next column to the left with carry, if success return true,
If char assigned & doesnt match, return false
If char unassigned & correct digit already used, return false
If char unassigned & correct digit unused,
assign it and recur on next column to left with carry, if success return true
return false to trigger backtracking
Source:
http://see.stanford.edu/materials/icspacs106b/H19-RecBacktrackExamples.pdf

Divide and Conquer | Set 1 (Introduction)

Like Greedy and Dynamic Programming, Divide and Conquer is an algorithmic paradigm. A typical Divide and Conquer algorithm solves a
problem using following three steps.
1. Divide: Break the given problem into subproblems of same type.
2. Conquer: Recursively solve these subproblems
3. Combine: Appropriately combine the answers
Following are some standard algorithms that are Divide and Conquer algorithms.
1) Binary Search is a searching algorithm. In each step, the algorithm compares the input element x with the value of the middle element in array.
If the values match, return the index of middle. Otherwise, if x is less than the middle element, then the algorithm recurs for left side of middle
element, else recurs for right side of middle element.
2) Quicksort is a sorting algorithm. The algorithm picks a pivot element, rearranges the array elements in such a way that all elements smaller than
the picked pivot element move to left side of pivot, and all greater elements move to right side. Finally, the algorithm recursively sorts the subarrays
on left and right of pivot element.
3) Merge Sort is also a sorting algorithm. The algorithm divides the array in two halves, recursively sorts them and finally merges the two sorted
halves.
4) Closest Pair of Points The problem is to find the closest pair of points in a set of points in x-y plane. The problem can be solved in O(n^2)
time by calculating distances of every pair of points and comparing the distances to find the minimum. The Divide and Conquer algorithm solves the
problem in O(nLogn) time.
5) Strassens Algorithm is an efficient algorithm to multiply two matrices. A simple method to multiply two matrices need 3 nested loops and is
O(n^3). Strassens algorithm multiplies two matrices in O(n^2.8974) time.
6) CooleyTukey Fast Fourier Transform (FFT) algorithm is the most common algorithm for FFT. It is a divide and conquer algorithm which
works in O(nlogn) time.
7) Karatsuba algorithm for fast multiplication it does multiplication of twon-digit numbers in at most
single-digit
multiplications in general (and exactly
whennis a power of 2). It is therefore faster than theclassical algorithm, which requiresn2single-digit
10
products. Ifn= 2 = 1024, in particular, the exact counts are 310= 59,049 and (210)2= 1,048,576, respectively.
We will publishing above algorithms in separate posts.
Divide and Conquer (D & C) vs Dynamic Programming (DP)
Both paradigms (D & C and DP) divide the given problem into subproblems and solve subproblems. How to choose one of them for a given
problem? Divide and Conquer should be used when same subproblems are not evaluated many times. Otherwise Dynamic Programming or
Memoization should be used. For example, Binary Search is a Divide and Conquer algorithm, we never evaluate the same subproblems again. On
the other hand, for calculating nth Fibonacci number, Dynamic Programming should be preferred (See this for details).
References
Algorithms by Sanjoy Dasgupta, Christos Papadimitriou, Umesh Vazirani
Introduction to Algorithms by Clifford Stein, Thomas H. Cormen, Charles E. Leiserson, Ronald L.
http://en.wikipedia.org/wiki/Karatsuba_algorithm

Write a program to calculate pow(x,n)

Below solution divides the problem into subproblems of size y/2 and call the subproblems recursively.
#include<stdio.h>
/* Function to calculate x raised to the power y */
int power(int x, unsigned int y)
{
if( y == 0)
return 1;
else if (y%2 == 0)
return power(x, y/2)*power(x, y/2);
else
return x*power(x, y/2)*power(x, y/2);
}
/* Program to test function power */
int main()
{
int x = 2;
unsigned int y = 3;
printf("%d", power(x, y));
getchar();
return 0;
}

Time Complexity: O(n)

Space Complexity: O(1)
Algorithmic Paradigm: Divide and conquer.
Above function can be optimized to O(logn) by calculating power(x, y/2) only once and storing it.
/* Function to calculate x raised to the power y in O(logn)*/
int power(int x, unsigned int y)
{
int temp;
if( y == 0)
return 1;
temp = power(x, y/2);
if (y%2 == 0)
return temp*temp;
else
return x*temp*temp;
}

Time Complexity of optimized solution: O(logn)

Let us extend the pow function to work for negative y and float x.
/* Extended version of power function that can work
for float x and negative y*/
#include<stdio.h>
float power(float x, int y)
{
float temp;
if( y == 0)
return 1;
temp = power(x, y/2);
if (y%2 == 0)
return temp*temp;
else
{
if(y > 0)
return x*temp*temp;
else
return (temp*temp)/x;
}
}
/* Program to test function power */
int main()
{
float x = 2;
int y = -3;
printf("%f", power(x, y));
getchar();

return 0;
}

Median of two sorted arrays

Question: There are 2 sorted arrays A and B of size n each. Write an algorithm to find the median of the array obtained after merging the above 2
arrays(i.e. array of length 2n). The complexity should be O(log(n))
Median: In probability theory and statistics, a median is described as the number separating the higher half of a sample, a population, or a
probability distribution, from the lower half.
The median of a finite list of numbers can be found by arranging all the numbers from lowest value to highest value and picking the middle one.
For getting the median of input array { 12, 11, 15, 10, 20 }, first sort the array. We get { 10, 11, 12, 15, 20 } after sorting. Median is the middle
element of the sorted array which is 12.
There are different conventions to take median of an array with even number of elements, one can take the mean of the two middle values, or first
middle value, or second middle value.
Let us see different methods to get the median of two sorted arrays of size n each. Since size of the set for which we are looking for median is even
(2n), we are taking average of middle two numbers in all below solutions.
Method 1 (Simply count while Merging)
Use merge procedure of merge sort. Keep track of count while comparing elements of two arrays. If count becomes n(For 2n elements), we have
reached the median. Take the average of the elements at indexes n-1 and n in the merged array. See the below implementation.
Implementation:
#include <stdio.h>
/* This function returns median of ar1[] and ar2[].
Assumptions in this function:
Both ar1[] and ar2[] are sorted arrays
Both have n elements */
int getMedian(int ar1[], int ar2[], int n)
{
int i = 0; /* Current index of i/p array ar1[] */
int j = 0; /* Current index of i/p array ar2[] */
int count;
int m1 = -1, m2 = -1;
/* Since there are 2n elements, median will be average
of elements at index n-1 and n in the array obtained after
merging ar1 and ar2 */
for (count = 0; count <= n; count++)
{
/*Below is to handle case where all elements of ar1[] are
smaller than smallest(or first) element of ar2[]*/
if (i == n)
{
m1 = m2;
m2 = ar2[0];
break;
}
/*Below is to handle case where all elements of ar2[] are
smaller than smallest(or first) element of ar1[]*/
else if (j == n)
{
m1 = m2;
m2 = ar1[0];
break;
}
if (ar1[i] < ar2[j])
{
m1 = m2; /* Store the prev median */
m2 = ar1[i];
i++;
}
else
{
m1 = m2; /* Store the prev median */
m2 = ar2[j];
j++;
}
}
return (m1 + m2)/2;
}

/* Driver program to test above function */

int main()
{
int ar1[] = {1, 12, 15, 26, 38};
int ar2[] = {2, 13, 17, 30, 45};
int n1 = sizeof(ar1)/sizeof(ar1[0]);
int n2 = sizeof(ar2)/sizeof(ar2[0]);
if (n1 == n2)
printf("Median is %d", getMedian(ar1, ar2, n1));
else
printf("Doesn't work for arrays of unequal size");
getchar();
return 0;
}

Time Complexity: O(n)

Method 2 (By comparing the medians of two arrays)

This method works by first getting medians of the two sorted arrays and then comparing them.
Let ar1 and ar2 be the input arrays.
Algorithm:
1) Calculate the medians m1 and m2 of the input arrays ar1[]
and ar2[] respectively.
2) If m1 and m2 both are equal then we are done.
return m1 (or m2)
3) If m1 is greater than m2, then median is present in one
of the below two subarrays.
a) From first element of ar1 to m1 (ar1[0...|_n/2_|])
b) From m2 to last element of ar2 (ar2[|_n/2_|...n-1])
4) If m2 is greater than m1, then median is present in one
of the below two subarrays.
a) From m1 to last element of ar1 (ar1[|_n/2_|...n-1])
b) From first element of ar2 to m2 (ar2[0...|_n/2_|])
5) Repeat the above process until size of both the subarrays
becomes 2.
6) If size of the two arrays is 2 then use below formula to get
the median.
Median = (max(ar1[0], ar2[0]) + min(ar1[1], ar2[1]))/2

Example:
ar1[] = {1, 12, 15, 26, 38}
ar2[] = {2, 13, 17, 30, 45}

For above two arrays m1 = 15 and m2 = 17

For the above ar1[] and ar2[], m1 is smaller than m2. So median is present in one of the following two subarrays.
[15, 26, 38] and [2, 13, 17]

Let us repeat the process for above two subarrays:

m1 = 26 m2 = 13.

m1 is greater than m2. So the subarrays become

[15, 26] and [13, 17]
Now size is 2, so median
=
=
=

= (max(ar1[0], ar2[0]) + min(ar1[1], ar2[1]))/2

(max(15, 13) + min(26, 17))/2
(15 + 17)/2
16

Implementation:
#include<stdio.h>
int max(int, int); /* to get maximum of two integers */
int min(int, int); /* to get minimum of two integeres */
int median(int [], int); /* to get median of a sorted array */

/* This function returns median of ar1[] and ar2[].

Assumptions in this function:
Both ar1[] and ar2[] are sorted arrays
Both have n elements */
int getMedian(int ar1[], int ar2[], int n)
{
int m1; /* For median of ar1 */
int m2; /* For median of ar2 */
/* return -1 for invalid input */
if (n <= 0)
return -1;
if (n == 1)
return (ar1[0] + ar2[0])/2;
if (n == 2)
return (max(ar1[0], ar2[0]) + min(ar1[1], ar2[1])) / 2;
m1 = median(ar1, n); /* get the median of the first array */
m2 = median(ar2, n); /* get the median of the second array */
/* If medians are equal then return either m1 or m2 */
if (m1 == m2)
return m1;
/* if m1 < m2 then median must exist in ar1[m1....] and ar2[....m2] */
if (m1 < m2)
{
if (n % 2 == 0)
return getMedian(ar1 + n/2 - 1, ar2, n - n/2 +1);
else
return getMedian(ar1 + n/2, ar2, n - n/2);
}
/* if m1 > m2 then median must exist in ar1[....m1] and ar2[m2...] */
else
{
if (n % 2 == 0)
return getMedian(ar2 + n/2 - 1, ar1, n - n/2 + 1);
else
return getMedian(ar2 + n/2, ar1, n - n/2);
}
}
/* Function to
int median(int
{
if (n%2 ==
return
else
return
}

get median of a sorted array */

arr[], int n)
0)
(arr[n/2] + arr[n/2-1])/2;
arr[n/2];

/* Driver program to test above function */

int main()
{
int ar1[] = {1, 2, 3, 6};
int ar2[] = {4, 6, 8, 10};
int n1 = sizeof(ar1)/sizeof(ar1[0]);
int n2 = sizeof(ar2)/sizeof(ar2[0]);
if (n1 == n2)
printf("Median is %d", getMedian(ar1, ar2, n1));
else
printf("Doesn't work for arrays of unequal size");
getchar();
return 0;
}
/* Utility functions */
int max(int x, int y)
{
return x > y? x : y;
}
int min(int x, int y)
{
return x > y? y : x;
}

Time Complexity: O(logn)

Algorithmic Paradigm: Divide and Conquer

Method 3 (By doing binary search for the median):

The basic idea is that if you are given two arrays ar1[] and ar2[] and know the length of each, you can check whether an element ar1[i] is the
median in constant time. Suppose that the median is ar1[i]. Since the array is sorted, it is greater than exactly i values in array ar1[]. Then if it is the
median, it is also greater than exactly j = n i 1 elements in ar2[].
It requires constant time to check if ar2[j] <= ar1[i] <= ar2[j + 1]. If ar1[i] is not the median, then depending on whether ar1[i] is greater or less
than ar2[j] and ar2[j + 1], you know that ar1[i] is either greater than or less than the median. Thus you can binary search for median in O(lg n)
worst-case time. For two arrays ar1 and ar2, first do binary search in ar1[]. If you reach at the end (left or right) of the first array and don't find
median, start searching in the second array ar2[].
1) Get the middle element of ar1[] using array indexes left and right.
Let index of the middle element be i.
2) Calculate the corresponding index j of ar2[]
j = n i 1
3) If ar1[i] >= ar2[j] and ar1[i] <= ar2[j+1] then ar1[i] and ar2[j]
are the middle elements.
return average of ar2[j] and ar1[i]
4) If ar1[i] is greater than both ar2[j] and ar2[j+1] then
do binary search in left half (i.e., arr[left ... i-1])
5) If ar1[i] is smaller than both ar2[j] and ar2[j+1] then
do binary search in right half (i.e., arr[i+1....right])
6) If you reach at any corner of ar1[] then do binary search in ar2[]

Example:
ar1[] = {1, 5, 7, 10, 13}
ar2[] = {11, 15, 23, 30, 45}

Middle element of ar1[] is 7. Let us compare 7 with 23 and 30, since 7 smaller than both 23 and 30, move to right in ar1[]. Do binary search in
{10, 13}, this step will pick 10. Now compare 10 with 15 and 23. Since 10 is smaller than both 15 and 23, again move to right. Only 13 is there
in right side now. Since 13 is greater than 11 and smaller than 15, terminate here. We have got the median as 12 (average of 11 and 13)
Implementation:
#include<stdio.h>
int getMedianRec(int ar1[], int ar2[], int left, int right, int n);
/* This function returns median of ar1[] and ar2[].
Assumptions in this function:
Both ar1[] and ar2[] are sorted arrays
Both have n elements */
int getMedian(int ar1[], int ar2[], int n)
{
return getMedianRec(ar1, ar2, 0, n-1, n);
}
/* A recursive function to get the median of ar1[] and ar2[]
using binary search */
int getMedianRec(int ar1[], int ar2[], int left, int right, int n)
{
int i, j;
/* We have reached at the end (left or right) of ar1[] */
if (left > right)
return getMedianRec(ar2, ar1, 0, n-1, n);
i = (left + right)/2;
j = n - i - 1; /* Index of ar2[] */
/* Recursion terminates here.*/
if (ar1[i] > ar2[j] && (j == n-1 || ar1[i] <= ar2[j+1]))
{
/* ar1[i] is decided as median 2, now select the median 1
(element just before ar1[i] in merged array) to get the
average of both*/
if (i == 0 || ar2[j] > ar1[i-1])
return (ar1[i] + ar2[j])/2;
else
return (ar1[i] + ar1[i-1])/2;
}
/*Search in left half of ar1[]*/

else if (ar1[i] > ar2[j] && j != n-1 && ar1[i] > ar2[j+1])
return getMedianRec(ar1, ar2, left, i-1, n);
/*Search in right half of ar1[]*/
else /* ar1[i] is smaller than both ar2[j] and ar2[j+1]*/
return getMedianRec(ar1, ar2, i+1, right, n);
}
/* Driver program to test above function */
int main()
{
int ar1[] = {1, 12, 15, 26, 38};
int ar2[] = {2, 13, 17, 30, 45};
int n1 = sizeof(ar1)/sizeof(ar1[0]);
int n2 = sizeof(ar2)/sizeof(ar2[0]);
if (n1 == n2)
printf("Median is %d", getMedian(ar1, ar2, n1));
else
printf("Doesn't work for arrays of unequal size");
getchar();
return 0;
}

Time Complexity: O(logn)

Algorithmic Paradigm: Divide and Conquer
The above solutions can be optimized for the cases when all elements of one array are smaller than all elements of other array. For example, in
method 3, we can change the getMedian() function to following so that these cases can be handled in O(1) time. Thanks to nutcracker for
suggesting this optimization.
/* This function returns median of ar1[] and ar2[].
Assumptions in this function:
Both ar1[] and ar2[] are sorted arrays
Both have n elements */
int getMedian(int ar1[], int ar2[], int n)
{
// If all elements of array 1 are smaller then
// median is average of last element of ar1 and
// first element of ar2
if (ar1[n-1] < ar2[0])
return (ar1[n-1]+ar2[0])/2;
// If all elements of array 1 are smaller then
// median is average of first element of ar1 and
// last element of ar2
if (ar2[n-1] < ar1[0])
return (ar2[n-1]+ar1[0])/2;
return getMedianRec(ar1, ar2, 0, n-1, n);
}

References:
http://en.wikipedia.org/wiki/Median
http://ocw.alfaisal.edu/NR/rdonlyres/Electrical-Engineering-and-Computer-Science/6-046JFall-2005/30C68118-E436-4FE3-8C796BAFBB07D935/0/ps9sol.pdf ds3etph5wn
Asked by Snehal

Count Inversions in an array

Inversion Count for an array indicates how far (or close) the array is from being sorted. If array is already sorted then inversion count is 0. If
array is sorted in reverse order that inversion count is the maximum.
Formally speaking, two elements a[i] and a[j] form an inversion if a[i] > a[j] and i < j Example:
The sequence 2, 4, 1, 3, 5 has three inversions (2, 1), (4, 1), (4, 3).
METHOD 1 (Simple)
For each element, count number of elements which are on right side of it and are smaller than it.
int getInvCount(int arr[], int n)
{
int inv_count = 0;
int i, j;
for(i = 0; i < n - 1; i++)
for(j = i+1; j < n; j++)
if(arr[i] > arr[j])
inv_count++;
return inv_count;
}
/* Driver progra to test above functions */
int main(int argv, char** args)
{
int arr[] = {1, 20, 6, 4, 5};
printf(" Number of inversions are %d \n", getInvCount(arr, 5));
getchar();
return 0;
}

Time Complexity: O(n^2)

METHOD 2(Enhance Merge Sort)

Suppose we know the number of inversions in the left half and right half of the array (let be inv1 and inv2), what kinds of inversions are not
accounted for in Inv1 + Inv2? The answer is the inversions we have to count during the merge step. Therefore, to get number of inversions, we
need to add number of inversions in left subarray, right subarray and merge().

How to get number of inversions in merge()?

In merge process, let i is used for indexing left sub-array and j for right sub-array. At any step in merge(), if a[i] is greater than a[j], then there are
(mid i) inversions. because left and right subarrays are sorted, so all the remaining elements in left-subarray (a[i+1], a[i+2] a[mid]) will be greater
than a[j]

The complete picture:

Implementation:
#include <stdio.h>
#include <stdlib.h>
int _mergeSort(int arr[], int temp[], int left, int right);
int merge(int arr[], int temp[], int left, int mid, int right);
/* This function sorts the input array and returns the
number of inversions in the array */
int mergeSort(int arr[], int array_size)
{
int *temp = (int *)malloc(sizeof(int)*array_size);
return _mergeSort(arr, temp, 0, array_size - 1);
}
/* An auxiliary recursive function that sorts the input array and
returns the number of inversions in the array. */
int _mergeSort(int arr[], int temp[], int left, int right)
{
int mid, inv_count = 0;
if (right > left)
{
/* Divide the array into two parts and call _mergeSortAndCountInv()
for each of the parts */
mid = (right + left)/2;
/* Inversion
and number
inv_count =
inv_count +=

count will be sum of inversions in left-part, right-part

of inversions in merging */
_mergeSort(arr, temp, left, mid);
_mergeSort(arr, temp, mid+1, right);

/*Merge the two parts*/

inv_count += merge(arr, temp, left, mid+1, right);
}
return inv_count;

}
/* This funt merges two sorted arrays and returns inversion count in
the arrays.*/
int merge(int arr[], int temp[], int left, int mid, int right)
{
int i, j, k;
int inv_count = 0;
i = left; /* i is index for
j = mid; /* i is index for
k = left; /* i is index for
while ((i <= mid - 1) && (j
{
if (arr[i] <= arr[j])
{
temp[k++] = arr[i++];
}
else
{
temp[k++] = arr[j++];

left subarray*/
right subarray*/
resultant merged subarray*/
<= right))

/*this is tricky -- see above explanation/diagram for merge()*/

inv_count = inv_count + (mid - i);
}
}
/* Copy the remaining elements of left subarray
(if there are any) to temp*/
while (i <= mid - 1)
temp[k++] = arr[i++];
/* Copy the remaining elements of right subarray
(if there are any) to temp*/
while (j <= right)
temp[k++] = arr[j++];
/*Copy back the merged elements to original array*/
for (i=left; i <= right; i++)
arr[i] = temp[i];
return inv_count;
}
/* Driver progra to test above functions */
int main(int argv, char** args)
{
int arr[] = {1, 20, 6, 4, 5};
printf(" Number of inversions are %d \n", mergeSort(arr, 5));
getchar();
return 0;
}

Note that above code modifies (or sorts) the input array. If we want to count only inversions then we need to create a copy of original array and
call mergeSort() on copy.
Time Complexity: O(nlogn)
Algorithmic Paradigm: Divide and Conquer
References:
http://www.cs.umd.edu/class/fall2009/cmsc451/lectures/Lec08-inversions.pdf
http://www.cp.eng.chula.ac.th/~piak/teaching/algo/algo2008/count-inv.htm

Divide and Conquer | Set 2 (Closest Pair of Points)

We are given an array of n points in the plane, and the problem is to find out the closest pair of points in the array. This problem arises in a number
of applications. For example, in air-traffic control, you may want to monitor planes that come too close together, since this may indicate a possible
collision. Recall the following formula for distance between two points p and q.

The Brute force solution is O(n^2), compute the distance between each pair and return the smallest. We can calculate the smallest distance in
O(nLogn) time using Divide and Conquer strategy. In this post, a O(n x (Logn)^2) approach is discussed. We will be discussing a O(nLogn)
approach in a separate post.
Algorithm
Following are the detailed steps of a O(n (Logn)^2) algortihm.
Input: An array of n points P[]
Output: The smallest distance between two points in the given array.
As a pre-processing step, input array is sorted according to x coordinates.
1) Find the middle point in the sorted array, we can take P[n/2] as middle point.
2) Divide the given array in two halves. The first subarray contains points from P[0] to P[n/2]. The second subarray contains points from P[n/2+1]
to P[n-1].
3) Recursively find the smallest distances in both subarrays. Let the distances be dl and dr. Find the minimum of dl and dr. Let the minimum be d.

4) From above 3 steps, we have an upper bound d of minimum distance. Now we need to consider the pairs such that one point in pair is from left
half and other is from right half. Consider the vertical line passing through passing through P[n/2] and find all points whose x coordinate is closer
than d to the middle vertical line. Build an array strip[] of all such points.

5) Sort the array strip[] according to y coordinates. This step is O(nLogn). It can be optimized to O(n) by recursively sorting and merging.
6) Find the smallest distance in strip[]. This is tricky. From first look, it seems to be a O(n^2) step, but it is actually O(n). It can be proved
geometrically that for every point in strip, we only need to check at most 7 points after it (note that strip is sorted according to Y coordinate). See
this for more analysis.
7) Finally return the minimum of d and distance calculated in above step (step 6)
Implementation
Following is C/C++ implementation of the above algorithm.
// A divide and conquer program in C/C++ to find the smallest distance from a
// given set of points.
#include <stdio.h>
#include <float.h>

#include <stdlib.h>
#include <math.h>
// A structure to represent a Point in 2D plane
struct Point
{
int x, y;
};
/* Following two functions are needed for library function qsort().
Refer: http://www.cplusplus.com/reference/clibrary/cstdlib/qsort/ */
// Needed to sort array of points according to X coordinate
int compareX(const void* a, const void* b)
{
Point *p1 = (Point *)a, *p2 = (Point *)b;
return (p1->x - p2->x);
}
// Needed to sort array of points according to Y coordinate
int compareY(const void* a, const void* b)
{
Point *p1 = (Point *)a, *p2 = (Point *)b;
return (p1->y - p2->y);
}
// A utility function to find the distance between two points
float dist(Point p1, Point p2)
{
return sqrt( (p1.x - p2.x)*(p1.x - p2.x) +
(p1.y - p2.y)*(p1.y - p2.y)
);
}
// A Brute Force method to return the smallest distance between two points
// in P[] of size n
float bruteForce(Point P[], int n)
{
float min = FLT_MAX;
for (int i = 0; i < n; ++i)
for (int j = i+1; j < n; ++j)
if (dist(P[i], P[j]) < min)
min = dist(P[i], P[j]);
return min;
}
// A utility function to find minimum of two float values
float min(float x, float y)
{
return (x < y)? x : y;
}
// A utility function to find the distance beween the closest points of
// strip of given size. All points in strip[] are sorted accordint to
// y coordinate. They all have an upper bound on minimum distance as d.
// Note that this method seems to be a O(n^2) method, but it's a O(n)
// method as the inner loop runs at most 6 times
float stripClosest(Point strip[], int size, float d)
{
float min = d; // Initialize the minimum distance as d
qsort(strip, size, sizeof(Point), compareY);
// Pick all points one by one and try the next points till the difference
// between y coordinates is smaller than d.
// This is a proven fact that this loop runs at most 6 times
for (int i = 0; i < size; ++i)
for (int j = i+1; j < size && (strip[j].y - strip[i].y) < min; ++j)
if (dist(strip[i],strip[j]) < min)
min = dist(strip[i], strip[j]);
return min;
}
// A recursive function to find the smallest distance. The array P contains
// all points sorted according to x coordinate
float closestUtil(Point P[], int n)
{
// If there are 2 or 3 points, then use brute force
if (n <= 3)
return bruteForce(P, n);

// Find the middle point

int mid = n/2;
Point midPoint = P[mid];
// Consider the vertical line passing through the middle point
// calculate the smallest distance dl on left of middle point and
// dr on right side
float dl = closestUtil(P, mid);
float dr = closestUtil(P + mid, n-mid);
// Find the smaller of two distances
float d = min(dl, dr);
// Build an array strip[] that contains points close (closer than d)
// to the line passing through the middle point
Point strip[n];
int j = 0;
for (int i = 0; i < n; i++)
if (abs(P[i].x - midPoint.x) < d)
strip[j] = P[i], j++;
// Find the closest points in strip. Return the minimum of d and closest
// distance is strip[]
return min(d, stripClosest(strip, j, d) );
}
// The main functin that finds the smallest distance
// This method mainly uses closestUtil()
float closest(Point P[], int n)
{
qsort(P, n, sizeof(Point), compareX);
// Use recursive function closestUtil() to find the smallest distance
return closestUtil(P, n);
}
// Driver program to test above functions
int main()
{
Point P[] = {{2, 3}, {12, 30}, {40, 50}, {5, 1}, {12, 10}, {3, 4}};
int n = sizeof(P) / sizeof(P[0]);
printf("The smallest distance is %f ", closest(P, n));
return 0;
}

Output:
The smallest distance is 1.414214

Time Complexity Let Time complexity of above algorithm be T(n). Let us assume that we use a O(nLogn) sorting algorithm. The above algorithm
divides all points in two sets and recursively calls for two sets. After dividing, it finds the strip in O(n) time, sorts the strip in O(nLogn) time and
finally finds the closest points in strip in O(n) time. So T(n) can expressed as follows
T(n) = 2T(n/2) + O(n) + O(nLogn) + O(n)
T(n) = 2T(n/2) + O(nLogn)
T(n) = T(n x Logn x Logn)
Notes
1) Time complexity can be improved to O(nLogn) by optimizing step 5 of the above algorithm. We will soon be discussing the optimized solution
in a separate post.
2) The code finds smallest distance. It can be easily modified to find the points with smallest distance.
3) The code uses quick sort which can be O(n^2) in worst case. To have the upper bound as O(n (Logn)^2), a O(nLogn) sorting algorithm like
merge sort or heap sort can be used
References:
http://www.cs.umd.edu/class/fall2013/cmsc451/Lects/lect10.pdf
http://www.youtube.com/watch?v=vS4Zn1a9KUc
http://www.youtube.com/watch?v=T3T7T8Ym20M
http://en.wikipedia.org/wiki/Closest_pair_of_points_problem

Divide and Conquer | Set 5 (Strassens Matrix Multiplication)

Given two square matrices A and B of size n x n each, find their multiplication matrix.
Naive Method
Following is a simple way to multiply two matrices.
void multiply(int A[][N], int B[][N], int C[][N])
{
for (int i = 0; i < N; i++)
{
for (int j = 0; j < N; j++)
{
C[i][j] = 0;
for (int k = 0; k < N; k++)
{
C[i][j] += A[i][k]*B[k][j];
}
}
}
}

Time Complexity of above method is O(N3).

Divide and Conquer
Following is simple Divide and Conquer method to multiply two square matrices.
1) Divide matrices A and B in 4 sub-matrices of size N/2 x N/2 as shown in the below diagram.
2) Calculate following values recursively. ae + bg, af + bh, ce + dg and cf + dh.

In the above method, we do 8 multiplications for matrices of size N/2 x N/2 and 4 additions. Addition of two matrices takes O(N2) time. So the
time complexity can be written as
T(N) = 8T(N/2) + O(N2)
From Master's Theorem, time complexity of above method is O(N3)
which is unfortunately same as the above naive method.

Simple Divide and Conquer also leads to O(N3), can there be a better way?
In the above divide and conquer method, the main component for high time complexity is 8 recursive calls. The idea of Strassens method is to
reduce the number of recursive calls to 7. Strassens method is similar to above simple divide and conquer method in the sense that this method also
divide matrices to sub-matrices of size N/2 x N/2 as shown in the above diagram, but in Strassens method, the four sub-matrices of result are
calculated using following formulae.

Time Complexity of Strassens Method

Addition and Subtraction of two matrices takes O(N2) time. So time complexity can be written as
T(N) = 7T(N/2) + O(N2)
From Master's Theorem, time complexity of above method is
O(NLog7) which is approximately O(N2.8074)

Generally Strassens Method is not preferred for practical applications for following reasons.
1) The constants used in Strassens method are high and for a typical application Naive method works better.
2) For Sparse matrices, there are better methods especially designed for them.
3) The submatrices in recursion take extra space.
4) Because of the limited precision of computer arithmetic on noninteger values, larger errors accumulate in Strassens algorithm than in Naive
Method (Source: CLRS Book)
References:
Introduction to Algorithms 3rd Edition by Clifford Stein, Thomas H. Cormen, Charles E. Leiserson, Ronald L. Rivest
https://www.youtube.com/watch?v=LOLebQ8nKHA
https://www.youtube.com/watch?v=QXY4RskLQcI

Closest Pair of Points | O(nlogn) Implementation

We are given an array of n points in the plane, and the problem is to find out the closest pair of points in the array. This problem arises in a number
of applications. For example, in air-traffic control, you may want to monitor planes that come too close together, since this may indicate a possible
collision. Recall the following formula for distance between two points p and q.
We have discussed a divide and conquer solution for this problem. The time complexity of the implementation provided in the previous post is O(n
(Logn)^2). In this post, we discuss an implementation with time complexity as O(nLogn).
Following is a recap of the algorithm discussed in the previous post.
1) We sort all points according to x coordinates.
2) Divide all points in two halves.
3) Recursively find the smallest distances in both subarrays.
4) Take the minimum of two smallest distances. Let the minimum be d.
5) Create an array strip[] that stores all points which are at most d distance away from the middle line dividing the two sets.
6) Find the smallest distance in strip[].
7) Return the minimum of d and the smallest distance calculated in above step 6.
The great thing about the above approach is, if the array strip[] is sorted according to y coordinate, then we can find the smallest distance in strip[]
in O(n) time. In the implementation discussed in previous post, strip[] was explicitly sorted in every recursive call that made the time complexity
O(n (Logn)^2), assuming that the sorting step takes O(nLogn) time.
In this post, we discuss an implementation where the time complexity is O(nLogn). The idea is to presort all points according to y coordinates. Let
the sorted array be Py[]. When we make recursive calls, we need to divide points of Py[] also according to the vertical line. We can do that by
simply processing every point and comparing its x coordinate with x coordinate of middle line.
Following is C++ implementation of O(nLogn) approach.
// A divide and conquer program in C++ to find the smallest distance from a
// given set of points.
#include <iostream>
#include <float.h>
#include <stdlib.h>
#include <math.h>
using namespace std;
// A structure to represent a Point in 2D plane
struct Point
{
int x, y;
};
/* Following two functions are needed for library function qsort().
Refer: http://www.cplusplus.com/reference/clibrary/cstdlib/qsort/ */
// Needed to sort array of points according to X coordinate
int compareX(const void* a, const void* b)
{
Point *p1 = (Point *)a, *p2 = (Point *)b;
return (p1->x - p2->x);
}
// Needed to sort array of points according to Y coordinate
int compareY(const void* a, const void* b)
{
Point *p1 = (Point *)a, *p2 = (Point *)b;
return (p1->y - p2->y);
}
// A utility function to find the distance between two points
float dist(Point p1, Point p2)
{
return sqrt( (p1.x - p2.x)*(p1.x - p2.x) +
(p1.y - p2.y)*(p1.y - p2.y)
);
}

// A Brute Force method to return the smallest distance between two points
// in P[] of size n
float bruteForce(Point P[], int n)
{
float min = FLT_MAX;
for (int i = 0; i < n; ++i)
for (int j = i+1; j < n; ++j)
if (dist(P[i], P[j]) < min)
min = dist(P[i], P[j]);
return min;
}
// A utility function to find minimum of two float values
float min(float x, float y)
{
return (x < y)? x : y;
}
// A utility function to find the distance beween the closest points of
// strip of given size. All points in strip[] are sorted accordint to
// y coordinate. They all have an upper bound on minimum distance as d.
// Note that this method seems to be a O(n^2) method, but it's a O(n)
// method as the inner loop runs at most 6 times
float stripClosest(Point strip[], int size, float d)
{
float min = d; // Initialize the minimum distance as d
// Pick all points one by one and try the next points till the difference
// between y coordinates is smaller than d.
// This is a proven fact that this loop runs at most 6 times
for (int i = 0; i < size; ++i)
for (int j = i+1; j < size && (strip[j].y - strip[i].y) < min; ++j)
if (dist(strip[i],strip[j]) < min)
min = dist(strip[i], strip[j]);
return min;
}
// A recursive function to find the smallest distance. The array Px contains
// all points sorted according to x coordinates and Py contains all points
// sorted according to y coordinates
float closestUtil(Point Px[], Point Py[], int n)
{
// If there are 2 or 3 points, then use brute force
if (n <= 3)
return bruteForce(Px, n);
// Find the middle point
int mid = n/2;
Point midPoint = Px[mid];
// Divide points in y sorted array around the vertical line.
// Assumption: All x coordinates are distinct.
Point Pyl[mid+1]; // y sorted points on left of vertical line
Point Pyr[n-mid-1]; // y sorted points on right of vertical line
int li = 0, ri = 0; // indexes of left and right subarrays
for (int i = 0; i < n; i++)
{
if (Py[i].x <= midPoint.x)
Pyl[li++] = Py[i];
else
Pyr[ri++] = Py[i];
}
// Consider the vertical line passing through the middle point
// calculate the smallest distance dl on left of middle point and
// dr on right side
float dl = closestUtil(Px, Pyl, mid);
float dr = closestUtil(Px + mid, Pyr, n-mid);
// Find the smaller of two distances
float d = min(dl, dr);
// Build an array strip[] that contains points close (closer than d)
// to the line passing through the middle point
Point strip[n];
int j = 0;
for (int i = 0; i < n; i++)
if (abs(Py[i].x - midPoint.x) < d)

strip[j] = Py[i], j++;

// Find the closest points in strip. Return the minimum of d and closest
// distance is strip[]
return min(d, stripClosest(strip, j, d) );
}
// The main functin that finds the smallest distance
// This method mainly uses closestUtil()
float closest(Point P[], int n)
{
Point Px[n];
Point Py[n];
for (int i = 0; i < n; i++)
{
Px[i] = P[i];
Py[i] = P[i];
}
qsort(Px, n, sizeof(Point), compareX);
qsort(Py, n, sizeof(Point), compareY);
// Use recursive function closestUtil() to find the smallest distance
return closestUtil(Px, Py, n);
}
// Driver program to test above functions
int main()
{
Point P[] = {{2, 3}, {12, 30}, {40, 50}, {5, 1}, {12, 10}, {3, 4}};
int n = sizeof(P) / sizeof(P[0]);
cout << "The smallest distance is " << closest(P, n);
return 0;
}

Output:
The smallest distance is 1.41421

Time Complexity:Let Time complexity of above algorithm be T(n). Let us assume that we use a O(nLogn) sorting algorithm. The above
algorithm divides all points in two sets and recursively calls for two sets. After dividing, it finds the strip in O(n) time. Also, it takes O(n) time to
divide the Py array around the mid vertical line. Finally finds the closest points in strip in O(n) time. So T(n) can expressed as follows
T(n) = 2T(n/2) + O(n) + O(n) + O(n)
T(n) = 2T(n/2) + O(n)
T(n) = T(nLogn)
References:
http://www.cs.umd.edu/class/fall2013/cmsc451/Lects/lect10.pdf
http://www.youtube.com/watch?v=vS4Zn1a9KUc
http://www.youtube.com/watch?v=T3T7T8Ym20M
http://en.wikipedia.org/wiki/Closest_pair_of_points_problem

How to check if two given line segments intersect?

Given two line segments (p1, q1) and (p2, q2), find if the given line segments intersect with each other.
Before we discuss solution, let us define notion of orientation. Orientation of an ordered triplet of points in theplane can be

counterclockwise
clockwise
colinear
The following diagram shows different possible orientations of (a, b, c)

Note the word ordered here. Orientation of (a, b, c) may be different from orientation of (c, b, a).
How is Orientation useful here?
Two segments (p1,q1) and (p2,q2) intersect if and only ifone of the following two conditions is verified
1.General Case:
(p1, q1, p2) and (p1, q1, q2) have differentorientations and
(p2, q2,p1) and (p2, q2,q1) have differentorientations
2. Special Case
(p1, q1, p2), (p1, q1, q2), (p2, q2, p1), and (p2, q2, q1) areall collinear and
the x-projections of (p1, q1) and (p2, q2) intersect
the y-projections of (p1, q1) and (p2, q2) intersect
Examples of General Case:

Examples of Special Case:

Following is C++ implementation based on above idea.

// A C++ program to check if two given line segments intersect
#include <iostream>
using namespace std;
struct Point
{

int x;
int y;
};
// Given three colinear
// point q lies on line
bool onSegment(Point p,
{
if (q.x <= max(p.x,
q.y <= max(p.y,
return true;

points p, q, r, the function checks if

segment 'pr'
Point q, Point r)
r.x) && q.x >= min(p.x, r.x) &&
r.y) && q.y >= min(p.y, r.y))

return false;
}
// To find orientation of ordered triplet (p, q, r).
// The function returns following values
// 0 --> p, q and r are colinear
// 1 --> Clockwise
// 2 --> Counterclockwise
int orientation(Point p, Point q, Point r)
{
// See 10th slides from following link for derivation of the formula
// http://www.dcs.gla.ac.uk/~pat/52233/slides/Geometry1x1.pdf
int val = (q.y - p.y) * (r.x - q.x) (q.x - p.x) * (r.y - q.y);
if (val == 0) return 0; // colinear
return (val > 0)? 1: 2; // clock or counterclock wise
}
// The main function that returns true if line segment 'p1q1'
// and 'p2q2' intersect.
bool doIntersect(Point p1, Point q1, Point p2, Point q2)
{
// Find the four orientations needed for general and
// special cases
int o1 = orientation(p1, q1, p2);
int o2 = orientation(p1, q1, q2);
int o3 = orientation(p2, q2, p1);
int o4 = orientation(p2, q2, q1);
// General case
if (o1 != o2 && o3 != o4)
return true;
// Special Cases
// p1, q1 and p2 are colinear and p2 lies on segment p1q1
if (o1 == 0 && onSegment(p1, p2, q1)) return true;
// p1, q1 and p2 are colinear and q2 lies on segment p1q1
if (o2 == 0 && onSegment(p1, q2, q1)) return true;
// p2, q2 and p1 are colinear and p1 lies on segment p2q2
if (o3 == 0 && onSegment(p2, p1, q2)) return true;
// p2, q2 and q1 are colinear and q1 lies on segment p2q2
if (o4 == 0 && onSegment(p2, q1, q2)) return true;
return false; // Doesn't fall in any of the above cases
}
// Driver program to test above functions
int main()
{
struct Point p1 = {1, 1}, q1 = {10, 1};
struct Point p2 = {1, 2}, q2 = {10, 2};
doIntersect(p1, q1, p2, q2)? cout << "Yes\n": cout << "No\n";
p1 = {10, 0}, q1 = {0, 10};
p2 = {0, 0}, q2 = {10, 10};
doIntersect(p1, q1, p2, q2)? cout << "Yes\n": cout << "No\n";
p1 = {-5, -5}, q1 = {0, 0};
p2 = {1, 1}, q2 = {10, 10};
doIntersect(p1, q1, p2, q2)? cout << "Yes\n": cout << "No\n";
return 0;
}

Output:
No
Yes
No

Sources:
http://www.dcs.gla.ac.uk/~pat/52233/slides/Geometry1x1.pdf
Introduction to Algorithms 3rd Edition by Clifford Stein, Thomas H. Cormen, Charles E. Leiserson, Ronald L. Rivest

How to check if a given point lies inside or outside a polygon?

Given a polygon and a point p, find if p lies inside the polygon or not. The points lying on the border are considered inside.

We strongly recommend to see the following post first.

How to check if two given line segments intersect?
Following is a simple idea to check whether a point is inside or outside.
1) Draw a horizontal line to the right of each point and extend it to infinity
1) Count the number of times the line intersects with polygon edges.
2) A point is inside the polygon if either count of intersections is odd or
point lies on an edge of polygon. If none of the conditions is true, then
point lies outside.

How to handle point g in the above figure?

Note that we should returns true if the point lies on the line or same as one of the vertices of the given polygon. To handle this, after checking if the
line from p to extreme intersects, we check whether p is colinear with vertices of current line of polygon. If it is coliear, then we check if the point p
lies on current side of polygon, if it lies, we return true, else false.
Following is C++ implementation of the above idea.
// A C++ program to check if a given point lies inside a given polygon
// Refer http://www.geeksforgeeks.org/check-if-two-given-line-segments-intersect/
// for explanation of functions onSegment(), orientation() and doIntersect()
#include <iostream>
using namespace std;
// Define Infinite (Using INT_MAX caused overflow problems)
#define INF 10000
struct Point
{
int x;
int y;
};
// Given three colinear points p, q, r, the function checks if
// point q lies on line segment 'pr'
bool onSegment(Point p, Point q, Point r)
{
if (q.x <= max(p.x, r.x) && q.x >= min(p.x, r.x) &&
q.y <= max(p.y, r.y) && q.y >= min(p.y, r.y))
return true;
return false;
}
// To find orientation of ordered triplet (p, q, r).
// The function returns following values
// 0 --> p, q and r are colinear

// 1 --> Clockwise
// 2 --> Counterclockwise
int orientation(Point p, Point q, Point r)
{
int val = (q.y - p.y) * (r.x - q.x) (q.x - p.x) * (r.y - q.y);
if (val == 0) return 0; // colinear
return (val > 0)? 1: 2; // clock or counterclock wise
}
// The function that returns true if line segment 'p1q1'
// and 'p2q2' intersect.
bool doIntersect(Point p1, Point q1, Point p2, Point q2)
{
// Find the four orientations needed for general and
// special cases
int o1 = orientation(p1, q1, p2);
int o2 = orientation(p1, q1, q2);
int o3 = orientation(p2, q2, p1);
int o4 = orientation(p2, q2, q1);
// General case
if (o1 != o2 && o3 != o4)
return true;
// Special Cases
// p1, q1 and p2 are colinear and p2 lies on segment p1q1
if (o1 == 0 && onSegment(p1, p2, q1)) return true;
// p1, q1 and p2 are colinear and q2 lies on segment p1q1
if (o2 == 0 && onSegment(p1, q2, q1)) return true;
// p2, q2 and p1 are colinear and p1 lies on segment p2q2
if (o3 == 0 && onSegment(p2, p1, q2)) return true;
// p2, q2 and q1 are colinear and q1 lies on segment p2q2
if (o4 == 0 && onSegment(p2, q1, q2)) return true;
return false; // Doesn't fall in any of the above cases
}
// Returns true if the point p lies inside the polygon[] with n vertices
bool isInside(Point polygon[], int n, Point p)
{
// There must be at least 3 vertices in polygon[]
if (n < 3) return false;
// Create a point for line segment from p to infinite
Point extreme = {INF, p.y};
// Count intersections of the above line with sides of polygon
int count = 0, i = 0;
do
{
int next = (i+1)%n;
// Check if the line segment from 'p' to 'extreme' intersects
// with the line segment from 'polygon[i]' to 'polygon[next]'
if (doIntersect(polygon[i], polygon[next], p, extreme))
{
// If the point 'p' is colinear with line segment 'i-next',
// then check if it lies on segment. If it lies, return true,
// otherwise false
if (orientation(polygon[i], p, polygon[next]) == 0)
return onSegment(polygon[i], p, polygon[next]);
count++;
}
i = next;
} while (i != 0);
// Return true if count is odd, false otherwise
return count&1; // Same as (count%2 == 1)
}
// Driver program to test above functions
int main()
{
Point polygon1[] = {{0, 0}, {10, 0}, {10, 10}, {0, 10}};
int n = sizeof(polygon1)/sizeof(polygon1[0]);

Point p = {20, 20};

isInside(polygon1, n, p)? cout << "Yes \n": cout << "No \n";
p = {5, 5};
isInside(polygon1, n, p)? cout << "Yes \n": cout << "No \n";
Point polygon2[] = {{0, 0}, {5, 5}, {5, 0}};
p = {3, 3};
n = sizeof(polygon2)/sizeof(polygon2[0]);
isInside(polygon2, n, p)? cout << "Yes \n": cout << "No \n";
p = {5, 1};
isInside(polygon2, n, p)? cout << "Yes \n": cout << "No \n";
p = {8, 1};
isInside(polygon2, n, p)? cout << "Yes \n": cout << "No \n";
Point polygon3[] = {{0, 0}, {10, 0}, {10, 10}, {0, 10}};
p = {-1,10};
n = sizeof(polygon3)/sizeof(polygon3[0]);
isInside(polygon3, n, p)? cout << "Yes \n": cout << "No \n";
return 0;
}

Output:
No
Yes
Yes
Yes
No
No

Time Complexity: O(n) where n is the number of vertices in the given polygon.
Source:
http://www.dcs.gla.ac.uk/~pat/52233/slides/Geometry1x1.pdf

Convex Hull | Set 1 (Jarviss Algorithm or Wrapping)

Given a set of points in the plane. the convex hull of the set is the smallest convex polygon that contains all the points of it.

We strongly recommend to see the following post first.

How to check if two given line segments intersect?
The idea of Jarviss Algorithm is simple, we start from the leftmost point (or point with minimum x coordinate value) and we keep wrapping points
in counterclockwise direction. The big question is, given a point p as current point, how to find the next point in output? The idea is to use
orientation() here. Next point is selected as the point that beats all other points at counterclockwise orientation, i.e., next point is q if for any other
point r, we have orientation(p, r, q) = counterclockwise. Following is the detailed algorithm.
1) Initialize p as leftmost point.
2) Do following while we dont come back to the first (or leftmost) point.
..a) The next point q is the point such that the triplet (p, q, r) is counterclockwise for any other point r.
..b) next[p] = q (Store q as next of p in the output convex hull).
..c) p = q (Set p as q for next iteration).
// A C++ program to find convex hull of a set of points
// Refer http://www.geeksforgeeks.org/check-if-two-given-line-segments-intersect/
// for explanation of orientation()
#include <iostream>
using namespace std;
// Define Infinite (Using INT_MAX caused overflow problems)
#define INF 10000
struct Point
{
int x;
int y;
};
// To find orientation of ordered triplet (p, q, r).
// The function returns following values
// 0 --> p, q and r are colinear
// 1 --> Clockwise
// 2 --> Counterclockwise
int orientation(Point p, Point q, Point r)
{
int val = (q.y - p.y) * (r.x - q.x) (q.x - p.x) * (r.y - q.y);
if (val == 0) return 0; // colinear
return (val > 0)? 1: 2; // clock or counterclock wise
}
// Prints convex hull of a set of n points.
void convexHull(Point points[], int n)
{
// There must be at least 3 points
if (n < 3) return;
// Initialize Result
int next[n];
for (int i = 0; i < n; i++)
next[i] = -1;
// Find the leftmost point
int l = 0;
for (int i = 1; i < n; i++)
if (points[i].x < points[l].x)
l = i;
// Start from leftmost point, keep moving counterclockwise
// until reach the start point again
int p = l, q;
do
{
// Search for a point 'q' such that orientation(p, i, q) is

// counterclockwise for all points 'i'

q = (p+1)%n;
for (int i = 0; i < n; i++)
if (orientation(points[p], points[i], points[q]) == 2)
q = i;
next[p] = q; // Add q to result as a next point of p
p = q; // Set p as q for next iteration
} while (p != l);
// Print Result
for (int i = 0; i < n; i++)
{
if (next[i] != -1)
cout << "(" << points[i].x << ", " << points[i].y << ")\n";
}
}
// Driver program to test above functions
int main()
{
Point points[] = {{0, 3}, {2, 2}, {1, 1}, {2, 1},
{3, 0}, {0, 0}, {3, 3}};
int n = sizeof(points)/sizeof(points[0]);
convexHull(points, n);
return 0;
}

Output: The output is points of the convex hull.

(0,
(3,
(0,
(3,

3)
0)
0)
3)

Time Complexity: For every point on the hull we examine all the other points to determine the next point. Time complexity is ?(m * n) where n is
number of input points and m is number of output or hull points (m <= n). In worst case, time complexity is O(n 2). The worst case occurs when all
the points are on the hull (m = n)
We will soon be discussing other algorithms for finding convex hulls.
Sources:
http://www.cs.uiuc.edu/~jeffe/teaching/373/notes/x05-convexhull.pdf
http://www.dcs.gla.ac.uk/~pat/52233/slides/Hull1x1.pdf

Convex Hull | Set 2 (Graham Scan)

Given a set of points in the plane. the convex hull of the set is the smallest convex polygon that contains all the points of it.

We strongly recommend to see the following post first.

How to check if two given line segments intersect?
We have discussed Jarviss Algorithm for Convex Hull. Worst case time complexity of Jarviss Algorithm is O(n^2). Using Grahams scan algorithm,
we can find Convex Hull in O(nLogn) time. Following is Grahams algorithm
Let points[0..n-1] be the input array.
1) Find the bottom-most point by comparing y coordinate of all points. If there are two points with same y value, then the point with smaller x
coordinate value is considered. Put the bottom-most point at first position.
2) Consider the remaining n-1 points and sort them by polor angle in counterclockwise order around points[0]. If polor angle of two points is
same, then put the nearest point first.
3) Create an empty stack S and push points[0], points[1] and points[2] to S.
4) Process remaining n-3 points one by one. Do following for every point points[i]
4.1) Keep removing points from stack while orientation of following 3 points is not counterclockwise (or they dont make a left turn).
a) Point next to top in stack
b) Point at the top of stack
c) points[i]
4.2) Push points[i] to S
5) Print contents of S
The above algorithm can be divided in two phases.
Phase 1 (Sort points): We first find the bottom-most point. The idea is to pre-process points be sorting them with respect to the bottom-most
point. Once the points are sorted, they form a simple closed path (See following diagram).

What should be the sorting criteria? computation of actual angles would be inefficient since trigonometric functions are not simple to evaluate. The
idea is to use the orientation to compare angles without actually computing them (See the compare() function below)
Phase 2 (Accept or Reject Points): Once we have the closed path, the next step is to traverse the path and remove concave points on this path.
How to decide which point to remove and which to keep? Again, orientation helps here. The first two points in sorted array are always part of
Convex Hull. For remaining points, we keep track of recent three points, and find the angle formed by them. Let the three points be prev(p),
curr(c) and next(n). If orientation of these points (considering them in same order) is not counterclockwise, we discard c, otherwise we keep it.
Following diagram shows step by step process of this phase (Source of these diagrams is Ref 2).

Following is C++ implementation of the above algorithm.

// A C++ program to find convex hull of a set of points
// Refer http://www.geeksforgeeks.org/check-if-two-given-line-segments-intersect/
// for explanation of orientation()
#include <iostream>
#include <stack>
#include <stdlib.h>
using namespace std;
struct Point
{
int x;
int y;
};
// A globle point needed for sorting points with reference to the first point
// Used in compare function of qsort()
Point p0;
// A utility function to find next to top in a stack
Point nextToTop(stack<Point> &S)
{
Point p = S.top();
S.pop();
Point res = S.top();
S.push(p);
return res;
}
// A utility function to swap two points
int swap(Point &p1, Point &p2)
{
Point temp = p1;
p1 = p2;
p2 = temp;
}
// A utility function to return square of distance between p1 and p2
int dist(Point p1, Point p2)
{
return (p1.x - p2.x)*(p1.x - p2.x) + (p1.y - p2.y)*(p1.y - p2.y);
}
// To find orientation of ordered triplet (p, q, r).
// The function returns following values
// 0 --> p, q and r are colinear
// 1 --> Clockwise
// 2 --> Counterclockwise
int orientation(Point p, Point q, Point r)
{
int val = (q.y - p.y) * (r.x - q.x) (q.x - p.x) * (r.y - q.y);
if (val == 0) return 0; // colinear
return (val > 0)? 1: 2; // clock or counterclock wise
}

// A function used by library function qsort() to sort an array of

// points with respect to the first point
int compare(const void *vp1, const void *vp2)
{
Point *p1 = (Point *)vp1;
Point *p2 = (Point *)vp2;
// Find orientation
int o = orientation(p0, *p1, *p2);
if (o == 0)
return (dist(p0, *p2) >= dist(p0, *p1))? -1 : 1;
return (o == 2)? -1: 1;
}
// Prints convex hull of a set of n points.
void convexHull(Point points[], int n)
{
// Find the bottommost point
int ymin = points[0].y, min = 0;
for (int i = 1; i < n; i++)
{
int y = points[i].y;
// Pick the bottom-most or chose the left most point in case of tie
if ((y < ymin) || (ymin == y && points[i].x < points[min].x))
ymin = points[i].y, min = i;
}
// Place the bottom-most point at first position
swap(points[0], points[min]);
// Sort n-1 points with respect to the first point. A point p1 comes
// before p2 in sorted ouput if p2 has larger polar angle (in
// counterclockwise direction) than p1
p0 = points[0];
qsort(&points[1], n-1, sizeof(Point), compare);
// Create an empty stack and push first three points to it.
stack<Point> S;
S.push(points[0]);
S.push(points[1]);
S.push(points[2]);
// Process remaining n-3 points
for (int i = 3; i < n; i++)
{
// Keep removing top while the angle formed by points next-to-top,
// top, and points[i] makes a non-left turn
while (orientation(nextToTop(S), S.top(), points[i]) != 2)
S.pop();
S.push(points[i]);
}
// Now stack has the output points, print contents of stack
while (!S.empty())
{
Point p = S.top();
cout << "(" << p.x << ", " << p.y <<")" << endl;
S.pop();
}
}
// Driver program to test above functions
int main()
{
Point points[] = {{0, 3}, {1, 1}, {2, 2}, {4, 4},
{0, 0}, {1, 2}, {3, 1}, {3, 3}};
int n = sizeof(points)/sizeof(points[0]);
convexHull(points, n);
return 0;
}

Output:
(0,
(4,
(3,
(0,

3)
4)
1)
0)

Time Complexity: Let n be the number of input points. The algorithm takes O(nLogn) time if we use a O(nLogn) sorting algorithm.
The first step (finding the bottom-most point) takes O(n) time. The second step (sorting points) takes O(nLogn) time. In third step, every element
is pushed and popped at most one time. So the third step to process points one by one takes O(n) time, assuming that the stack operations take
O(1) time. Overall complexity is O(n) + O(nLogn) + O(n) which is O(nLogn)

Given n line segments, find if any two segments intersect

We have discussed the problem to detect if two given line segments intersect or not. In this post, we extend the problem. Here we are given n line
segments and we need to find out if any two line segments intersect or not.
Naive Algorithm A naive solution to solve this problem is to check every pair of lines and check if the pair intersects or not. We can check two
line segments in O(1) time. Therefore, this approach takes O(n2).
Sweep Line Algorithm: We can solve this problem in O(nLogn) time using Sweep Line Algorithm. The algorithm first sorts the end points along
the x axis from left to right, then it passes a vertical line through all points from left to right and checks for intersections. Following are detailed
steps.
1) Let there be n given lines. There must be 2n end points to represent the n lines. Sort all points according to x coordinates. While sorting maintain
a flag to indicate whether this point is left point of its line or right point.
2) Start from the leftmost point. Do following for every point
..a) If the current point is a left point of its line segment, check for intersection of its line segment with the segments just above and below it. And
add its line to active line segments (line segments for which left end point is seen, but right end point is not seen yet). Note that we consider only
those neighbors which are still active.
.b) If the current point is a right point, remove its line segment from active list and check whether its two active neighbors (points just above and
below) intersect with each other.
The step 2 is like passing a vertical line from all points starting from the leftmost point to the rightmost point. That is why this algorithm is called
Sweep Line Algorithm. The Sweep Line technique is useful in many other geometric algorithms like calculating the 2D Voronoi diagram
What data structures should be used for efficient implementation?
In step 2, we need to store all active line segments. We need to do following operations efficiently:
a) Insert a new line segment
b) Delete a line segment
c) Find predecessor and successor according to y coordinate values
The obvious choice for above operations is Self-Balancing Binary Search Tree like AVL Tree, Red Black Tree. With a Self-Balancing BST, we
can do all of the above operations in O(Logn) time.
Also, in step 1, instead of sorting, we can use min heap data structure. Building a min heap takes O(n) time and every extract min operation takes
O(Logn) time (See this).
PseudoCode:
The following pseudocode doesnt use heap. It simply sort the array.
sweepLineIntersection(Points[0..2n-1]):
1. Sort Points[] from left to right (according to x coordinate)
2. Create an empty Self-Balancing BST T. It will contain all active line
Segments ordered by y coordinate.
// Process all 2n points
3. for i = 0 to 2n-1
// If this point is left end of its line
if (Points[i].isLeft)
T.insert(Points[i].line()) // Insert into the tree
// Check if this points intersects with its predecessor and successor
if ( doIntersect(Points[i].line(), T.pred(Points[i].line()) )
return true
if ( doIntersect(Points[i].line(), T.succ(Points[i].line()) )
return true
else // If it's a right end of its line
// Check if its predecessor and successor intersect with each other
if ( doIntersect(T.pred(Points[i].line(), T.succ(Points[i].line()))
return true
T.delete(Points[i].line()) // Delete from tree
4. return False

Example:
Let us consider the following example taken from here. There are 5 line segments 1, 2, 3, 4 and 5. The dotted green lines show sweep lines.

Following are steps followed by the algorithm. All points from left to right are processed one by one. We maintain a self-balancing binary search
tree.
Left end point of line segment1 is processed: 1 is inserted into the Tree. The tree contains 1. No intersection.
Left end point of line segment2 is processed: Intersection of 1 and 2 is checked. 2 is inserted into the Tree. No intersection.The tree contains 1,
2.
Left end point of line segment 3is processed: Intersection of 3 with 1 is checked. No intersection.3 is inserted into the Tree. The tree contains 2,
1, 3.
Right end point of line segment1 is processed: 1 is deleted from the Tree. Intersection of 2 and 3 is checked. Intersection of 2 and 3 is
reported. The tree contains 2, 3. Note that the above pseudocode returns at this point. We can continue from here to report all intersection points.
Left end point of line segment4 is processed: Intersections of line 4 with lines 2 and 3 are checked. No intersection. 4 is inserted into the Tree.
The tree contains 2, 4, 3.
Left end point of line segment5 is processed: Intersection of 5 with 3 is checked. No intersection. 4 is inserted into the Tree. The tree contains
2,4, 3, 5.
Right end point of line segment5 is processed:5 is deleted from the Tree. The tree contains 2, 4, 3.
Right end point of line segment4 is processed:4 is deleted from the Tree. The tree contains 2, 4, 3. Intersection of 2 with3 is checked.
Intersection of 2 with 3 is reported. The tree contains 2, 3. Note that the intersection of 2 and 3 is reported again. We can add some logic to
check for duplicates.
Right end point of line segment2 and 3 are processed: Both are deleted from tree and tree becomes empty.
Time Complexity: The first step is sorting which takes O(nLogn) time. The second step process 2n points and for processing every point, it takes
O(Logn) time. Therefore, overall time complexity is O(nLogn)
References:
http://www.cs.uiuc.edu/~jeffe/teaching/373/notes/x06-sweepline.pdf
http://courses.csail.mit.edu/6.006/spring11/lectures/lec24.pdf
http://www.youtube.com/watch?v=dePDHVovJlE
http://www.eecs.wsu.edu/~cook/aa/lectures/l25/node10.html

Check whether a given point lies inside a triangle or not

Given three corner points of a triangle, and one more point P. Write a function to check whether P lies within the triangle or not.
For example, consider the following program, the function should return true for P(10, 15) and false for P'(30, 15)
B(10,30)
/ \
/ \
/
\
/ P \
P'
/
\
A(0,0) ----------- C(20,0)

Source: Microsoft Interview Question

Solution:
Let the coordinates of three corners be (x1, y1), (x2, y2) and (x3, y3). And coordinates of the given point P be (x, y)
1) Calculate area of the given triangle, i.e., area of the triangle ABC in the above diagram. Area A = [ x1(y2 y3) + x2(y3 y1) + x3(y1-y2)]/2
2) Calculate area of the triangle PAB. We can use the same formula for this. Let this area be A1.
3) Calculate area of the triangle PBC. Let this area be A2.
4) Calculate area of the triangle PAC. Let this area be A3.
5) If P lies inside the triangle, then A1 + A2 + A3 must be equal to A.
#include <stdio.h>
#include <stdlib.h>
/* A utility function to calculate area of triangle formed by (x1, y1),
(x2, y2) and (x3, y3) */
float area(int x1, int y1, int x2, int y2, int x3, int y3)
{
return abs((x1*(y2-y3) + x2*(y3-y1)+ x3*(y1-y2))/2.0);
}
/* A function to check whether point P(x, y) lies inside the triangle formed
by A(x1, y1), B(x2, y2) and C(x3, y3) */
bool isInside(int x1, int y1, int x2, int y2, int x3, int y3, int x, int y)
{
/* Calculate area of triangle ABC */
float A = area (x1, y1, x2, y2, x3, y3);
/* Calculate area of triangle PBC */
float A1 = area (x, y, x2, y2, x3, y3);
/* Calculate area of triangle PAC */
float A2 = area (x1, y1, x, y, x3, y3);
/* Calculate area of triangle PAB */
float A3 = area (x1, y1, x2, y2, x, y);
/* Check if sum of A1, A2 and A3 is same as A */
return (A == A1 + A2 + A3);
}
/* Driver program to test above function */
int main()
{
/* Let us check whether the point P(10, 15) lies inside the triangle
formed by A(0, 0), B(20, 0) and C(10, 30) */
if (isInside(0, 0, 20, 0, 10, 30, 10, 15))
printf ("Inside");
else
printf ("Not Inside");
return 0;
}

Ouptut:
Inside

Exercise: Given coordinates of four corners of a rectangle, and a point P. Write a function to check whether P lies inside the given rectangle or
not.

How to check if given four points form a square

Given coordinates of four points in a plane, find if the four points form a square or not.
To check for square, we need to check for following.
a) All fours sides formed by points are same.
b) The angle between any two sides is 90 degree. (This condition is required as Quadrilateral also has same sides.

The idea is to pick any point and calculate its distance from rest of the points. Let the picked picked point be p. To form a square, distance of two
points must be same from p, let this distance be d. The distance from one point must be different from that d and must be equal to ?2 times d. Let
this point with different distance be q.
The above condition is not good enough as the the point with different distance can be on the other side. We also need to check that q is at same
distance from 2 other points and this distance is same as d.
Below is C++ implementation of above idea.
// A C++ program to check if four given points form a square or not.
#include<iostream>
using namespace std;
// Structure of a point in 2D space
struct Point
{
int x, y;
};
// A utility function to find square of distance
// from point 'p' to poitn 'q'
int distSq(Point p, Point q)
{
return (p.x - q.x)*(p.x - q.x) +
(p.y - q.y)*(p.y - q.y);
}
// This function returns true
// square, otherwise false
bool isSquare(Point p1, Point
{
int d2 = distSq(p1, p2);
int d3 = distSq(p1, p3);
int d4 = distSq(p1, p4);
//
//
//
//
//
if
{

if (p1, p2, p3, p4) form a

p2, Point p3, Point p4)
// from p1 to p2
// from p1 to p3
// from p1 to p4

If lengths if (p1, p2) and (p1, p3) are same, then

following conditions must met to form a square.
1) Square of length of (p1, p4) is same as twice
the square of (p1, p2)
2) p4 is at same distance from p2 and p3
(d2 == d3 && 2*d2 == d4)
int d = distSq(p2, p4);
return (d == distSq(p3, p4) && d == d2);

}
// The below two cases are similar to above case
if (d3 == d4 && 2*d3 == d2)
{
int d = distSq(p2, p3);
return (d == distSq(p2, p4) && d == d3);
}
if (d2 == d4 && 2*d2 == d3)
{
int d = distSq(p2, p3);
return (d == distSq(p3, p4) && d == d2);
}
return false;
}

// Driver program to test above function

int main()
{
Point p1 = {20, 10}, p2 = {10, 20},
p3 = {20, 20}, p4 = {10, 10};
isSquare(p1, p2, p3, p4)? cout << "Yes": cout << "No";
return 0;
}

Output:
Yes

Write an Efficient Method to Check if a Number is Multiple of 3

The very first solution that comes to our mind is the one that we learned in school. If sum of digits in a number is multiple of 3 then number is
multiple of 3 e.g., for 612 sum of digits is 9 so its a multiple of 3. But this solution is not efficient. You have to get all decimal digits one by one, add
them and then check if sum is multiple of 3.
There is a pattern in binary representation of the number that can be used to find if number is a multiple of 3. If difference between count of odd set
bits (Bits set at odd positions) and even set bits is multiple of 3 then is the number.
Example: 23 (00..10111)
1) Get count of all set bits at odd positions (For 23 its 3).
2) Get count of all set bits at even positions (For 23 its 1).
3) If difference of above two counts is a multiple of 3 then number is also a multiple of 3.
(For 23 its 2 so 23 is not a multiple of 3)
Take some more examples like 21, 15, etc
Algorithm: isMutlipleOf3(n)
1) Make n positive if n is negative.
2) If number is 0 then return 1
3) If number is 1 then return 0
4) Initialize: odd_count = 0, even_count = 0
5) Loop while n != 0
a) If rightmost bit is set then increment odd count.
b) Right-shift n by 1 bit
c) If rightmost bit is set then increment even count.
d) Right-shift n by 1 bit
6) return isMutlipleOf3(odd_count - even_count)

Proof:
Above can be proved by taking the example of 11 in decimal numbers. (In this context 11 in decimal numbers is same as 3 in binary numbers)
If difference between sum of odd digits and even digits is multiple of 11 then decimal number is multiple of 11. Lets see how.
Lets take the example of 2 digit numbers in decimal
AB = 11A -A + B = 11A + (B A)
So if (B A) is a multiple of 11 then is AB.
Let us take 3 digit numbers.
ABC = 99A + A + 11B B + C = (99A + 11B) + (A + C B)
So if (A + C B) is a multiple of 11 then is (A+C-B)
Let us take 4 digit numbers now.
ABCD = 1001A + D + 11C C + 999B + B A
= (1001A 999B + 11C) + (D + B A -C )
So, if (B + D A C) is a multiple of 11 then is ABCD.
This can be continued for all decimal numbers.
Above concept can be proved for 3 in binary numbers in the same way.
Time Complexity: O(logn)
Program:
#include<stdio.h>
/* Fnction to check if n is a multiple of 3*/
int isMultipleOf3(int n)
{
int odd_count = 0;
int even_count = 0;
/* Make no positive if +n is multiple of 3
then is -n. We are doing this to avoid
stack overflow in recursion*/
if(n < 0) n = -n;
if(n == 0) return 1;
if(n == 1) return 0;
while(n)
{
/* If odd bit is set then

increment odd counter */

if(n & 1)
odd_count++;
n = n>>1;
/* If even bit is set then
increment even counter */
if(n & 1)
even_count++;
n = n>>1;
}
return isMultipleOf3(abs(odd_count - even_count));
}
/* Program to test function isMultipleOf3 */
int main()
{
int num = 23;
if (isMultipleOf3(num))
printf("num is multiple of 3");
else
printf("num is not a multiple of 3");
getchar();
return 0;
}

Efficient way to multiply with 7

We can multiply a number by 7 using bitwise operator. First left shift the number by 3 bits (you will get 8n) then subtract the original numberfrom
the shifted number and return the difference (8n n).
Program:
# include<stdio.h>
int multiplyBySeven(unsigned int n)
{
/* Note the inner bracket here. This is needed
because precedence of '-' operator is higher
than '<<' */
return ((n<<3) - n);
}
/* Driver program to test above function */
int main()
{
unsigned int n = 4;
printf("%u", multiplyBySeven(n));
getchar();
return 0;
}

Time Complexity: O(1)

Space Complexity: O(1)
Note: Works only for positive integers.
Same concept can be used for fast multiplication by 9 or other numbers.

Write a program to print all permutations of a given string

A permutation, also called an arrangement number or order, is a rearrangement of the elements of an ordered list S into a one-to-one
correspondence with S itself. A string of length n has n! permutation.
Source: Mathword(http://mathworld.wolfram.com/Permutation.html)
Below are the permutations of string ABC.
ABC, ACB, BAC, BCA, CAB, CBA
Here is a solution using backtracking.

C/C++
// C program to print all permutations with duplicates allowed
#include <stdio.h>
#include <string.h>
/* Function to swap values at two pointers */
void swap(char *x, char *y)
{
char temp;
temp = *x;
*x = *y;
*y = temp;
}
/* Function to print permutations of string
This function takes three parameters:
1. String
2. Starting index of the string
3. Ending index of the string. */
void permute(char *a, int l, int r)
{
int i;
if (l == r)
printf("%s\n", a);
else
{
for (i = l; i <= r; i++)
{
swap((a+l), (a+i));
permute(a, l+1, r);
swap((a+l), (a+i)); //backtrack
}
}
}
/* Driver program to test above functions */
int main()
{
char str[] = "ABC";
int n = strlen(str);
permute(str, 0, n-1);
return 0;
}

Python
# Python program to print all permutations with
# duplicates allowed

# Function to swap values

def swap(a,l,r):
t = a[l]
a[l] = a[r]
a[r] = t
return a
def toList(string):
List = []
for x in string:
List.append(x)
return List
def toString(List):
return ''.join(List)
# Function to print permutations of string
# This function takes three parameters:
# 1. String
# 2. Starting index of the string
# 3. Ending index of the string.
def permute(a, l, r):
if l==r:
print toString(a)
else:
for i in xrange(l,r+1):
a = swap(a,l,i)
permute(a, l+1, r)
a = swap(a,l,i) # backtrack
# Driver program to test the above function
string = "ABC"
n = len(string)
a = toList(string)
permute(a, 0, n-1)
# This code is contributed by Bhavya Jain

ABC
ACB
BAC
BCA
CBA
CAB

Algorithm Paradigm: Backtracking

Time Complexity: O(n*n!)

Lucky Numbers
Lucky numbers are subset of integers. Rather than going into much theory, let us see the process of arriving at lucky numbers,
Take the set of integers
1,2,3,4,5,6,7,8,9,10,11,12,14,15,16,17,18,19,
First, delete every second number, we get following reduced set.
1,3,5,7,9,11,13,15,17,19,
Now, delete every third number, we get
1, 3, 7, 9, 13, 15, 19,..
Continue this process indefinitely
Any number that does NOT get deleted due to above process is called lucky.
Therefore, set of lucky numbers is 1, 3, 7, 13,
Now, given an integer n, write a function to say whether this number is lucky or not.
bool isLucky(int n)

Algorithm:
Before every iteration, if we calculate position of the given no, then in a given iteration, we can determine if the no will be deleted. Suppose
calculated position for the given no. is P before some iteration, and each Ith no. is going to be removed in this iteration, if P < I then input no is
lucky, if P is such that P%I == 0 (I is a divisor of P), then input no is not lucky. Recursive Way:
#include <stdio.h>
#define bool int
/* Returns 1 if n is a lucky no. ohterwise returns 0*/
bool isLucky(int n)
{
static int counter = 2;
/*variable next_position is just for readability of
the program we can remove it and use n only */
int next_position = n;
if(counter > n)
return 1;
if(n%counter == 0)
return 0;
/*calculate next position of input no*/
next_position -= next_position/counter;
counter++;
return isLucky(next_position);
}
/*Driver function to test above function*/
int main()
{
int x = 5;
if( isLucky(x) )
printf("%d is a lucky no.", x);
else
printf("%d is not a lucky no.", x);
getchar();
}

Example:
Lets us take an example of 19
1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,15,17,18,19,20,21,
1,3,5,7,9,11,13,15,17,19,..
1,3,7,9,13,15,19,.
1,3,7,13,15,19,
1,3,7,13,19,
In next step every 6th no .in sequence will be deleted. 19 will not be deleted after this step because position of 19 is 5th after this step. Therefore,
19 is lucky. Lets see how above C code finds out:
Current function call

Position after this call

Counter for next call

Next Call

isLucky(19 )

10

isLucky(10)

isLucky(10)
isLucky(7)
isLucky(6)

7
6
5

4
5
6

isLucky(7)
isLucky(6)
isLucky(5)

When isLucky(6) is called, it returns 1 (because counter > n).

Iterative Way:
Please see this comment for another simple and elegant implementation of the above algorithm.

Write a program to add two numbers in base 14

Asked by Anshya.
Below are the different ways to add base 14 numbers.
Method 1
Thanks to Raj for suggesting this method.
1. Convert both i/p base 14 numbers to base 10.
2. Add numbers.
3. Convert the result back to base 14.

Method 2
Just add the numbers in base 14 in same way we add in base 10. Add numerals of both numbers one by one from right to left. If there is a carry
while adding two numerals, consider the carry for adding next numerals.
Let us consider the presentation of base 14 numbers same as hexadecimal numbers
A
B
C
D

-->
-->
-->
-->

10
11
12
13

Example:
num1 =
num2 =

1 2 A
C D 3

1. Add A and 3, we get 13(D). Since 13 is smaller than

14, carry becomes 0 and resultant numeral becomes D
2. Add 2, D and carry(0). we get 15. Since 15 is greater
than 13, carry becomes 1 and resultant numeral is 15 - 14 = 1
3. Add 1, C and carry(1). we get 14. Since 14 is greater
than 13, carry becomes 1 and resultant numeral is 14 - 14 = 0
Finally, there is a carry, so 1 is added as leftmost numeral and the result becomes
101D

Implementation of Method 2
# include <stdio.h>
# include <stdlib.h>
# define bool int
int getNumeralValue(char );
char getNumeral(int );
/* Function to add two numbers in base 14 */
char *sumBase14(char *num1, char *num2)
{
int l1 = strlen(num1);
int l2 = strlen(num2);
char *res;
int i;
int nml1, nml2, res_nml;
bool carry = 0;
if(l1 != l2)
{
printf("Function doesn't support numbers of different"
" lengths. If you want to add such numbers then"
" prefix smaller number with required no. of zeroes");
getchar();
assert(0);
}
/* Note the size of the allocated memory is one
more than i/p lenghts for the cases where we
have carry at the last like adding D1 and A1 */
res = (char *)malloc(sizeof(char)*(l1 + 1));
/* Add all numerals from right to left */
for(i = l1-1; i >= 0; i--)
{
/* Get decimal values of the numerals of

i/p numbers*/
nml1 = getNumeralValue(num1[i]);
nml2 = getNumeralValue(num2[i]);
/* Add decimal values of numerals and carry */
res_nml = carry + nml1 + nml2;
/* Check if we have carry for next addition
of numerals */
if(res_nml >= 14)
{
carry = 1;
res_nml -= 14;
}
else
{
carry = 0;
}
res[i+1] = getNumeral(res_nml);
}
/* if there is no carry after last iteration
then result should not include 0th character
of the resultant string */
if(carry == 0)
return (res + 1);
/* if we have carry after last iteration then
result should include 0th character */
res[0] = '1';
return res;
}
/* Function to get value of a numeral
For example it returns 10 for input 'A'
1 for '1', etc */
int getNumeralValue(char num)
{
if( num >= '0' && num <= '9')
return (num - '0');
if( num >= 'A' && num <= 'D')
return (num - 'A' + 10);
/* If we reach this line caller is giving
invalid character so we assert and fail*/
assert(0);
}
/* Function to get numeral for a value.
For example it returns 'A' for input 10
'1' for 1, etc */
char getNumeral(int val)
{
if( val >= 0 && val <= 9)
return (val + '0');
if( val >= 10 && val <= 14)
return (val + 'A' - 10);
/* If we reach this line caller is giving
invalid no. so we assert and fail*/
assert(0);
}
/*Driver program to test above functions*/
int main()
{
char *num1 = "DC2";
char *num2 = "0A3";
printf("Result is %s", sumBase14(num1, num2));
getchar();
return 0;
}

Notes:
Above approach can be used to add numbers in any base. We dont have to do string operations if base is smaller than 10.
You can try extending the above program for numbers of different lengths.
Please comment if you find any bug in the program or a better approach to do the same.

Babylonian method for square root

Algorithm:
This method can be derived from (but predates) NewtonRaphson method.
1 Start with an arbitrary positive start value x (the closer to the
root, the better).
2 Initialize y = 1.
3. Do following until desired approximation is achieved.
a) Get the next approximation for root using average of x and y
b) Set y = n/x

Implementation:
/*Returns the square root of n. Note that the function */
float squareRoot(float n)
{
/*We are using n itself as initial approximation
This can definitely be improved */
float x = n;
float y = 1;
float e = 0.000001; /* e decides the accuracy level*/
while(x - y > e)
{
x = (x + y)/2;
y = n/x;
}
return x;
}
/* Driver program to test above function*/
int main()
{
int n = 50;
printf ("Square root of %d is %f", n, squareRoot(n));
getchar();
}

Example:
n = 4 /*n itself is used for initial approximation*/
Initialize x = 4, y = 1
Next Approximation x = (x + y)/2 (= 2.500000),
y = n/x (=1.600000)
Next Approximation x = 2.050000,
y = 1.951220
Next Approximation x = 2.000610,
y = 1.999390
Next Approximation x = 2.000000,
y = 2.000000
Terminate as (x - y) > e now.

If we are sure that n is a perfect square, then we can use following method. The method can go in infinite loop for non-perfect-square numbers.
For example, for 3 the below while loop will never terminate.
/*Returns the square root of n. Note that the function
will not work for numbers which are not perfect squares*/
unsigned int squareRoot(int n)
{
int x = n;
int y = 1;
while(x > y)
{
x = (x + y)/2;
y = n/x;
}
return x;
}
/* Driver program to test above function*/
int main()
{
int n = 49;
printf (" root of %d is %d", n, squareRoot(n));
getchar();
}

References;
http://en.wikipedia.org/wiki/Square_root
http://en.wikipedia.org/wiki/Babylonian_method#Babylonian_method
Asked by Snehal

Multiply two integers without using multiplication, division and bitwise operators,
and no loops
Asked by Kapil
By making use of recursion, we can multiply two integers with the given constraints.
To multiply x and y, recursively add x y times.
Thanks to geek4u for suggesting this method.
#include<stdio.h>
/* function to multiply two numbers x and y*/
int multiply(int x, int y)
{
/* 0 multiplied with anything gives 0 */
if(y == 0)
return 0;
/* Add x one by one */
if(y > 0 )
return (x + multiply(x, y-1));
/* the case where y is negative */
if(y < 0 )
return -multiply(x, -y);
}
int main()
{
printf("\n %d", multiply(5, -11));
getchar();
return 0;
}

Time Complexity: O(y) where y is the second argument to function multiply().

Print all combinations of points that can compose a given number

You can win three kinds of basketball points, 1 point, 2 points, and 3 points. Given a total score n, print out all the combination to compose n.
Examples:
For n = 1, the program should print following:
1
For n = 2, the program should print following:
11
2
For n = 3, the program should print following:
111
12
21
3
For n = 4, the program should print following:
1111
112
121
13
211
22
31
and so on
Algorithm:
At first position we can have three numbers 1 or 2 or 3.
First put 1 at first position and recursively call for n-1.
Then put 2 at first position and recursively call for n-2.
Then put 3 at first position and recursively call for n-3.
If n becomes 0 then we have formed a combination that compose n, so print the current combination.
Below is a generalized implementation. In the below implementation, we can change MAX_POINT if there are higher points (more than 3) in the
basketball game.
#define MAX_POINT 3
#define ARR_SIZE 100
#include<stdio.h>
/* Utility function to print array arr[] */
void printArray(int arr[], int arr_size);
/* The function prints all combinations of numbers 1, 2, ...MAX_POINT
that sum up to n.
i is used in recursion keep track of index in arr[] where next
element is to be added. Initital value of i must be passed as 0 */
void printCompositions(int n, int i)
{
/* array must be static as we want to keep track
of values stored in arr[] using current calls of
printCompositions() in function call stack*/
static int arr[ARR_SIZE];
if (n == 0)
{
printArray(arr, i);
}
else if(n > 0)
{
int k;
for (k = 1; k <= MAX_POINT; k++)
{
arr[i]= k;
printCompositions(n-k, i+1);
}
}
}

/* UTILITY FUNCTIONS */
/* Utility function to print array arr[] */
void printArray(int arr[], int arr_size)
{
int i;
for (i = 0; i < arr_size; i++)
printf("%d ", arr[i]);
printf("\n");
}
/* Driver function to test above functions */
int main()
{
int n = 5;
printf("Differnt compositions formed by 1, 2 and 3 of %d are\n", n);
printCompositions(n, 0);
getchar();
return 0;
}

Asked by Aloe

Write you own Power without using multiplication(*) and division(/) operators
Method 1 (Using Nested Loops)
We can calculate power by using repeated addition.
For example to calculate 5^6.
1) First 5 times add 5, we get 25. (5^2)
2) Then 5 times add 25, we get 125. (5^3)
3) Then 5 time add 125, we get 625 (5^4)
4) Then 5 times add 625, we get 3125 (5^5)
5) Then 5 times add 3125, we get 15625 (5^6)
/* Works only if a >= 0 and b >= 0 */
int pow(int a, int b)
{
if (b == 0)
return 1;
int answer = a;
int increment = a;
int i, j;
for(i = 1; i < b; i++)
{
for(j = 1; j < a; j++)
{
answer += increment;
}
increment = answer;
}
return answer;
}
/* driver program to test above function */
int main()
{
printf("\n %d", pow(5, 3));
getchar();
return 0;
}

Method 2 (Using Recursion)

Recursively add a to get the multiplication of two numbers. And recursively multiply to get a raise to the power b.
#include<stdio.h>
/* A recursive function to get a^b
Works only if a >= 0 and b >= 0 */
int pow(int a, int b)
{
if(b)
return multiply(a, pow(a, b-1));
else
return 1;
}
/* A recursive function to get x*y */
int multiply(int x, int y)
{
if(y)
return (x + multiply(x, y-1));
else
return 0;
}
/* driver program to test above functions */
int main()
{
printf("\n %d", pow(5, 3));
getchar();
return 0;
}

Program for Fibonacci numbers

The Fibonacci numbers are the numbers in the following integer sequence.
0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89, 141, ..
In mathematical terms, the sequence Fn of Fibonacci numbers is defined by the recurrence relation
Fn = Fn-1 + Fn-2

with seed values

F0 = 0 and F1 = 1.

Write a function int fib(int n) that returns Fn. For example, if n = 0, then fib() should return 0. If n = 1, then it should return 1. For n > 1, it should
return Fn-1 + Fn-2
Following are different methods to get the nth Fibonacci number.
Method 1 ( Use recursion )
A simple method that is a direct recusrive implementation mathematical recurance relation given above.
#include<stdio.h>
int fib(int n)
{
if (n <= 1)
return n;
return fib(n-1) + fib(n-2);
}
int main ()
{
int n = 9;
printf("%d", fib(n));
getchar();
return 0;
}

Time Complexity: T(n) = T(n-1) + T(n-2) which is exponential.

We can observe that this implementation does a lot of repeated work (see the following recursion tree). So this is a bad implementation for nth
Fibonacci number.
fib(5)
/
\
fib(4)
fib(3)
/
\
/
\
fib(3)
fib(2)
fib(2)
fib(1)
/
\
/
\
/
\
fib(2) fib(1) fib(1) fib(0) fib(1) fib(0)
/
\
fib(1) fib(0)

Extra Space: O(n) if we consider the function call stack size, otherwise O(1).
Method 2 ( Use Dynamic Programming )
We can avoid the repeated work done is the method 1 by storing the Fibonacci numbers calculated so far.
#include<stdio.h>
int fib(int n)
{
/* Declare an array to store Fibonacci numbers. */
int f[n+1];
int i;
/* 0th and 1st number of the series are 0 and 1*/
f[0] = 0;
f[1] = 1;
for (i = 2; i <= n; i++)
{
/* Add the previous 2 numbers in the series
and store it */
f[i] = f[i-1] + f[i-2];
}

return f[n];
}
int main ()
{
int n = 9;
printf("%d", fib(n));
getchar();
return 0;
}

Time Complexity: O(n)

Extra Space: O(n)
Method 3 ( Space Otimized Method 2 )
We can optimize the space used in method 2 by storing the previous two numbers only because that is all we need to get the next Fibannaci
number in series.
#include<stdio.h>
int fib(int n)
{
int a = 0, b = 1, c, i;
if( n == 0)
return a;
for (i = 2; i <= n; i++)
{
c = a + b;
a = b;
b = c;
}
return b;
}
int main ()
{
int n = 9;
printf("%d", fib(n));
getchar();
return 0;
}

Time Complexity: O(n)

Extra Space: O(1)
Method 4 ( Using power of the matrix {{1,1},{1,0}} )
This another O(n) which relies on the fact that if we n times multiply the matrix M = {{1,1},{1,0}} to itself (in other words calculate power(M, n
)), then we get the (n+1)th Fibonacci number as the element at row and column (0, 0) in the resultant matrix.
The matrix representation gives the following closed expression for the Fibonacci numbers:

#include <stdio.h>
/* Helper function that multiplies 2 matricies F and M of size 2*2, and
puts the multiplication result back to F[][] */
void multiply(int F[2][2], int M[2][2]);
/* Helper function that calculates F[][] raise to the power n and puts the
result in F[][]
Note that this function is desinged only for fib() and won't work as general
power function */
void power(int F[2][2], int n);
int fib(int n)
{
int F[2][2] = {{1,1},{1,0}};
if (n == 0)
return 0;
power(F, n-1);
return F[0][0];
}
void multiply(int F[2][2], int M[2][2])
{
int x = F[0][0]*M[0][0] + F[0][1]*M[1][0];
int y = F[0][0]*M[0][1] + F[0][1]*M[1][1];

int z = F[1][0]*M[0][0] + F[1][1]*M[1][0];

int w = F[1][0]*M[0][1] + F[1][1]*M[1][1];
F[0][0]
F[0][1]
F[1][0]
F[1][1]

=
=
=
=

x;
y;
z;
w;

}
void power(int F[2][2], int n)
{
int i;
int M[2][2] = {{1,1},{1,0}};
// n - 1 times multiply the matrix to {{1,0},{0,1}}
for (i = 2; i <= n; i++)
multiply(F, M);
}
/* Driver program to test above function */
int main()
{
int n = 9;
printf("%d", fib(n));
getchar();
return 0;
}

Time Complexity: O(n)

Extra Space: O(1)
Method 5 ( Optimized Method 4 )
The method 4 can be optimized to work in O(Logn) time complexity. We can do recursive multiplication to get power(M, n) in the prevous
method (Similar to the optimization done in this post)
#include <stdio.h>
void multiply(int F[2][2], int M[2][2]);
void power(int F[2][2], int n);
/* function that returns nth Fibonacci number */
int fib(int n)
{
int F[2][2] = {{1,1},{1,0}};
if (n == 0)
return 0;
power(F, n-1);
return F[0][0];
}
/* Optimized version of power() in method 4 */
void power(int F[2][2], int n)
{
if( n == 0 || n == 1)
return;
int M[2][2] = {{1,1},{1,0}};
power(F, n/2);
multiply(F, F);
if (n%2 != 0)
multiply(F, M);
}
void multiply(int F[2][2],
{
int x = F[0][0]*M[0][0]
int y = F[0][0]*M[0][1]
int z = F[1][0]*M[0][0]
int w = F[1][0]*M[0][1]
F[0][0]
F[0][1]
F[1][0]
F[1][1]

=
=
=
=

int M[2][2])
+
+
+
+

F[0][1]*M[1][0];
F[0][1]*M[1][1];
F[1][1]*M[1][0];
F[1][1]*M[1][1];

x;
y;
z;
w;

}
/* Driver program to test above function */

int main()
{
int n = 9;
printf("%d", fib(9));
getchar();
return 0;
}

Time Complexity: O(Logn)

Extra Space: O(Logn) if we consider the function call stack size, otherwise O(1).
References:
http://en.wikipedia.org/wiki/Fibonacci_number
http://www.ics.uci.edu/~eppstein/161/960109.html

Average of a stream of numbers

Difficulty Level: Rookie
Given a stream of numbers, print average (or mean) of the stream at every point. For example, let us consider the stream as 10, 20, 30, 40, 50,
60,
Average of 1 numbers
Average of 2 numbers
Average of 3 numbers
Average of 4 numbers
Average of 5 numbers
Average of 6 numbers
..................

is
is
is
is
is
is

10.00
15.00
20.00
25.00
30.00
35.00

To print mean of a stream, we need to find out how to find average when a new number is being added to the stream. To do this, all we need is
count of numbers seen so far in the stream, previous average and new number. Let n be the count, prev_avg be the previous average and x be the
new number being added. The average after including x number can be written as (prev_avg*n + x)/(n+1).
#include <stdio.h>
// Returns the new average after including x
float getAvg(float prev_avg, int x, int n)
{
return (prev_avg*n + x)/(n+1);
}
// Prints average of a stream of numbers
void streamAvg(float arr[], int n)
{
float avg = 0;
for(int i = 0; i < n; i++)
{
avg = getAvg(avg, arr[i], i);
printf("Average of %d numbers is %f \n", i+1, avg);
}
return;
}
// Driver program to test above functions
int main()
{
float arr[] = {10, 20, 30, 40, 50, 60};
int n = sizeof(arr)/sizeof(arr[0]);
streamAvg(arr, n);
return 0;
}

The above function getAvg() can be optimized using following changes. We can avoid the use of prev_avg and number of elements by using static
variables (Assuming that only this function is called for average of stream). Following is the oprimnized version.
#include <stdio.h>
// Returns the new average after including x
float getAvg (int x)
{
static int sum, n;
sum += x;
return (((float)sum)/++n);
}
// Prints average of a stream of numbers
void streamAvg(float arr[], int n)
{
float avg = 0;
for(int i = 0; i < n; i++)
{
avg = getAvg(arr[i]);
printf("Average of %d numbers is %f \n", i+1, avg);
}
return;
}
// Driver program to test above functions
int main()
{
float arr[] = {10, 20, 30, 40, 50, 60};

int n = sizeof(arr)/sizeof(arr[0]);
streamAvg(arr, n);
return 0;
}

Thanks to Abhijeet Deshpande for suggesting this optimized version.

Count numbers that dont contain 3

Given a number n, write a function that returns count of numbers from 1 to n that dont contain digit 3 in their decimal representation.
Examples:
Input: n = 10
Output: 9
Input: n = 45
Output: 31
// Numbers 3, 13, 23, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 43 contain digit 3.
Input: n = 578
Ouput: 385

Solution:
We can solve it recursively. Let count(n) be the function that counts such numbers.
'msd' --> the most significant digit in n
'd' --> number of digits in n.
count(n) = n if n < 3
count(n) = n - 1 if 3 <= n < 10
count(n) = count(msd) * count(10^(d-1) - 1) +
count(msd) +
count(n % (10^(d-1)))
if n > 10 and msd is not 3
count(n) = count( msd * (10^(d-1)) - 1)
if n > 10 and msd is 3
Let us understand the solution with n = 578.
count(578) = 4*count(99) + 4 + count(78)
The middle term 4 is added to include numbers 100, 200, 400 and 500.
Let us take n = 35 as another example.
count(35) = count (3*10 - 1) = count(29)
#include <stdio.h>
/* returns count of numbers which are in range from 1 to n and don't contain 3
as a digit */
int count(int n)
{
// Base cases (Assuming n is not negative)
if (n < 3)
return n;
if (n >= 3 && n < 10)
return n-1;
// Calculate 10^(d-1) (10 raise to the power d-1) where d is
// number of digits in n. po will be 100 for n = 578
int po = 1;
while (n/po > 9)
po = po*10;
// find the most significant digit (msd is 5 for 578)
int msd = n/po;
if (msd != 3)
// For 578, total will be 4*count(10^2 - 1) + 4 + count(78)
return count(msd)*count(po - 1) + count(msd) + count(n%po);
else
// For 35, total will be equal to count(29)
return count(msd*po - 1);
}
// Driver program to test above function
int main()
{
printf ("%d ", count(578));
return 0;
}

Output:
385

Magic Square
A magic square of order n is an arrangement of n^2 numbers, usually distinct integers, in a square, such that the n numbers in all rows, all columns,
and both diagonals sum to the same constant. A magic square contains the integers from 1 to n^2.
The constant sum in every row, column and diagonal is called the magic constant or magic sum, M. The magic constant of a normal magic square
depends only on n and has the following value:
M = n(n^2+1)/2
For normal magic squares of order n = 3, 4, 5, , the magic constants are: 15, 34, 65, 111, 175, 260,
In this post, we will discuss how programmatically we can generate a magic square of size n. Before we go further, consider the below examples:
Magic Square of size 3
----------------------2 7 6
9 5 1
4 3 8
Sum in each row & each column = 3*(3^2+1)/2 = 15
Magic Square of size 5
---------------------9 3 22 16 15
2 21 20 14 8
25 19 13 7 1
18 12 6 5 24
11 10 4 23 17
Sum in each row & each column = 5*(5^2+1)/2 = 65
Magic Square of size 7
---------------------20 12 4 45 37 29 28
11 3 44 36 35 27 19
2 43 42 34 26 18 10
49 41 33 25 17 9 1
40 32 24 16 8 7 48
31 23 15 14 6 47 39
22 21 13 5 46 38 30
Sum in each row & each column = 7*(7^2+1)/2 = 175

Did you find any pattern in which the numbers are stored?
In any magic square, the first number i.e. 1 is stored at position (n/2, n-1). Let this position be (i,j). The next number is stored at position (i-1, j+1)
where we can consider each row & column as circular array i.e. they wrap around.
Three conditions hold:
1. The position of next number is calculated by decrementing row number of previous number by 1, and incrementing the column number of
previous number by 1. At any time, if the calculated row position becomes -1, it will wrap around to n-1. Similarly, if the calculated column
position becomes n, it will wrap around to 0.
2. If the magic square already contains a number at the calculated position, calculated column position will be decremented by 2, and calculated
row position will be incremented by 1.
3. If the calculated row position is -1 & calculated column position is n, the new position would be: (0, n-2).
Example:
Magic Square of size 3
---------------------2 7 6
9 5 1
4 3 8
Steps:
1. position of number 1 = (3/2, 3-1) = (1, 2)
2. position of number 2 = (1-1, 2+1) = (0, 0)
3. position of number 3 = (0-1, 0+1) = (3-1, 1) = (2, 1)
4. position of number 4 = (2-1, 1+1) = (1, 2)
Since, at this position, 1 is there. So, apply condition 2.
new position=(1+1,2-2)=(2,0)
5. position of number 5=(2-1,0+1)=(1,1)
6. position of number 6=(1-1,1+1)=(0,2)
7. position of number 7 = (0-1, 2+1) = (-1,3) // this is tricky, see condition 3
new position = (0, 3-2) = (0,1)
8. position of number 8=(0-1,1+1)=(-1,2)=(2,2) //wrap around

9. position of number 9=(2-1,2+1)=(1,3)=(1,0) //wrap around

Based on the above approach, following is the working code:

#include<stdio.h>
#include<string.h>
// A function to generate odd sized magic squares
void generateSquare(int n)
{
int magicSquare[n][n];
// set all slots as 0
memset(magicSquare, 0, sizeof(magicSquare));
// Initialize position for 1
int i = n/2;
int j = n-1;
// One by one put all values in magic square
for (int num=1; num <= n*n; )
{
if (i==-1 && j==n) //3rd condition
{
j = n-2;
i = 0;
}
else
{
//1st condition helper if next number goes to out of square's right side
if (j == n)
j = 0;
//1st condition helper if next number is goes to out of square's upper side
if (i < 0)
i=n-1;
}
if (magicSquare[i][j]) //2nd condition
{
j -= 2;
i++;
continue;
}
else
magicSquare[i][j] = num++; //set number
j++; i--; //1st condition
}
// print magic square
printf("The Magic Square for n=%d:\nSum of each row or column %d:\n\n",
n, n*(n*n+1)/2);
for(i=0; i<n; i++)
{
for(j=0; j<n; j++)
printf("%3d ", magicSquare[i][j]);
printf("\n");
}
}
// Driver program to test above function
int main()
{
int n = 7; // Works only when n is odd
generateSquare (n);
return 0;
}

Output:
The Magic Square for n=7:
Sum of each row or column 175:
20
11
2
49
40
31
22

12
3
43
41
32
23
21

4
44
42
33
24
15
13

45
36
34
25
16
14
5

37
35
26
17
8
6
46

29
27
18
9
7
47
38

28
19
10
1
48
39
30

NOTE: This approach works only for odd values of n.

References:
http://en.wikipedia.org/wiki/Magic_square

Sieve of Eratosthenes
Given a number n, print all primes smaller than or equal to n. It is also given that n is a small number.
For example, if n is 10, the output should be 2, 3, 5, 7?. If n is 20, the output should be 2, 3, 5, 7, 11, 13, 17, 19?.
The sieve of Eratosthenes is one of the most efficient ways to find all primes smaller than n when n is smaller than 10 million or so (Ref Wiki).
Following is the algorithm to find all the prime numbers less than or equal to a given integernby Eratosthenes method:
1. Create a list of consecutive integers from 2 ton: (2, 3, 4, ,n).
2. Initially, letpequal 2, the first prime number.
3. Starting fromp, count up in increments ofpand mark each of these numbers greater thanpitself in the list. These numbers will be 2p, 3p, 4p,
etc.; note that some of them may have already been marked.
4. Find the first number greater thanpin the list that is not marked. If there was no such number, stop. Otherwise, letpnow equal this number
(which is the next prime), and repeat from step 3.
When the algorithm terminates, all the numbers in the list that are not marked are prime.
Explanation with Example:
Let us take an example when n = 50. So we need to print all print numbers smaller than or equal to 50.
We create a list of all numbers from 2 to 50.

According to the algorithm we will mark all the numbers which are divisible by 2.

Now we move to our next unmarked number 3 and mark all the numbers which are multiples of 3.

We move to our next unmarked number 5 and mark all multiples of 5.

We continue this process and our final table will look like below:

So the prime numbers are the unmarked ones: 2, 3, 5, 7, 11, 13, 17, 19, 23, 29, 31, 37, 41, 43, 47.
Thanks to Krishan Kumar for providing above explanation.
Implementation:
Following is C++ implementation of the above algorithm. In the following implementation, a boolean array arr[] of size n is used to mark multiples
of prime numbers.
#include <stdio.h>
#include <string.h>
// marks all mutiples of 'a' ( greater than 'a' but less than equal to 'n') as 1.
void markMultiples(bool arr[], int a, int n)
{
int i = 2, num;
while ( (num = i*a) <= n )
{

arr[ num-1 ] = 1; // minus 1 because index starts from 0.

++i;
}
}
// A function to print all prime numbers smaller than n
void SieveOfEratosthenes(int n)
{
// There are no prime numbers smaller than 2
if (n >= 2)
{
// Create an array of size n and initialize all elements as 0
bool arr[n];
memset(arr, 0, sizeof(arr));
/* Following property is maintained in the below for loop
arr[i] == 0 means i + 1 is prime
arr[i] == 1 means i + 1 is not prime */
for (int i=1; i<n; ++i)
{
if ( arr[i] == 0 )
{
//(i+1) is prime, print it and mark its multiples
printf("%d ", i+1);
markMultiples(arr, i+1, n);
}
}
}
}
// Driver Program to test above function
int main()
{
int n = 30;
printf("Following are the prime numbers below %d\n", n);
SieveOfEratosthenes(n);
return 0;
}

Output:
Following are the prime numbers below 30
2 3 5 7 11 13 17 19 23 29

References:
http://en.wikipedia.org/wiki/Sieve_of_Eratosthenes

Find day of the week for a given date

Write a function that calculates the day of the week for any particular date in the past or future. A typical application is to calculate the day of the
week on which someone was born or some other special event occurred.
Following is a simple C function suggested by Sakamoto, Lachman, Keith and Craver to calculate day. The following function returns 0 for
Sunday, 1 for Monday, etc.
/* A program to find day of a given date */
#include<stdio.h>
int dayofweek(int d, int m, int y)
{
static int t[] = { 0, 3, 2, 5, 0, 3, 5, 1, 4, 6, 2, 4 };
y -= m < 3;
return ( y + y/4 - y/100 + y/400 + t[m-1] + d) % 7;
}
/* Driver function to test above function */
int main()
{
int day = dayofweek(30, 8, 2010);
printf ("%d", day);
return 0;
}

Output: 1 (Monday)
See this for explanation of the above function.
References:
http://en.wikipedia.org/wiki/Determination_of_the_day_of_the_week

DFA based division

Deterministic Finite Automaton (DFA) can be used to check whether a number num is divisible by k or not. If the number is not divisible,
remainder can also be obtained using DFA.
We consider the binary representation of num and build a DFA with k states. The DFA has transition function for both 0 and 1. Once the DFA is
built, we process num over the DFA to get remainder.
Let us walk through an example. Suppose we want to check whether a given number num is divisible by 3 or not. Any number can be written in
the form: num = 3*a + b where a is the quotient and b is the remainder.
For 3, there can be 3 states in DFA, each corresponding to remainder 0, 1 and 2. And each state can have two transitions corresponding 0 and 1
(considering the binary representation of given num).
The transition function F(p, x) = q tells that on reading alphabet x, we move from state p to state q. Let us name the states as 0, 1 and 2. The initial
state will always be 0. The final state indicates the remainder. If the final state is 0, the number is divisible.

In the above diagram, double circled state is final state.

1. When we are at state 0 and read 0, we remain at state 0.
2. When we are at state 0 and read 1, we move to state 1, why? The number so formed(1) in decimal gives remainder 1.
3. When we are at state 1 and read 0, we move to state 2, why? The number so formed(10) in decimal gives remainder 2.
4. When we are at state 1 and read 1, we move to state 0, why? The number so formed(11) in decimal gives remainder 0.
5. When we are at state 2 and read 0, we move to state 1, why? The number so formed(100) in decimal gives remainder 1.
6. When we are at state 2 and read 1, we remain at state 2, why? The number so formed(101) in decimal gves remainder 2.
The transition table looks like following:
state 0 1
_____________
0
0 1
1
2 0
2
1 2

Let us check whether 6 is divisible by 3?

Binary representation of 6 is 110
state = 0
1. state=0, we read 1, new state=1
2. state=1, we read 1, new state=0
3. state=0, we read 0, new state=0
Since the final state is 0, the number is divisible by 3.
Let us take another example number as 4
state=0
1. state=0, we read 1, new state=1
2. state=1, we read 0, new state=2
3. state=2, we read 0, new state=1
Since, the final state is not 0, the number is not divisible by 3. The remainder is 1.
Note that the final state gives the remainder.
We can extend the above solution for any value of k. For a value k, the states would be 0, 1, . , k-1. How to calculate the transition if the decimal
equivalent of the binary bits seen so far, crosses the range k? If we are at state p, we have read p (in decimal). Now we read 0, new read number
becomes 2*p. If we read 1, new read number becomes 2*p+1. The new state can be obtained by subtracting k from these values (2p or 2p+1)
where 0 <= p < k. Based on the above approach, following is the working code:
#include <stdio.h>
#include <stdlib.h>
// Function to build DFA for divisor k
void preprocess(int k, int Table[][2])
{
int trans0, trans1;
// The following loop calculates the two transitions for each state,

// starting from state 0

for (int state=0; state<k; ++state)
{
// Calculate next state for bit 0
trans0 = state<<1;
Table[state][0] = (trans0 < k)? trans0: trans0-k;
// Calculate next state for bit 1
trans1 = (state<<1) + 1;
Table[state][1] = (trans1 < k)? trans1: trans1-k;
}
}
// A recursive utility function that takes a num and DFA (transition
// table) as input and process num bit by bit over DFA
void isDivisibleUtil(int num, int* state, int Table[][2])
{
// process "num" bit by bit from MSB to LSB
if (num != 0)
{
isDivisibleUtil(num>>1, state, Table);
*state = Table[*state][num&1];
}
}
// The main function that divides num by k and returns the remainder
int isDivisible (int num, int k)
{
// Allocate memory for transition table. The table will have k*2 entries
int (*Table)[2] = (int (*)[2])malloc(k*sizeof(*Table));
// Fill the transition table
preprocess(k, Table);
// Process num over DFA and get the remainder
int state = 0;
isDivisibleUtil(num, &state, Table);
// Note that the final value of state is the remainder
return state;
}
// Driver program to test above functions
int main()
{
int num = 47; // Number to be divided
int k = 5; // Divisor
int remainder = isDivisible (num, k);
if (remainder == 0)
printf("Divisible\n");
else
printf("Not Divisible: Remainder is %d\n", remainder);
return 0;
}

Output:
Not Divisible: Remainder is 2

DFA based division can be useful if we have a binary stream as input and we want to check for divisibility of the decimal value of stream at any
time.

Generate integer from 1 to 7 with equal probability

Given a function foo() that returns integers from 1 to 5 with equal probability, write a function that returns integers from 1 to 7 with equal
probability using foo() only. Minimize the number of calls to foo() method. Also, use of any other library function is not allowed and no floating
point arithmetic allowed.
Solution:
We know foo() returns integers from 1 to 5. How we can ensure that integers from 1 to 7 occur with equal probability?
If we somehow generate integers from 1 to a-multiple-of-7 (like 7, 14, 21, ) with equal probability, we can use modulo division by 7 followed by
adding 1 to get the numbers from 1 to 7 with equal probability.
We can generate from 1 to 21 with equal probability using the following expression.
5*foo() + foo() -5

Let us see how above expression can be used.

1. For each value of first foo(), there can be 5 possible combinations for values of second foo(). So, there are total 25 combinations possible.
2. The range of values returned by the above equation is 1 to 25, each integer occurring exactly once.
3. If the value of the equation comes out to be less than 22, return modulo division by 7 followed by adding 1. Else, again call the method
recursively. The probability of returning each integer thus becomes 1/7.
The below program shows that the expression returns each integer from 1 to 25 exactly once.
#include <stdio.h>
int main()
{
int first, second;
for ( first=1; first<=5; ++first )
for ( second=1; second<=5; ++second )
printf ("%d \n", 5*first + second - 5);
return 0;
}

Output:
1
2
.
.
24
25

The below program depicts how we can use foo() to return 1 to 7 with equal probability.
#include <stdio.h>
int foo() // given method that returns 1 to 5 with equal probability
{
// some code here
}
int my_rand() // returns 1 to 7 with equal probability
{
int i;
i = 5*foo() + foo() - 5;
if (i < 22)
return i%7 + 1;
return my_rand();
}
int main()
{
printf ("%d ", my_rand());
return 0;
}

Given a number, find the next smallest palindrome

Given a number, find the next smallest palindrome larger than this number. For example, if the input number is 2 3 5 4 5?, the output should be 2 3
6 3 2?. And if the input number is 9 9 9?, the output should be 1 0 0 1?.
The input is assumed to be an array. Every entry in array represents a digit in input number. Let the array be num[] and size of array be n
There can be three different types of inputs that need to be handled separately.
1) The input number is palindrome and has all 9s. For example 9 9 9?. Output should be 1 0 0 1?
2) The input number is not palindrome. For example 1 2 3 4?. Output should be 1 3 3 1?
3) The input number is palindrome and doesnt have all 9s. For example 1 2 2 1?. Output should be 1 3 3 1?.
Solution for input type 1 is easy. The output contains n + 1 digits where the corner digits are 1, and all digits between corner digits are 0.
Now let us first talk about input type 2 and 3. How to convert a given number to a greater palindrome? To understand the solution, let us first
define the following two terms:
Left Side: The left half of given number. Left side of 1 2 3 4 5 6? is 1 2 3? and left side of 1 2 3 4 5? is 1 2?
Right Side: The right half of given number. Right side of 1 2 3 4 5 6? is 4 5 6? and right side of 1 2 3 4 5? is 4 5?
To convert to palindrome, we can either take the mirror of its left side or take mirror of its right side. However, if we take the mirror of the right
side, then the palindrome so formed is not guaranteed to be next larger palindrome. So, we must take the mirror of left side and copy it to right
side. But there are some cases that must be handled in different ways. See the following steps.
We will start with two indices i and j. i pointing to the two middle elements (or pointing to two elements around the middle element in case of n
being odd). We one by one move i and j away from each other.
Step 1. Initially, ignore the part of left side which is same as the corresponding part of right side. For example, if the number is 8 3 4 2 2 4 6 9?,
we ignore the middle four elements. i now points to element 3 and j now points to element 6.
Step 2. After step 1, following cases arise:
Case 1: Indices i & j cross the boundary.
This case occurs when the input number is palindrome. In this case, we just add 1 to the middle digit (or digits in case n is even) propagate the
carry towards MSB digit of left side and simultaneously copy mirror of the left side to the right side.
For example, if the given number is 1 2 9 2 1?, we increment 9 to 10 and propagate the carry. So the number becomes 1 3 0 3 1?
Case 2: There are digits left between left side and right side which are not same. So, we just mirror the left side to the right side & try to minimize
the number formed to guarantee the next smallest palindrome.
In this case, there can be two sub-cases.
2.1) Copying the left side to the right side is sufficient, we dont need to increment any digits and the result is just mirror of left side. Following are
some examples of this sub-case.
Next palindrome for 7 8 3 3 2 2? is 7 8 3 3 8 7?
Next palindrome for 1 2 5 3 2 2? is 1 2 5 5 2 1?
Next palindrome for 1 4 5 8 7 6 7 8 3 2 2? is 1 4 5 8 7 6 7 8 5 4 1?
How do we check for this sub-case? All we need to check is the digit just after the ignored part in step 1. This digit is highlighted in above
examples. If this digit is greater than the corresponding digit in right side digit, then copying the left side to the right side is sufficient and we dont
need to do anything else.
2.2) Copying the left side to the right side is NOT sufficient. This happens when the above defined digit of left side is smaller. Following are some
examples of this case.
Next palindrome for 7 1 3 3 2 2? is 7 1 4 4 1 7?
Next palindrome for 1 2 3 4 6 2 8? is 1 2 3 5 3 2 1?
Next palindrome for 9 4 1 8 7 9 7 8 3 2 2? is 9 4 1 8 8 0 8 8 1 4 9?
We handle this subcase like Case 1. We just add 1 to the middle digit (or digits in ase n is even) propagate the carry towards MSB digit of left
side and simultaneously copy mirror of the left side to the right side.
#include <stdio.h>
// A utility function to print an array
void printArray (int arr[], int n);
// A utility function to check if num has all 9s
int AreAll9s (int num[], int n );
// Returns next palindrome of a given number num[].
// This function is for input type 2 and 3
void generateNextPalindromeUtil (int num[], int n )
{
// find the index of mid digit
int mid = n/2;

// A bool variable to check if copy of left side to right is sufficient or not

bool leftsmaller = false;
// end of left side is always 'mid -1'
int i = mid - 1;
// Begining of right side depends if n is odd or even
int j = (n % 2)? mid + 1 : mid;
// Initially, ignore the middle same digits
while (i >= 0 && num[i] == num[j])
i--,j++;
// Find if the middle digit(s) need to be incremented or not (or copying left
// side is not sufficient)
if ( i < 0 || num[i] < num[j])
leftsmaller = true;
// Copy the mirror of left to tight
while (i >= 0)
{
num[j] = num[i];
j++;
i--;
}
// Handle the case where middle digit(s) must be incremented.
// This part of code is for CASE 1 and CASE 2.2
if (leftsmaller == true)
{
int carry = 1;
i = mid - 1;
// If there are odd digits, then increment
// the middle digit and store the carry
if (n%2 == 1)
{
num[mid] += carry;
carry = num[mid] / 10;
num[mid] %= 10;
j = mid + 1;
}
else
j = mid;
// Add 1 to the rightmost digit of the left side, propagate the carry
// towards MSB digit and simultaneously copying mirror of the left side
// to the right side.
while (i >= 0)
{
num[i] += carry;
carry = num[i] / 10;
num[i] %= 10;
num[j++] = num[i--]; // copy mirror to right
}
}
}
// The function that prints next palindrome of a given number num[]
// with n digits.
void generateNextPalindrome( int num[], int n )
{
int i;
printf("\nNext palindrome is:\n");
// Input type 1: All the digits are 9, simply o/p 1
// followed by n-1 0's followed by 1.
if( AreAll9s( num, n ) )
{
printf( "1 ");
for( i = 1; i < n; i++ )
printf( "0 " );
printf( "1" );
}
// Input type 2 and 3
else
{
generateNextPalindromeUtil ( num, n );

// print the result

printArray (num, n);
}
}
// A utility function to check if num has all 9s
int AreAll9s( int* num, int n )
{
int i;
for( i = 0; i < n; ++i )
if( num[i] != 9 )
return 0;
return 1;
}
/* Utility that prints out an array on a line */
void printArray(int arr[], int n)
{
int i;
for (i=0; i < n; i++)
printf("%d ", arr[i]);
printf("\n");
}
// Driver Program to test above function
int main()
{
int num[] = {9, 4, 1, 8, 7, 9, 7, 8, 3, 2, 2};
int n = sizeof (num)/ sizeof(num[0]);
generateNextPalindrome( num, n );
return 0;
}

Output:
Next palindrome is:
9 4 1 8 8 0 8 8 1 4 9

Make a fair coin from a biased coin

You are given a function foo() that represents a biased coin. When foo() is called, it returns 0 with 60% probability, and 1 with 40% probability.
Write a new function that returns 0 and 1 with 50% probability each. Your function should use only foo(), no other library method.
Solution:
We know foo() returns 0 with 60% probability. How can we ensure that 0 and 1 are returned with 50% probability?
The solution is similar to this post. If we can somehow get two cases with equal probability, then we are done. We call foo() two times. Both calls
will return 0 with 60% probability. So the two pairs (0, 1) and (1, 0) will be generated with equal probability from two calls of foo(). Let us see
how.
(0, 1): The probability to get 0 followed by 1 from two calls of foo() = 0.6 * 0.4 = 0.24
(1, 0): The probability to get 1 followed by 0 from two calls of foo() = 0.4 * 0.6 = 0.24
So the two cases appear with equal probability. The idea is to return consider only the above two cases, return 0 in one case, return 1 in
other case. For other cases [(0, 0) and (1, 1)], recur until you end up in any of the above two cases.
The below program depicts how we can use foo() to return 0 and 1 with equal probability.
#include <stdio.h>
int foo() // given method that returns 0 with 60% probability and 1 with 40%
{
// some code here
}
// returns both 0 and 1 with 50% probability
int my_fun()
{
int val1 = foo();
int val2 = foo();
if (val1 == 0 && val2 == 1)
return 0; // Will reach here with 0.24 probability
if (val1 == 1 && val2 == 0)
return 1; // // Will reach here with 0.24 probability
return my_fun(); // will reach here with (1 - 0.24 - 0.24) probability
}
int main()
{
printf ("%d ", my_fun());
return 0;
}

References:
http://en.wikipedia.org/wiki/Fair_coin#Fair_results_from_a_biased_coin

Check divisibility by 7
Given a number, check if it is divisible by 7. You are not allowed to use modulo operator, floating point arithmetic is also not allowed.
A simple method is repeated subtraction. Following is another interesting method.
Divisibility by 7 can be checked by a recursive method. A number of the form 10a + b is divisible by 7 if and only if a 2b is divisible by 7. In other
words, subtract twice the last digit from the number formed by the remaining digits. Continue to do this until a small number.
Example: the number 371: 37 (21) = 37 2 = 35; 3 (2 5) = 3 10 = -7; thus, since -7 is divisible by 7, 371 is divisible by 7.
Following is C implementation of the above method
// A Program to check whether a number is divisible by 7
#include <stdio.h>
int isDivisibleBy7( int num )
{
// If number is negative, make it positive
if( num < 0 )
return isDivisibleBy7( -num );
// Base cases
if( num == 0 || num == 7 )
return 1;
if( num < 10 )
return 0;
// Recur for ( num / 10 - 2 * num % 10 )
return isDivisibleBy7( num / 10 - 2 * ( num - num / 10 * 10 ) );
}
// Driver program to test above function
int main()
{
int num = 616;
if( isDivisibleBy7(num ) )
printf( "Divisible" );
else
printf( "Not Divisible" );
return 0;
}

Output:
Divisible

How does this work? Let b be the last digit of a number n and let a be the number we get when we split off b.
The representation of the number may also be multiplied by any number relatively prime to the divisor without changing its divisibility. After
observing that 7 divides 21, we can perform the following:
10.a + b

after multiplying by 2, this becomes

20.a + 2.b

and then
21.a - a + 2.b

Eliminating the multiple of 21 gives

-a + 2b

and multiplying by -1 gives

a - 2b

There are other interesting methods to check divisibility by 7 and other numbers. See following Wiki page for details.
References:
http://en.wikipedia.org/wiki/Divisibility_rule

Find the largest multiple of 3

Given an array of non-negative integers. Find the largest multiple of 3 that can be formed from array elements.
For example, if the input array is {8, 1, 9}, the output should be 9 8 1?, and if the input array is {8, 1, 7, 6, 0}, output should be 8 7 6 0?.
Method 1 (Brute Force)
The simple & straight forward approach is to generate all the combinations of the elements and keep track of the largest number formed which is
divisible by 3.
Time Complexity: O(n x 2^n). There will be 2^n combinations of array elements. To compare each combination with the largest number so far
may take O(n) time.
Auxiliary Space: O(n) // to avoid integer overflow, the largest number is assumed to be stored in the form of array.
Method 2 (Tricky)
This problem can be solved efficiently with the help of O(n) extra space. This method is based on the following facts about numbers which are
multiple of 3.
1) A number is multiple of 3 if and only if the sum of digits of number is multiple of 3. For example, let us consider 8760, it is a multiple of 3
because sum of digits is 8 + 7+ 6+ 0 = 21, which is a multiple of 3.
2) If a number is multiple of 3, then all permutations of it are also multiple of 3. For example, since 6078 is a multiple of 3, the numbers 8760,
7608, 7068, .. are also multiples of 3.
3) We get the same remainder when we divide the number and sum of digits of the number. For example, if divide number 151 and sum of it digits
7, by 3, we get the same remainder 1.
What is the idea behind above facts?
The value of 10%3 and 100%3 is 1. The same is true for all the higher powers of 10, because 3 divides 9, 99, 999, etc.
Let us consider a 3 digit number n to prove above facts. Let the first, second and third digits of n be a, b and c respectively. n can be written as
n = 100.a + 10.b + c

Since (10^x)%3 is 1 for any x, the above expression gives the same remainder as following expression
1.a + 1.b + c

So the remainder obtained by sum of digits and n is same.

Following is a solution based on the above observation.
1. Sort the array in non-decreasing order.
2. Take three queues. One for storing elements which on dividing by 3 gives remainder as 0.The second queue stores digits which on dividing by 3
gives remainder as 1. The third queue stores digits which on dividing by 3 gives remainder as 2. Call them as queue0, queue1 and queue2
3. Find the sum of all the digits.
4. Three cases arise:
4.1 The sum of digits is divisible by 3. Dequeue all the digits from the three queues. Sort them in non-increasing order. Output the array.
4.2 The sum of digits produces remainder 1 when divided by 3.
Remove one item from queue1. If queue1 is empty, remove two items from queue2. If queue2 contains less than two items, the number is not
possible.
4.3 The sum of digits produces remainder 2 when divided by 3.
Remove one item from queue2. If queue2 is empty, remove two items from queue1. If queue1 contains less than two items, the number is not
possible.
5. Finally empty all the queues into an auxiliary array. Sort the auxiliary array in non-increasing order. Output the auxiliary array.
Based on the above, below is C implementation:
The below code works only if the input arrays has numbers from 0 to 9. It can be easily extended for any positive integer array. We
just have to modify the part where we sort the array in decreasing order, at the end of code.
/* A program to find the largest multiple of 3 from an array of elements */
#include <stdio.h>
#include <stdlib.h>
// A queue node

typedef struct Queue

{
int front;
int rear;
int capacity;
int* array;
} Queue;
// A utility function to create a queue with given capacity
Queue* createQueue( int capacity )
{
Queue* queue = (Queue *) malloc (sizeof(Queue));
queue->capacity = capacity;
queue->front = queue->rear = -1;
queue->array = (int *) malloc (queue->capacity * sizeof(int));
return queue;
}
// A utility function to check if queue is empty
int isEmpty (Queue* queue)
{
return queue->front == -1;
}
// A function to add an item to queue
void Enqueue (Queue* queue, int item)
{
queue->array[ ++queue->rear ] = item;
if ( isEmpty(queue) )
++queue->front;
}
// A function to remove an item from queue
int Dequeue (Queue* queue)
{
int item = queue->array[ queue->front ];
if( queue->front == queue->rear )
queue->front = queue->rear = -1;
else
queue->front++;
return item;
}
// A utility function to print array contents
void printArr (int* arr, int size)
{
int i;
for (i = 0; i< size; ++i)
printf ("%d ", arr[i]);
}
/* Following two functions are needed for library function qsort().
Refer following link for help of qsort()
http://www.cplusplus.com/reference/clibrary/cstdlib/qsort/ */
int compareAsc( const void* a, const void* b )
{
return *(int*)a > *(int*)b;
}
int compareDesc( const void* a, const void* b )
{
return *(int*)a < *(int*)b;
}
// This function puts all elements of 3 queues in the auxiliary array
void populateAux (int* aux, Queue* queue0, Queue* queue1,
Queue* queue2, int* top )
{
// Put all items of first queue in aux[]
while ( !isEmpty(queue0) )
aux[ (*top)++ ] = Dequeue( queue0 );
// Put all items of second queue in aux[]
while ( !isEmpty(queue1) )
aux[ (*top)++ ] = Dequeue( queue1 );
// Put all items of third queue in aux[]
while ( !isEmpty(queue2) )
aux[ (*top)++ ] = Dequeue( queue2 );
}

// The main function that finds the largest possible multiple of

// 3 that can be formed by arr[] elements
int findMaxMultupleOf3( int* arr, int size )
{
// Step 1: sort the array in non-decreasing order
qsort( arr, size, sizeof( int ), compareAsc );
// Create 3 queues to store numbers with remainder 0, 1
// and 2 respectively
Queue* queue0 = createQueue( size );
Queue* queue1 = createQueue( size );
Queue* queue2 = createQueue( size );
// Step 2 and 3 get the sum of numbers and place them in
// corresponding queues
int i, sum;
for ( i = 0, sum = 0; i < size; ++i )
{
sum += arr[i];
if ( (arr[i] % 3) == 0 )
Enqueue( queue0, arr[i] );
else if ( (arr[i] % 3) == 1 )
Enqueue( queue1, arr[i] );
else
Enqueue( queue2, arr[i] );
}
// Step 4.2: The sum produces remainder 1
if ( (sum % 3) == 1 )
{
// either remove one item from queue1
if ( !isEmpty( queue1 ) )
Dequeue( queue1 );
// or remove two items from queue2
else
{
if ( !isEmpty( queue2 ) )
Dequeue( queue2 );
else
return 0;
if ( !isEmpty( queue2 ) )
Dequeue( queue2 );
else
return 0;
}
}
// Step 4.3: The sum produces remainder 2
else if ((sum % 3) == 2)
{
// either remove one item from queue2
if ( !isEmpty( queue2 ) )
Dequeue( queue2 );
// or remove two items from queue1
else
{
if ( !isEmpty( queue1 ) )
Dequeue( queue1 );
else
return 0;
if ( !isEmpty( queue1 ) )
Dequeue( queue1 );
else
return 0;
}
}
int aux[size], top = 0;
// Empty all the queues into an auxiliary array.
populateAux (aux, queue0, queue1, queue2, &top);
// sort the array in non-increasing order
qsort (aux, top, sizeof( int ), compareDesc);
// print the result
printArr (aux, top);

return top;
}
// Driver program to test above functions
int main()
{
int arr[] = {8, 1, 7, 6, 0};
int size = sizeof(arr)/sizeof(arr[0]);
if (findMaxMultupleOf3( arr, size ) == 0)
printf( "Not Possible" );
return 0;
}

The above method can be optimized in following ways.

1) We can use Heap Sort or Merge Sort to make the time complexity O(nLogn).
2) We can avoid extra space for queues. We know at most two items will be removed from the input array. So we can keep track of two items in
two variables.
3) At the end, instead of sorting the array again in descending order, we can print the ascending sorted array in reverse order. While printing in
reverse order, we can skip the two elements to be removed.
Time Complexity: O(nLogn), assuming a O(nLogn) algorithm is used for sorting.

Lexicographic rank of a string

Given a string, find its rank among all its permutations sorted lexicographically. For example, rank of abc is 1, rank of acb is 2, and rank of cba is
6.
For simplicity, let us assume that the string does not contain any duplicated characters.
One simple solution is to initialize rank as 1, generate all permutations in lexicographic order. After generating a permutation, check if the generated
permutation is same as given string, if same, then return rank, if not, then increment the rank by 1. The time complexity of this solution will be
exponential in worst case. Following is an efficient solution.
Let the given string be STRING. In the input string, S is the first character. There are total 6 characters and 4 of them are smaller than S. So there
can be 4 * 5! smaller strings where first character is smaller than S, like following
RXXXXX
IXXXXX
N XXXXX
GXXXXX
Now let us Fix S and find the smaller strings staring with S.
Repeat the same process for T, rank is 4*5! + 4*4! +
Now fix T and repeat the same process for R, rank is 4*5! + 4*4! + 3*3! +
Now fix R and repeat the same process for I, rank is 4*5! + 4*4! + 3*3! + 1*2! +
Now fix I and repeat the same process for N, rank is 4*5! + 4*4! + 3*3! + 1*2! + 1*1! +
Now fix N and repeat the same process for G, rank is 4*5! + 4*4 + 3*3! + 1*2! + 1*1! + 0*0!
Rank = 4*5! + 4*4! + 3*3! + 1*2! + 1*1! + 0*0! = 597
Since the value of rank starts from 1, the final rank = 1 + 597 = 598
#include <stdio.h>
#include <string.h>
// A utility function to find factorial of n
int fact(int n)
{
return (n <= 1)? 1 :n * fact(n-1);
}
// A utility function to count smaller characters on right
// of arr[low]
int findSmallerInRight(char* str, int low, int high)
{
int countRight = 0, i;
for (i = low+1; i <= high; ++i)
if (str[i] < str[low])
++countRight;
return countRight;
}
// A function to find rank of a string in all permutations
// of characters
int findRank (char* str)
{
int len = strlen(str);
int mul = fact(len);
int rank = 1;
int countRight;
int i;
for (i = 0; i < len; ++i)
{
mul /= len - i;
// count number of chars smaller than str[i]
// fron str[i+1] to str[len-1]
countRight = findSmallerInRight(str, i, len-1);

rank += countRight * mul ;

}
return rank;
}
// Driver program to test above function
int main()
{
char str[] = "string";
printf ("%d", findRank(str));
return 0;
}

Output
598

The time complexity of the above solution is O(n^2). We can reduce the time complexity to O(n) by creating an auxiliary array of size 256. See
following code.
// A O(n) solution for finding rank of string
#include <stdio.h>
#include <string.h>
#define MAX_CHAR 256
// A utility function to find factorial of n
int fact(int n)
{
return (n <= 1)? 1 :n * fact(n-1);
}
// Construct a count array where value at every index
// contains count of smaller characters in whole string
void populateAndIncreaseCount (int* count, char* str)
{
int i;
for( i = 0; str[i]; ++i )
++count[ str[i] ];
for( i = 1; i < 256; ++i )
count[i] += count[i-1];
}
// Removes a character ch from count[] array
// constructed by populateAndIncreaseCount()
void updatecount (int* count, char ch)
{
int i;
for( i = ch; i < MAX_CHAR; ++i )
--count[i];
}
// A function to find rank of a string in all permutations
// of characters
int findRank (char* str)
{
int len = strlen(str);
int mul = fact(len);
int rank = 1, i;
int count[MAX_CHAR] = {0}; // all elements of count[] are initialized with 0
// Populate the count array such that count[i] contains count of
// characters which are present in str and are smaller than i
populateAndIncreaseCount( count, str );
for (i = 0; i < len; ++i)
{
mul /= len - i;
// count number of chars smaller than str[i]
// fron str[i+1] to str[len-1]
rank += count[ str[i] - 1] * mul;
// Reduce count of characters greater than str[i]
updatecount (count, str[i]);
}
return rank;
}

// Driver program to test above function

int main()
{
char str[] = "string";
printf ("%d", findRank(str));
return 0;
}

The above programs dont work for duplicate characters. To make them work for duplicate characters, find all the characters that are smaller
(include equal this time also), do the same as above but, this time divide the rank so formed by p! where p is the count of occurrences of the
repeating character.

Print all permutations in sorted (lexicographic) order

Given a string, print all permutations of it in sorted order. For example, if the input string is ABC, then output should be ABC, ACB, BAC, BCA,
CAB, CBA.
We have discussed a program to print all permutations in this post, but here we must print the permutations in increasing order.
Following are the steps to print the permutations lexicographic-ally
1. Sort the given string in non-decreasing order and print it. The first permutation is always the string sorted in non-decreasing order.
2. Start generating next higher permutation. Do it until next higher permutation is not possible. If we reach a permutation where all characters are
sorted in non-increasing order, then that permutation is the last permutation.
Steps to generate the next higher permutation:
1. Take the previously printed permutation and find the rightmost character in it, which is smaller than its next character. Let us call this character as
first character.
2. Now find the ceiling of the first character. Ceiling is the smallest character on right of first character, which is greater than first character. Let us
call the ceil character as second character.
3. Swap the two characters found in above 2 steps.
4. Sort the substring (in non-decreasing order) after the original index of first character.
Let us consider the string ABCDEF. Let previously printed permutation be DCFEBA. The next permutation in sorted order should be DEABCF.
Let us understand above steps to find next permutation. The first character will be C. The second character will be E. After swapping these two,
we get DEFCBA. The final step is to sort the substring after the character original index of first character. Finally, we get DEABCF.
Following is C++ implementation of the algorithm.
// Program to print all permutations of a string in sorted order.
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
/* Following function is needed for library function qsort(). Refer
http://www.cplusplus.com/reference/clibrary/cstdlib/qsort/ */
int compare (const void *a, const void * b)
{ return ( *(char *)a - *(char *)b ); }
// A utility function two swap two characters a and b
void swap (char* a, char* b)
{
char t = *a;
*a = *b;
*b = t;
}
// This function finds the index of the smallest character
// which is greater than 'first' and is present in str[l..h]
int findCeil (char str[], char first, int l, int h)
{
// initialize index of ceiling element
int ceilIndex = l;
// Now iterate through rest of the elements and find
// the smallest character greater than 'first'
for (int i = l+1; i <= h; i++)
if (str[i] > first && str[i] < str[ceilIndex])
ceilIndex = i;
return ceilIndex;
}
// Print all permutations of str in sorted order
void sortedPermutations ( char str[] )
{
// Get size of string
int size = strlen(str);
// Sort the string in increasing order
qsort( str, size, sizeof( str[0] ), compare );
// Print permutations one by one
bool isFinished = false;

while ( ! isFinished )
{
// print this permutation
printf ("%s \n", str);
// Find the rightmost character which is smaller than its next
// character. Let us call it 'first char'
int i;
for ( i = size - 2; i >= 0; --i )
if (str[i] < str[i+1])
break;
// If there is no such chracter, all are sorted in decreasing order,
// means we just printed the last permutation and we are done.
if ( i == -1 )
isFinished = true;
else
{
// Find the ceil of 'first char' in right of first character.
// Ceil of a character is the smallest character greater than it
int ceilIndex = findCeil( str, str[i], i + 1, size - 1 );
// Swap first and second characters
swap( &str[i], &str[ceilIndex] );
// Sort the string on right of 'first char'
qsort( str + i + 1, size - i - 1, sizeof(str[0]), compare );
}
}
}
// Driver program to test above function
int main()
{
char str[] = "ABCD";
sortedPermutations( str );
return 0;
}

Output:
ABCD
ABDC
....
....
DCAB
DCBA

The upper bound on time complexity of the above program is O(n^2 x n!). We can optimize step 4 of the above algorithm for finding next
permutation. Instead of sorting the subarray after the first character, we can reverse the subarray, because the subarray we get after swapping is
always sorted in non-increasing order. This optimization makes the time complexity as O(n x n!). See following optimized code.
// An optimized version that uses reverse instead of sort for
// finding the next permutation
// A utility function to reverse a string str[l..h]
void reverse(char str[], int l, int h)
{
while (l < h)
{
swap(&str[l], &str[h]);
l++;
h--;
}
}
// Print all permutations of str in sorted order
void sortedPermutations ( char str[] )
{
// Get size of string
int size = strlen(str);
// Sort the string in increasing order
qsort( str, size, sizeof( str[0] ), compare );
// Print permutations one by one
bool isFinished = false;
while ( ! isFinished )
{
// print this permutation

printf ("%s \n", str);

// Find the rightmost character which is smaller than its next
// character. Let us call it 'first char'
int i;
for ( i = size - 2; i >= 0; --i )
if (str[i] < str[i+1])
break;
// If there is no such chracter, all are sorted in decreasing order,
// means we just printed the last permutation and we are done.
if ( i == -1 )
isFinished = true;
else
{
// Find the ceil of 'first char' in right of first character.
// Ceil of a character is the smallest character greater than it
int ceilIndex = findCeil( str, str[i], i + 1, size - 1 );
// Swap first and second characters
swap( &str[i], &str[ceilIndex] );
// reverse the string on right of 'first char'
reverse( str, i + 1, size - 1 );
}
}
}

The above programs print duplicate permutation when characters are repeated. We can avoid it by keeping track of the previous permutation.
While printing, if the current permutation is same as previous permutation, we wont print it.

Shuffle a given array

Given an array, write a program to generate a random permutation of array elements. This question is also asked as shuffle a deck of cards or
randomize a given array.

Let the given array be arr[]. A simple solution is to create an auxiliary array temp[] which is initially a copy of arr[]. Randomly select an element
from temp[], copy the randomly selected element to arr[0] and remove the selected element from temp[]. Repeat the same process n times and
keep copying elements to arr[1], arr[2], . The time complexity of this solution will be O(n^2).
FisherYates shuffle Algorithm works in O(n) time complexity. The assumption here is, we are given a function rand() that generates random
number in O(1) time.
The idea is to start from the last element, swap it with a randomly selected element from the whole array (including last). Now consider the array
from 0 to n-2 (size reduced by 1), and repeat the process till we hit the first element.
Following is the detailed algorithm
To shuffle an array a of n elements (indices 0..n-1):
for i from n - 1 downto 1 do
j = random integer with 0 <= j <= i
exchange a[j] and a[i]

Following is C++ implementation of this algorithm.

// C Program to shuffle a given array
#include <stdio.h>
#include <stdlib.h>
#include <time.h>
// A utility function to swap to integers
void swap (int *a, int *b)
{
int temp = *a;
*a = *b;
*b = temp;
}
// A utility function to print an array
void printArray (int arr[], int n)
{
for (int i = 0; i < n; i++)
printf("%d ", arr[i]);
printf("\n");
}
// A function to generate a random permutation of arr[]
void randomize ( int arr[], int n )
{
// Use a different seed value so that we don't get same
// result each time we run this program
srand ( time(NULL) );
// Start from the last element and swap one by one. We don't
// need to run for the first element that's why i > 0
for (int i = n-1; i > 0; i--)
{
// Pick a random index from 0 to i
int j = rand() % (i+1);
// Swap arr[i] with the element at random index
swap(&arr[i], &arr[j]);
}
}
// Driver program to test above function.
int main()
{
int arr[] = {1, 2, 3, 4, 5, 6, 7, 8};
int n = sizeof(arr)/ sizeof(arr[0]);

randomize (arr, n);

printArray(arr, n);
return 0;
}

Output:
7 8 4 6 3 1 2 5

The above function assumes that rand() generates a random number.

Time Complexity: O(n), assuming that the function rand() takes O(1) time.
How does this work?
The probability that ith element (including the last one) goes to last position is 1/n, because we randomly pick an element in first iteration.
The probability that ith element goes to second last position can be proved to be 1/n by dividing it in two cases.
Case 1: i = n-1 (index of last element):
The probability of last element going to second last position is = (probability that last element doesn't stay at its original position) x (probability that
the index picked in previous step is picked again so that the last element is swapped)
So the probability = ((n-1)/n) x (1/(n-1)) = 1/n
Case 2: 0 < i < n-1 (index of non-last):
The probability of ith element going to second position = (probability that ith element is not picked in previous iteration) x (probability that ith
element is picked in this iteration)
So the probability = ((n-1)/n) x (1/(n-1)) = 1/n
We can easily generalize above proof for any other position.

Space and time efficient Binomial Coefficient

Write a function that takes two parameters n and k and returns the value of Binomial Coefficient C(n, k). For example, your function should return
6 for n = 4 and k = 2, and it should return 10 for n = 5 and k = 2.
We have discussed a O(n*k) time and O(k) extra space algorithm in this post. The value of C(n, k) can be calculated in O(k) time and O(1) extra
space.
C(n, k) = n! / (n-k)! * k!
= [n * (n-1) *....* 1] / [ ( (n-k) * (n-k-1) * .... * 1) *
( k * (k-1) * .... * 1 ) ]
After simplifying, we get
C(n, k) = [n * (n-1) * .... * (n-k+1)] / [k * (k-1) * .... * 1]
Also, C(n, k) = C(n, n-k) // we can change r to n-r if r > n-r

Following implementation uses above formula to calculate C(n, k)

// Program to calculate C(n ,k)
#include <stdio.h>
// Returns value of Binomial Coefficient C(n, k)
int binomialCoeff(int n, int k)
{
int res = 1;
// Since C(n, k) = C(n, n-k)
if ( k > n - k )
k = n - k;
// Calculate value of [n * (n-1) *---* (n-k+1)] / [k * (k-1) *----* 1]
for (int i = 0; i < k; ++i)
{
res *= (n - i);
res /= (i + 1);
}
return res;
}
/* Drier program to test above function*/
int main()
{
int n = 8, k = 2;
printf ("Value of C(%d, %d) is %d ", n, k, binomialCoeff(n, k) );
return 0;
}
Value of C(8, 2) is 28

Time Complexity: O(k)

Auxiliary Space: O(1)

Reservoir Sampling
Reservoir sampling is a family of randomized algorithms for randomly choosing k samples from a list of n items, where n is either a very large or
unknown number. Typically n is large enough that the list doesnt fit into main memory. For example, a list of search queries in Google and
Facebook.
So we are given a big array (or stream) of numbers (to simplify), and we need to write an efficient function to randomly select k numbers where 1
<= k <= n. Let the input array be stream[].
A simple solution is to create an array reservoir[] of maximum size k. One by one randomly select an item from stream[0..n-1]. If the selected
item is not previously selected, then put it in reservoir[]. To check if an item is previously selected or not, we need to search the item in
reservoir[]. The time complexity of this algorithm will be O(k^2). This can be costly if k is big. Also, this is not efficient if the input is in the form of
a stream.
It can be solved in O(n) time. The solution also suits well for input in the form of stream. The idea is similar to this post. Following are the steps.
1) Create an array reservoir[0..k-1] and copy first k items of stream[] to it.
2) Now one by one consider all items from (k+1)th item to nth item.
a) Generate a random number from 0 to i where i is index of current item in stream[]. Let the generated random number is j.
b) If j is in range 0 to k-1, replace reservoir[j] with arr[i]
Following is C implementation of the above algorithm.
// An efficient program to randomly select k items from a stream of items
#include <stdio.h>
#include <stdlib.h>
#include <time.h>
// A utility function to print an array
void printArray(int stream[], int n)
{
for (int i = 0; i < n; i++)
printf("%d ", stream[i]);
printf("\n");
}
// A function to randomly select k items from stream[0..n-1].
void selectKItems(int stream[], int n, int k)
{
int i; // index for elements in stream[]
// reservoir[] is the output array. Initialize it with
// first k elements from stream[]
int reservoir[k];
for (i = 0; i < k; i++)
reservoir[i] = stream[i];
// Use a different seed value so that we don't get
// same result each time we run this program
srand(time(NULL));
// Iterate from the (k+1)th element to nth element
for (; i < n; i++)
{
// Pick a random index from 0 to i.
int j = rand() % (i+1);
// If the randomly picked index is smaller than k, then replace
// the element present at the index with new element from stream
if (j < k)
reservoir[j] = stream[i];
}
printf("Following are k randomly selected items \n");
printArray(reservoir, k);
}
// Driver program to test above function.
int main()
{
int stream[] = {1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12};
int n = sizeof(stream)/sizeof(stream[0]);
int k = 5;
selectKItems(stream, n, k);
return 0;
}

Output:
Following are k randomly selected items
6 2 11 8 12

Time Complexity: O(n)

How does this work?
To prove that this solution works perfectly, we must prove that the probability that any item stream[i] where 0 <= i < n will be in final
reservoir[] is k/n. Let us divide the proof in two cases as first k items are treated differently.
Case 1: For last n-k stream items, i.e., for stream[i] where k <= i < n
For every such stream item stream[i], we pick a random index from 0 to i and if the picked index is one of the first k indexes, we replace the
element at picked index with stream[i]
To simplify the proof, let us first consider the last item. The probability that the last item is in final reservoir = The probability that one of the first k
indexes is picked for last item = k/n (the probability of picking one of the k items from a list of size n)
Let us now consider the second last item. The probability that the second last item is in final reservoir[] = [Probability that one of the first k
indexes is picked in iteration for stream[n-2]] X [Probability that the index picked in iteration for stream[n-1] is not same as index picked for
stream[n-2] ] = [k/(n-1)]*[(n-1)/n] = k/n.
Similarly, we can consider other items for all stream items from stream[n-1] to stream[k] and generalize the proof.
Case 2: For first k stream items, i.e., for stream[i] where 0 <= i < k
The first k items are initially copied to reservoir[] and may be removed later in iterations for stream[k] to stream[n].
The probability that an item from stream[0..k-1] is in final array = Probability that the item is not picked when items stream[k], stream[k+1], .
stream[n-1] are considered = [k/(k+1)] x [(k+1)/(k+2)] x [(k+2)/(k+3)] x x [(n-1)/n] = k/n
References:
http://en.wikipedia.org/wiki/Reservoir_sampling

Pascals Triangle
Pascals triangle is a triangular array of the binomial coefficients. Write a function that takes an integer value n as input and prints first n lines of the
Pascals triangle. Following are the first 6 rows of Pascals Triangle.
1
1
1
1
1
1

1
2
3
4
5

1
3 1
6 4 1
10 10 5 1

Method 1 ( O(n^3) time complexity )

Number of entries in every line is equal to line number. For example, the first line has 1, the second line has 1 1?, the third line has 1 2 1?,.. and so
on. Every entry in a line is value of a Binomial Coefficient. The value of ith entry in line number line is C(line, i). The value can be calculated using
following formula.
C(line, i)

= line! / ( (line-i)! * i! )

A simple method is to run two loops and calculate the value of Binomial Coefficient in inner loop.
// A simple O(n^3) program for Pascal's Triangle
#include <stdio.h>
// See http://www.geeksforgeeks.org/archives/25621 for details of this function
int binomialCoeff(int n, int k);
// Function to print first n lines of Pascal's Triangle
void printPascal(int n)
{
// Iterate through every line and print entries in it
for (int line = 0; line < n; line++)
{
// Every line has number of integers equal to line number
for (int i = 0; i <= line; i++)
printf("%d ", binomialCoeff(line, i));
printf("\n");
}
}
// See http://www.geeksforgeeks.org/archives/25621 for details of this function
int binomialCoeff(int n, int k)
{
int res = 1;
if (k > n - k)
k = n - k;
for (int i = 0; i < k; ++i)
{
res *= (n - i);
res /= (i + 1);
}
return res;
}
// Driver program to test above function
int main()
{
int n = 7;
printPascal(n);
return 0;
}

Time complexity of this method is O(n^3). Following are optimized methods.

Method 2( O(n^2) time and O(n^2) extra space )

If we take a closer at the triangle, we observe that every entry is sum of the two values above it. So we can create a 2D array that stores
previously generated values. To generate a value in a line, we can use the previously stored values from array.

// A O(n^2) time and O(n^2) extra space method for Pascal's Triangle
void printPascal(int n)
{
int arr[n][n]; // An auxiliary array to store generated pscal triangle values
// Iterate through every line and print integer(s) in it
for (int line = 0; line < n; line++)
{
// Every line has number of integers equal to line number
for (int i = 0; i <= line; i++)
{
// First and last values in every row are 1
if (line == i || i == 0)
arr[line][i] = 1;
else // Other values are sum of values just above and left of above
arr[line][i] = arr[line-1][i-1] + arr[line-1][i];
printf("%d ", arr[line][i]);
}
printf("\n");
}
}

This method can be optimized to use O(n) extra space as we need values only from previous row. So we can create an auxiliary array of size n
and overwrite values. Following is another method uses only O(1) extra space.

Method 3 ( O(n^2) time and O(1) extra space )

This method is based on method 1. We know that ith entry in a line number line is Binomial Coefficient C(line, i) and all lines start with value 1.
The idea is to calculate C(line, i) using C(line, i-1). It can be calculated in O(1) time using the following.
C(line, i) = line! / ( (line-i)! * i! )
C(line, i-1) = line! / ( (line - i + 1)! * (i-1)! )
We can derive following expression from above two expressions.
C(line, i) = C(line, i-1) * (line - i + 1) / i
So C(line, i) can be calculated from C(line, i-1) in O(1) time
// A O(n^2) time and O(1) extra space function for Pascal's Triangle
void printPascal(int n)
{
for (int line = 1; line <= n; line++)
{
int C = 1; // used to represent C(line, i)
for (int i = 1; i <= line; i++)
{
printf("%d ", C); // The first value in a line is always 1
C = C * (line - i) / i;
}
printf("\n");
}
}

So method 3 is the best method among all, but it may cause integer overflow for large values of n as it multiplies two integers to obtain values.

Select a random number from stream, with O(1) space

Given a stream of numbers, generate a random number from the stream. You are allowed to use only O(1) space and the input is in the form of
stream, so cant store the previously seen numbers.
So how do we generate a random number from the whole stream such that the probability of picking any number is 1/n. with O(1) extra space?
This problem is a variation of Reservoir Sampling. Here the value of k is 1.
1) Initialize count as 0, count is used to store count of numbers seen so far in stream.
2) For each number x from stream, do following
..a) Increment count by 1.
..b) If count is 1, set result as x, and return result.
..c) Generate a random number from 0 to count-1?. Let the generated random number be i.
..d) If i is equal to count 1?, update the result as x.
// An efficient program to randomly select a number from stream of numbers.
#include <stdio.h>
#include <stdlib.h>
#include <time.h>
// A function to randomly select a item from stream[0], stream[1], .. stream[i-1]
int selectRandom(int x)
{
static int res;
// The resultant random number
static int count = 0; //Count of numbers visited so far in stream
count++; // increment count of numbers seen so far
// If this is the first element from stream, return it
if (count == 1)
res = x;
else
{
// Generate a random number from 0 to count - 1
int i = rand() % count;
// Replace the prev random number with new number with 1/count probability
if (i == count - 1)
res = x;
}
return res;
}
// Driver program to test above function.
int main()
{
int stream[] = {1, 2, 3, 4};
int n = sizeof(stream)/sizeof(stream[0]);
// Use a different seed value for every run.
srand(time(NULL));
for (int i = 0; i < n; ++i)
printf("Random number from first %d numbers is %d \n",
i+1, selectRandom(stream[i]));
return 0;
}

Output:
Random
Random
Random
Random

number
number
number
number

from
from
from
from

first
first
first
first

1
2
3
4

numbers
numbers
numbers
numbers

is
is
is
is

1
1
3
4

Auxiliary Space: O(1)

How does this work
We need to prove that every element is picked with 1/n probability where n is the number of items seen so far. For every new stream item x, we
pick a random number from 0 to count -1?, if the picked number is count-1?, we replace the previous result with x.
To simplify proof, let us first consider the last element, the last element replaces the previously stored result with 1/n probability. So probability of
getting last element as result is 1/n.
Let us now talk about second last element. When second last element processed first time, the probability that it replaced the previous result is
1/(n-1). The probability that previous result stays when nth item is considered is (n-1)/n. So probability that the second last element is picked in last

iteration is [1/(n-1)] * [(n-1)/n] which is 1/n.

Similarly, we can prove for third last element and others.
References:
Reservoir Sampling

Find the largest multiple of 2, 3 and 5

An array of size n is given. The array contains digits from 0 to 9. Generate the largest number using the digits in the array such that the number is
divisible by 2, 3 and 5.
For example, if the arrays is {1, 8, 7, 6, 0}, output must be: 8760. And if the arrays is {7, 7, 7, 6}, output must be: no number can be formed.
Source: Amazon Interview | Set 7
This problem is a variation of Find the largest multiple of 3.
Since the number has to be divisible by 2 and 5, it has to have last digit as 0. So if the given array doesnt contain any zero, then no solution exists.
Once a 0 is available, extract 0 from the given array. Only thing left is, the number should be is divisible by 3 and the largest of all. Which has been
discussed here.

Efficient program to calculate e^x

The value of Exponential Function e^x can be expressed using following Taylor Series.
e^x = 1 + x/1! + x^2/2! + x^3/3! + ......

How to efficiently calculate the sum of above series?

The series can be re-written as
e^x = 1 + (x/1) (1 + (x/2) (1 + (x/3) (........) ) )

Let the sum needs to be calculated for n terms, we can calculate sum using following loop.
for (i = n - 1, sum = 1; i > 0; --i )
sum = 1 + x * sum / i;

Following is implementation of the above idea.

// Efficient program to calculate e raise to the power x
#include <stdio.h>
//Returns approximate value of e^x using sum of first n terms of Taylor Series
float exponential(int n, float x)
{
float sum = 1.0f; // initialize sum of series
for (int i = n - 1; i > 0; --i )
sum = 1 + x * sum / i;
return sum;
}
// Driver program to test above function
int main()
{
int n = 10;
float x = 1.0f;
printf("e^x = %f", exponential(n, x));
return 0;
}

Output:
e^x = 2.718282

Measure one litre using two vessels and infinite water supply
There are two vessels of capacities a and b respectively. We have infinite water supply. Give an efficient algorithm to make exactly 1 litre of water
in one of the vessels. You can throw all the water from any vessel any point of time. Assume that a and b are Coprimes.
Following are the steps:
Let V1 be the vessel of capacity a and V2 be the vessel of capacity b and a is smaller than b.
1) Do following while the amount of water in V1 is not 1.
.a) If V1 is empty, then completely fill V1
.b) Transfer water from V1 to V2. If V2 becomes full, then keep the remaining water in V1 and empty V2
2) V1 will have 1 litre after termination of loop in step 1. Return.
Following is C++ implementation of the above algorithm.
/*
1.
2.
2.
3.
4.
5.
6.
7.

Sample run of the Algo for V1 with capacity

Fill V1:
V1 =
Transfer from V1 to V2, and fill V1: V1 =
Transfer from V1 to V2, and fill V1: V1 =
Transfer from V1 to V2, and empty V2: V1 =
Transfer from V1 to V2, and fill V1: V1 =
Transfer from V1 to V2, and fill V1: V1 =
Transfer from V1 to V2, and empty V2: V1 =
Stop as V1 now contains 1 litre.

3 and
3, V2
3, V2
3, V2
2, V2
3, V2
3, V2
1, V2

V2 with capacity 7
= 0
= 3
= 6
= 0
= 2
= 5
= 0

Note that V2 was made empty in steps 3 and 6 because it became full */
#include <iostream>
using namespace std;
// A utility function to get GCD of two numbers
int gcd(int a, int b) { return b? gcd(b, a % b) : a; }
// Class to represent a Vessel
class Vessel
{
// A vessel has capacity, and current amount of water in it
int capacity, current;
public:
// Constructor: initializes capacity as given, and current as 0
Vessel(int capacity) { this->capacity = capacity; current = 0; }
// The main function to fill one litre in this vessel. Capacity of V2
// must be greater than this vessel and two capacities must be co-prime
void makeOneLitre(Vessel &V2);
// Fills vessel with given amount and returns the amount of water
// transferred to it. If the vessel becomes full, then the vessel
// is made empty.
int transfer(int amount);
};
// The main function to fill one litre in this vessel. Capacity
// of V2 must be greater than this vessel and two capacities
// must be coprime
void Vessel:: makeOneLitre(Vessel &V2)
{
// solution exists iff a and b are co-prime
if (gcd(capacity, V2.capacity) != 1)
return;
while (current != 1)
{
// fill A (smaller vessel)
if (current == 0)
current = capacity;
cout << "Vessel 1: " << current << "
<< V2.current << endl;

Vessel 2: "

// Transfer water from V1 to V2 and reduce current of V1 by

// the amount equal to transferred water
current = current - V2.transfer(current);
}
// Finally, there will be 1 litre in vessel 1
cout << "Vessel 1: " << current << " Vessel 2: "
<< V2.current << endl;
}

// Fills vessel with given amount and returns the amount of water
// transferred to it. If the vessel becomes full, then the vessel
// is made empty
int Vessel::transfer(int amount)
{
// If the vessel can accommodate the given amount
if (current + amount < capacity)
{
current += amount;
return amount;
}
// If the vessel cannot accommodate the given amount, then
// store the amount of water transferred
int transferred = capacity - current;
// Since the vessel becomes full, make the vessel
// empty so that it can be filled again
current = 0;
return transferred;
}
// Driver program to test above function
int main()
{
int a = 3, b = 7; // a must be smaller than b
// Create two vessels of capacities a and b
Vessel V1(a), V2(b);
// Get 1 litre in first vessel
V1.makeOneLitre(V2);
return 0;
}

Output:
Vessel
Vessel
Vessel
Vessel
Vessel
Vessel
Vessel

1:
1:
1:
1:
1:
1:
1:

3
3
3
2
3
3
1

Vessel
Vessel
Vessel
Vessel
Vessel
Vessel
Vessel

2:
2:
2:
2:
2:
2:
2:

0
3
6
0
2
5
0

How does this work?

To prove that the algorithm works, we need to proof that after certain number of iterations in the while loop, we will get 1 litre in V1.
Let a be the capacity of vessel V1 and b be the capacity of V2. Since we repeatedly transfer water from V1 to V2 until V2 becomes full, we will
have a b (mod a) water in V1 when V2 becomes full first time . Once V2 becomes full, it is emptied. We will have a 2b (mod a) water in V1 when
V2 is full second time. We repeat the above steps, and get a nb (mod a) water in V1 after the vessel V2 is filled and emptied n times. We need to
prove that the value of a nb (mod a) will be 1 for a finite integer n. To prove this, let us consider the following property of coprime numbers.
For any two coprime integers a and b, the integer b has a multiplicative inverse modulo a. In other words, there exists an integer y such that b*y ?
1(mod)a (See 3rd point here). After (a 1)*y iterations, we will have a [(a-1)*y*b (mod a)] water in V1, the value of this expression is a [(a 1) * 1]
mod a which is 1. So the algorithm converges and we get 1 litre in V1.

Efficient program to print all prime factors of a given number

Given a number n, write an efficient function to print all prime factors of n. For example, if the input number is 12, then output should be 2 2 3?.
And if the input number is 315, then output should be 3 3 5 7?.
Following are the steps to find all prime factors.
1) While n is divisible by 2, print 2 and divide n by 2.
2) After step 1, n must be odd. Now start a loop from i = 3 to square root of n. While i divides n, print i and divide n by i, increment i by 2 and
continue.
3) If n is a prime number and is greater than 2, then n will not become 1 by above two steps. So print n if it is greater than 2.
// Program to print all prime factors
# include <stdio.h>
# include <math.h>
// A function to print all prime factors of a given number n
void primeFactors(int n)
{
// Print the number of 2s that divide n
while (n%2 == 0)
{
printf("%d ", 2);
n = n/2;
}
// n must be odd at this point. So we can skip one element (Note i = i +2)
for (int i = 3; i <= sqrt(n); i = i+2)
{
// While i divides n, print i and divide n
while (n%i == 0)
{
printf("%d ", i);
n = n/i;
}
}
// This condition is to handle the case whien n is a prime number
// greater than 2
if (n > 2)
printf ("%d ", n);
}
/* Driver program to test above function */
int main()
{
int n = 315;
primeFactors(n);
return 0;
}

Output:
3 3 5 7

How does this work?

The steps 1 and 2 take care of composite numbers and step 3 takes care of prime numbers. To prove that the complete algorithm works, we need
to prove that steps 1 and 2 actually take care of composite numbers. This is clear that step 1 takes care of even numbers. And after step 1, all
remaining prime factor must be odd (difference of two prime factors must be at least 2), this explains why i is incremented by 2.
Now the main part is, the loop runs till square root of n not till. To prove that this optimization works, let us consider the following property of
composite numbers.
Every composite number has at least one prime factor less than or equal to square root of itself.
This property can be proved using counter statement. Let a and b be two factors of n such that a*b = n. If both are greater than ?n, then a.b > ?n,
* ?n, which contradicts the expression a * b = n.
In step 2 of the above algorithm, we run a loop and do following in loop
a) Find the least prime factor i (must be less than ?n,)
b) Remove all occurrences i from n by repeatedly dividing n by i.
c) Repeat steps a and b for divided n and i = i + 2. The steps a and b are repeated till n becomes either 1 or a prime number.

Print all possible combinations of r elements in a given array of size n

Given an array of size n, generate and print all possible combinations of r elements in array. For example, if input array is {1, 2, 3, 4} and r is 2,
then output should be {1, 2}, {1, 3}, {1, 4}, {2, 3}, {2, 4} and {3, 4}.
Following are two methods to do this.
Method 1 (Fix Elements and Recur)
We create a temporary array data[] which stores all outputs one by one. The idea is to start from first index (index = 0) in data[], one by one fix
elements at this index and recur for remaining indexes. Let the input array be {1, 2, 3, 4, 5} and r be 3. We first fix 1 at index 0 in data[], then
recur for remaining indexes, then we fix 2 at index 0 and recur. Finally, we fix 3 and recur for remaining indexes. When number of elements in
data[] becomes equal to r (size of a combination), we print data[].
Following diagram shows recursion tree for same input.

Following is C++ implementation of above approach.

C
// Program to print all combination of size r in an array of size n
#include <stdio.h>
void combinationUtil(int arr[], int data[], int start, int end,
int index, int r);
// The main function that prints all combinations of size r
// in arr[] of size n. This function mainly uses combinationUtil()
void printCombination(int arr[], int n, int r)
{
// A temporary array to store all combination one by one
int data[r];
// Print all combination using temprary array 'data[]'
combinationUtil(arr, data, 0, n-1, 0, r);
}
/* arr[] ---> Input Array
data[] ---> Temporary array to store current combination
start & end ---> Staring and Ending indexes in arr[]
index ---> Current index in data[]
r ---> Size of a combination to be printed */
void combinationUtil(int arr[], int data[], int start, int end,
int index, int r)
{
// Current combination is ready to be printed, print it
if (index == r)
{
for (int j=0; j<r; j++)
printf("%d ", data[j]);
printf("\n");
return;
}
// replace index with all possible elements. The condition
// "end-i+1 >= r-index" makes sure that including one element
// at index will make a combination with remaining elements
// at remaining positions
for (int i=start; i<=end && end-i+1 >= r-index; i++)
{
data[index] = arr[i];
combinationUtil(arr, data, i+1, end, index+1, r);
}
}
// Driver program to test above functions
int main()

{
int arr[] = {1, 2, 3, 4, 5};
int r = 3;
int n = sizeof(arr)/sizeof(arr[0]);
printCombination(arr, n, r);
}

Java
// Java program to print all combination of size r in an array of size n
import java.io.*;
class Permutation {
/* arr[] ---> Input Array
data[] ---> Temporary array to store current combination
start & end ---> Staring and Ending indexes in arr[]
index ---> Current index in data[]
r ---> Size of a combination to be printed */
static void combinationUtil(int arr[], int data[], int start,
int end, int index, int r)
{
// Current combination is ready to be printed, print it
if (index == r)
{
for (int j=0; j<r; j++)
System.out.print(data[j]+" ");
System.out.println("");
return;
}
// replace index with all possible elements. The condition
// "end-i+1 >= r-index" makes sure that including one element
// at index will make a combination with remaining elements
// at remaining positions
for (int i=start; i<=end && end-i+1 >= r-index; i++)
{
data[index] = arr[i];
combinationUtil(arr, data, i+1, end, index+1, r);
}
}
// The main function that prints all combinations of size r
// in arr[] of size n. This function mainly uses combinationUtil()
static void printCombination(int arr[], int n, int r)
{
// A temporary array to store all combination one by one
int data[]=new int[r];
// Print all combination using temprary array 'data[]'
combinationUtil(arr, data, 0, n-1, 0, r);
}
/*Driver function to check for above function*/
public static void main (String[] args) {
int arr[] = {1, 2, 3, 4, 5};
int r = 3;
int n = arr.length;
printCombination(arr, n, r);
}
}
/* This code is contributed by Devesh Agrawal */

1
1
1
1
1
1
2
2
2
3

2
2
2
3
3
4
3
3
4
4

3
4
5
4
5
5
4
5
5
5

How to handle duplicates?

Note that the above method doesnt handle duplicates. For example, if input array is {1, 2, 1} and r is 2, then the program prints {1, 2} and {2, 1}

as two different combinations. We can avoid duplicates by adding following two additional things to above code.
1) Add code to sort the array before calling combinationUtil() in printCombination()
2) Add following lines at the end of for loop in combinationUtil()
// Since the elements are sorted, all occurrences of an element
// must be together
while (arr[i] == arr[i+1])
i++;

See this for an implementation that handles duplicates.

Method 2 (Include and Exclude every element)

Like the above method, We create a temporary array data[]. The idea here is similar to Subset Sum Problem. We one by one consider every
element of input array, and recur for two cases:
1) The element is included in current combination (We put the element in data[] and increment next available index in data[])
2) The element is excluded in current combination (We do not put the element and do not change index)
When number of elements in data[] become equal to r (size of a combination), we print it.
This method is mainly based on Pascals Identity, i.e. ncr = n-1cr + n-1cr-1
Following is C++ implementation of method 2.

C
// Program to print all combination of size r in an array of size n
#include<stdio.h>
void combinationUtil(int arr[],int n,int r,int index,int data[],int i);
// The main function that prints all combinations of size r
// in arr[] of size n. This function mainly uses combinationUtil()
void printCombination(int arr[], int n, int r)
{
// A temporary array to store all combination one by one
int data[r];
// Print all combination using temprary array 'data[]'
combinationUtil(arr, n, r, 0, data, 0);
}
/* arr[] ---> Input Array
n
---> Size of input array
r
---> Size of a combination to be printed
index ---> Current index in data[]
data[] ---> Temporary array to store current combination
i
---> index of current element in arr[]
*/
void combinationUtil(int arr[], int n, int r, int index, int data[], int i)
{
// Current cobination is ready, print it
if (index == r)
{
for (int j=0; j<r; j++)
printf("%d ",data[j]);
printf("\n");
return;
}
// When no more elements are there to put in data[]
if (i >= n)
return;
// current is included, put next at next location
data[index] = arr[i];
combinationUtil(arr, n, r, index+1, data, i+1);
// current is excluded, replace it with next (Note that
// i+1 is passed, but index is not changed)
combinationUtil(arr, n, r, index, data, i+1);
}
// Driver program to test above functions
int main()
{

int arr[] = {1, 2, 3, 4, 5};

int r = 3;
int n = sizeof(arr)/sizeof(arr[0]);
printCombination(arr, n, r);
return 0;
}

Java
// Java program to print all combination of size r in an array of size n
import java.io.*;
class Permutation {
/* arr[] ---> Input Array
data[] ---> Temporary array to store current combination
start & end ---> Staring and Ending indexes in arr[]
index ---> Current index in data[]
r ---> Size of a combination to be printed */
static void combinationUtil(int arr[], int n, int r, int index,
int data[], int i)
{
// Current combination is ready to be printed, print it
if (index == r)
{
for (int j=0; j<r; j++)
System.out.print(data[j]+" ");
System.out.println("");
return;
}
// When no more elements are there to put in data[]
if (i >= n)
return;
// current is included, put next at next location
data[index] = arr[i];
combinationUtil(arr, n, r, index+1, data, i+1);
// current is excluded, replace it with next (Note that
// i+1 is passed, but index is not changed)
combinationUtil(arr, n, r, index, data, i+1);
}
// The main function that prints all combinations of size r
// in arr[] of size n. This function mainly uses combinationUtil()
static void printCombination(int arr[], int n, int r)
{
// A temporary array to store all combination one by one
int data[]=new int[r];
// Print all combination using temprary array 'data[]'
combinationUtil(arr, n, r, 0, data, 0);
}
/*Driver function to check for above function*/
public static void main (String[] args) {
int arr[] = {1, 2, 3, 4, 5};
int r = 3;
int n = arr.length;
printCombination(arr, n, r);
}
}
/* This code is contributed by Devesh Agrawal */

1
1
1
1
1
1
2
2
2
3

2
2
2
3
3
4
3
3
4
4

3
4
5
4
5
5
4
5
5
5

How to handle duplicates in method 2?

Like method 1, we can following two things to handle duplicates.

1) Add code to sort the array before calling combinationUtil() in printCombination()

2) Add following lines between two recursive calls of combinationUtil() in combinationUtil()
// Since the elements are sorted, all occurrences of an element
// must be together
while (arr[i] == arr[i+1])
i++;

See this for an implementation that handles duplicates.

Random number generator in arbitrary probability distribution fashion

Given n numbers, each with some frequency of occurrence. Return a random number with probability proportional to its frequency of occurrence.
Example:
Let following be the given numbers.
arr[] = {10, 30, 20, 40}
Let following be the frequencies of given numbers.
freq[] = {1, 6, 2, 1}
The output should be
10 with probability
30 with probability
20 with probability
40 with probability

1/10
6/10
2/10
1/10

It is quite clear that the simple random number generator wont work here as it doesnt keep track of the frequency of occurrence.
We need to somehow transform the problem into a problem whose solution is known to us.
One simple method is to take an auxiliary array (say aux[]) and duplicate the numbers according to their frequency of occurrence. Generate a
random number(say r) between 0 to Sum-1(including both), where Sum represents summation of frequency array (freq[] in above example).
Return the random number aux[r] (Implementation of this method is left as an exercise to the readers).
The limitation of the above method discussed above is huge memory consumption when frequency of occurrence is high. If the input is 997, 8761
and 1, this method is clearly not efficient.
How can we reduce the memory consumption? Following is detailed algorithm that uses O(n) extra space where n is number of elements in input
arrays.
1. Take an auxiliary array (say prefix[]) of size n.
2. Populate it with prefix sum, such that prefix[i] represents sum of numbers from 0 to i.
3. Generate a random number(say r) between 1 to Sum(including both), where Sum represents summation of input frequency array.
4. Find index of Ceil of random number generated in step #3 in the prefix array. Let the index be indexc.
5. Return the random number arr[indexc], where arr[] contains the input n numbers.
Before we go to the implementation part, let us have quick look at the algorithm with an example:
arr[]: {10, 20, 30}
freq[]: {2, 3, 1}
Prefix[]: {2, 5, 6}
Since last entry in prefix is 6, all possible values of r are [1, 2, 3, 4, 5, 6]
1: Ceil is 2. Random number generated is 10.
2: Ceil is 2. Random number generated is 10.
3: Ceil is 5. Random number generated is 20.
4: Ceil is 5. Random number generated is 20.
5: Ceil is 5. Random number generated is 20.
6. Ceil is 6. Random number generated is 30.
In the above example
10 is generated with probability 2/6.
20 is generated with probability 3/6.
30 is generated with probability 1/6.
How does this work?
Any number input[i] is generated as many times as its frequency of occurrence because there exists count of integers in range(prefix[i 1], prefix[i]]
is input[i]. Like in the above example 3 is generated thrice, as there exists 3 integers 3, 4 and 5 whose ceil is 5.
//C program to generate random numbers according to given frequency distribution
#include <stdio.h>
#include <stdlib.h>
// Utility function to find ceiling of r in arr[l..h]
int findCeil(int arr[], int r, int l, int h)
{
int mid;
while (l < h)
{
mid = l + ((h - l) >> 1); // Same as mid = (l+h)/2
(r > arr[mid]) ? (l = mid + 1) : (h = mid);
}
return (arr[l] >= r) ? l : -1;

}
// The main function that returns a random number from arr[] according to
// distribution array defined by freq[]. n is size of arrays.
int myRand(int arr[], int freq[], int n)
{
// Create and fill prefix array
int prefix[n], i;
prefix[0] = freq[0];
for (i = 1; i < n; ++i)
prefix[i] = prefix[i - 1] + freq[i];
// prefix[n-1] is sum of all frequencies. Generate a random number
// with value from 1 to this sum
int r = (rand() % prefix[n - 1]) + 1;
// Find index of ceiling of r in prefix arrat
int indexc = findCeil(prefix, r, 0, n - 1);
return arr[indexc];
}
// Driver program to test above functions
int main()
{
int arr[] = {1, 2, 3, 4};
int freq[] = {10, 5, 20, 100};
int i, n = sizeof(arr) / sizeof(arr[0]);
// Use a different seed value for every run.
srand(time(NULL));
// Let us generate 10 random numbers accroding to
// given distribution
for (i = 0; i < 5; i++)
printf("%d\n", myRand(arr, freq, n));
return 0;
}

Output: May be different for different runs

4
3
4
4
4

How to check if a given number is Fibonacci number?

Given a number n, how to check if n is a Fibonacci number.
A simple way is to generate Fibonacci numbers until the generated number is greater than or equal to n. Following is an interesting property about
Fibonacci numbers that can also be used to check if a given number is Fibonacci or not.
A number is Fibonacci if and only if one or both of (5*n2 + 4) or (5*n2 4) is a perfect square (Source: Wiki). Following is a simple program
based on this concept.
// C++ program to check if x is a perfect square
#include <iostream>
#include <math.h>
using namespace std;
// A utility function that returns true if x is perfect square
bool isPerfectSquare(int x)
{
int s = sqrt(x);
return (s*s == x);
}
// Returns true if n is a Fibinacci Number, else false
bool isFibonacci(int n)
{
// n is Fibinacci if one of 5*n*n + 4 or 5*n*n - 4 or both
// is a perferct square
return isPerfectSquare(5*n*n + 4) ||
isPerfectSquare(5*n*n - 4);
}
// A utility function to test above functions
int main()
{
for (int i = 1; i <= 10; i++)
isFibonacci(i)? cout << i << " is a Fibonacci Number \n":
cout << i << " is a not Fibonacci Number \n" ;
return 0;
}

Output:
1 is a Fibonacci Number
2 is a Fibonacci Number
3 is a Fibonacci Number
4 is a not Fibonacci Number
5 is a Fibonacci Number
6 is a not Fibonacci Number
7 is a not Fibonacci Number
8 is a Fibonacci Number
9 is a not Fibonacci Number
10 is a not Fibonacci Number

Russian Peasant Multiplication

Given two integers, write a function to multiply them without using multiplication operator.
There are many other ways to multiply two numbers (For example, see this). One interesting method is the Russian peasant algorithm. The idea is
to double the first number and halve the second number repeatedly till the second number doesnt become 1. In the process, whenever the second
number become odd, we add the first number to result (result is initialized as 0)
The following is simple algorithm.
Let the two given numbers be 'a' and 'b'
1) Initialize result 'res' as 0.
2) Do following while 'b' is greater than 0
a) If 'b' is odd, add 'a' to 'res'
b) Double 'a' and halve 'b'
3) Return 'res'.
#include <iostream>
using namespace std;
// A method to multiply two numbers using Russian Peasant method
unsigned int russianPeasant(unsigned int a, unsigned int b)
{
int res = 0; // initialize result
// While second number doesn't become 1
while (b > 0)
{
// If second number becomes odd, add the first number to result
if (b & 1)
res = res + a;
// Double the first number and halve the second number
a = a << 1;
b = b >> 1;
}
return res;
}
// Driver program to test above function
int main()
{
cout << russianPeasant(18, 1) << endl;
cout << russianPeasant(20, 12) << endl;
return 0;
}

Output:
18
240

How does this work?

The value of a*b is same as (a*2)*(b/2) if b is even, otherwise the value is same as ((a*2)*(b/2) + a). In the while loop, we keep multiplying a
with 2 and keep dividing b by 2. If b becomes odd in loop, we add a to res. When value of b becomes 1, the value of res + a, gives us the result.
Note that when b is a power of 2, the res would remain 0 and a would have the multiplication. See the reference for more information.
Reference:
http://mathforum.org/dr.math/faq/faq.peasant.html

Count all possible groups of size 2 or 3 that have sum as multiple of 3

Given an unsorted integer (positive values only) array of size n, we can form a group of two or three, the group should be such that the sum of all
elements in that group is a multiple of 3. Count all possible number of groups that can be generated in this way.
Input: arr[] = {3, 6, 7, 2, 9}
Output: 8
// Groups are {3,6}, {3,9}, {9,6}, {7,2}, {3,6,9},
//
{3,7,2}, {7,2,6}, {7,2,9}
Input: arr[] = {2, 1, 3, 4}
Output: 4
// Groups are {2,1}, {2,4}, {2,1,3}, {2,4,3}

The idea is to see remainder of every element when divided by 3. A set of elements can form a group only if sun of their remainders is multiple of
3. Since the task is to enumerate groups, we count all elements with different remainders.
1. Hash all elements in a count array based on remainder, i.e,
for all elements a[i], do c[a[i]%3]++;
2. Now c[0] contains the number of elements which when divided
by 3 leave remainder 0 and similarly c[1] for remainder 1
and c[2] for 2.
3. Now for group of 2, we have 2 possibilities
a. 2 elements of remainder 0 group. Such possibilities are
c[0]*(c[0]-1)/2
b. 1 element of remainder 1 and 1 from remainder 2 group
Such groups are c[1]*c[2].
4. Now for group of 3,we have 4 possibilities
a. 3 elements from remainder group 0.
No. of such groups are c[0]C3
b. 3 elements from remainder group 1.
No. of such groups are c[1]C3
c. 3 elements from remainder group 2.
No. of such groups are c[2]C3
d. 1 element from each of 3 groups.
No. of such groups are c[0]*c[1]*c[2].
5. Add all the groups in steps 3 and 4 to obtain the result.
#include<stdio.h>
// Returns count of all possible groups that can be formed from elements
// of a[].
int findgroups(int arr[], int n)
{
// Create an array C[3] to store counts of elements with remainder
// 0, 1 and 2. c[i] would store count of elements with remainder i
int c[3] = {0}, i;
int res = 0; // To store the result
// Count elements with remainder 0, 1 and 2
for (i=0; i<n; i++)
c[arr[i]%3]++;
// Case 3.a: Count groups of size 2 from 0 remainder elements
res += ((c[0]*(c[0]-1))>>1);
// Case 3.b: Count groups of size 2 with one element with 1
// remainder and other with 2 remainder
res += c[1] * c[2];
// Case 4.a: Count groups of size 3 with all 0 remainder elements
res += (c[0] * (c[0]-1) * (c[0]-2))/6;
// Case 4.b: Count groups of size 3 with all 1 remainder elements
res += (c[1] * (c[1]-1) * (c[1]-2))/6;
// Case 4.c: Count groups of size 3 with all 2 remainder elements
res += ((c[2]*(c[2]-1)*(c[2]-2))/6);
// Case 4.c: Count groups of size 3 with different remainders
res += c[0]*c[1]*c[2];
// Return total count stored in res
return res;
}
// Driver program to test above functions
int main()

{
int arr[] = {3, 6, 7, 2, 9};
int n = sizeof(arr)/sizeof(arr[0]);
printf("Required number of groups are %d\n", findgroups(arr,n));
return 0;
}

Output:
Required number of groups are 8

Time Complexity: O(n)

Auxiliary Space: O(1)

C Program for Tower of Hanoi

Tower of Hanoi is a mathematical puzzle where we have three rods and n disks. The objective of the puzzle is to move the entire stack to another
rod, obeying the following simple rules:
1) Only one disk can be moved at a time.
2) Each move consists of taking the upper disk from one of the stacks and placing it on top of another stack i.e. a disk can only be moved if it is
the uppermost disk on a stack.
3) No disk may be placed on top of a smaller disk.
#include <stdio.h>
// C recursive function to solve tower of hanoi puzzle
void towerOfHanoi(int n, char fromrod, char torod, char auxrod)
{
if (n == 1)
{
printf("\n Move disk 1 from rod %c to rod %c", fromrod, torod);
return;
}
towerOfHanoi(n-1, fromrod, auxrod, torod);
printf("\n Move disk %d from rod %c to rod %c", n, fromrod, torod);
towerOfHanoi(n-1, auxrod, torod, fromrod);
}
int main()
{
int n = 4; // Number of disks
towerOfHanoi(n, 'A', 'C', 'B'); // A, B and C are names of rods
return 0;
}

Output:
Move
Move
Move
Move
Move
Move
Move
Move
Move
Move
Move
Move
Move
Move
Move

disk
disk
disk
disk
disk
disk
disk
disk
disk
disk
disk
disk
disk
disk
disk

1
2
1
3
1
2
1
4
1
2
1
3
1
2
1

from
from
from
from
from
from
from
from
from
from
from
from
from
from
from

rod
rod
rod
rod
rod
rod
rod
rod
rod
rod
rod
rod
rod
rod
rod

A
A
B
A
C
C
A
A
B
B
C
B
A
A
B

to
to
to
to
to
to
to
to
to
to
to
to
to
to
to

rod
rod
rod
rod
rod
rod
rod
rod
rod
rod
rod
rod
rod
rod
rod

B
C
C
B
A
B
B
C
C
A
A
C
B
C
C

For n disks, total 2n 1 moves are required.

References:
http://en.wikipedia.org/wiki/Tower_of_Hanoi

Horners Method for Polynomial Evaluation

Given a polynomial of the form cnxn + cn-1xn-1 + cn-2xn-2 + + c1x + c0 and a value of x, find the value of polynomial for a given value of x. Here
cn, cn-1, .. are integers (may be negative) and n is a positive integer.
Input is in the form of an array say poly[] where poly[0] represents coefficient for xn and poly[1] represents coefficient for xn-1 and so on.
Examples:
// Evaluate value of 2x3 - 6x2 + 2x - 1 for x = 3
Input: poly[] = {2, -6, 2, -1}, x = 3
Output: 5
// Evaluate value of 2x3 + 3x + 1 for x = 2
Input: poly[] = {2, 0, 3, 1}, x = 2
Output: 23

A naive way to evaluate a polynomial is to one by one evaluate all terms. First calculate xn, multiply the value with cn, repeat the same steps for
other terms and return the sum. Time complexity of this approach is O(n2) if we use a simple loop for evaluation of xn. Time complexity can be
improved to O(nLogn) if we use O(Logn) approach for evaluation of xn.
Horners method can be used to evaluate polynomial in O(n) time. To understand the method, let us consider the example of 2x3 6x2 + 2x 1. The
polynomial can be evaluated as ((2x 6)x + 2)x 1. The idea is to initialize result as coefficient of xn which is 2 in this case, repeatedly multiply result
with x and add next coefficient to result. Finally return result.
Following is C++ implementation of Horners Method.
#include <iostream>
using namespace std;
// returns value of poly[0]x(n-1) + poly[1]x(n-2) + .. + poly[n-1]
int horner(int poly[], int n, int x)
{
int result = poly[0]; // Initialize result
// Evaluate value of polynomial using Horner's method
for (int i=1; i<n; i++)
result = result*x + poly[i];
return result;
}
// Driver program to test above function.
int main()
{
// Let us evaluate value of 2x3 - 6x2 + 2x - 1 for x = 3
int poly[] = {2, -6, 2, -1};
int x = 3;
int n = sizeof(poly)/sizeof(poly[0]);
cout << "Value of polynomial is " << horner(poly, n, x);
return 0;
}

Output:
Value of polynomial is 5

Time Complexity: O(n)

Count trailing zeroes in factorial of a number

Given an integer n, write a function that returns count of trailing zeroes in n!.
Examples:
Input: n = 5
Output: 1
Factorial of 5 is 20 which has one trailing 0.
Input: n = 20
Output: 4
Factorial of 20 is 2432902008176640000 which has
4 trailing zeroes.
Input: n = 100
Output: 24

A simple method is to first calculate factorial of n, then count trailing 0s in the result (We can count trailing 0s by repeatedly dividing the factorial by
10 till the remainder is 0).
The above method can cause overflow for a slightly bigger numbers as factorial of a number is a big number (See factorial of 20 given in above
examples). The idea is to consider prime factors of a factorial n. A trailing zero is always produced by prime factors 2 and 5. If we can count the
number of 5s and 2s, our task is done. Consider the following examples.
n = 5: There is one 5 and 3 2s in prime factors of 5! (2 * 2 * 2 * 3 * 5). So count of trailing 0s is 1.
n = 11: There are two 5s and three 2s in prime factors of 11! (2 8 * 34 * 52 * 7). So count of trailing 0s is 2.
We can easily observe that the number of 2s in prime factors is always more than or equal to the number of 5s. So if we count 5s in prime factors,
we are done. How to count total number of 5s in prime factors of n!? A simple way is to calculate floor(n/5). For example, 7! has one 5, 10!
has two 5s. It is done yet, there is one more thing to consider. Numbers like 25, 125, etc have more than one 5. For example if we consider 28!,
we get one extra 5 and number of 0s become 6. Handling this is simple, first divide n by 5 and remove all single 5s, then divide by 25 to remove
extra 5s and so on. Following is the summarized formula for counting trailing 0s.
Trailing 0s in n! = Count of 5s in prime factors of n!
= floor(n/5) + floor(n/25) + floor(n/125) + ....

Following is C++ program based on above formula.

// C++ program to count trailing 0s in n!
#include <iostream>
using namespace std;
// Function to return trailing 0s in factorial of n
int findTrailingZeros(int n)
{
// Initialize result
int count = 0;
// Keep dividing n by powers of 5 and update count
for (int i=5; n/i>=1; i *= 5)
count += n/i;
return count;
}
// Driver program to test above function
int main()
{
int n = 100;
cout << "Count of trailing 0s in " << 100
<< "! is " << findTrailingZeros(n);
return 0;
}

Output:
Count of trailing 0s in 100! is 24

Program for nth Catalan Number

Catalan numbers are a sequence of natural numbers that occurs in many interesting counting problems like following.
1) Count the number of expressions containing n pairs of parentheses which are correctly matched. For n = 3, possible expressions are ((())), ()
(()), ()()(), (())(), (()()).
2) Count the number of possible Binary Search Trees with n keys (See this)
3) Count the number of full binary trees (A rooted binary tree is full if every vertex has either two children or no children) with n+1 leaves.
See this for more applications.
The first few Catalan numbers for n = 0, 1, 2, 3, are 1, 1, 2, 5, 14, 42, 132, 429, 1430, 4862,
Recursive Solution
Catalan numbers satisfy the following recursive formula.
Following is C++ implementation of above recursive formula.
#include<iostream>
using namespace std;
// A recursive function to find nth catalan number
unsigned long int catalan(unsigned int n)
{
// Base case
if (n <= 1) return 1;
// catalan(n) is sum of catalan(i)*catalan(n-i-1)
unsigned long int res = 0;
for (int i=0; i<n; i++)
res += catalan(i)*catalan(n-i-1);
return res;
}
// Driver program to test above function
int main()
{
for (int i=0; i<10; i++)
cout << catalan(i) << " ";
return 0;
}

Output :
1 1 2 5 14 42 132 429 1430 4862

Time complexity of above implementation is equivalent to nth catalan number.

The value of nth catalan number is exponential that makes the time complexity exponential.

Dynamic Programming Solution

We can observe that the above recursive implementation does a lot of repeated work (we can the same by drawing recursion tree). Since there
are overlapping subproblems, we can use dynamic programming for this. Following is a Dynamic programming based implementation in C++.
#include<iostream>
using namespace std;
// A dynamic programming based function to find nth
// Catalan number
unsigned long int catalanDP(unsigned int n)
{
// Table to store results of subproblems
unsigned long int catalan[n+1];
// Initialize first two values in table
catalan[0] = catalan[1] = 1;
// Fill entries in catalan[] using recursive formula

for (int i=2; i<=n; i++)

{
catalan[i] = 0;
for (int j=0; j<i; j++)
catalan[i] += catalan[j] * catalan[i-j-1];
}
// Return last entry
return catalan[n];
}
// Driver program to test above function
int main()
{
for (int i = 0; i < 10; i++)
cout << catalanDP(i) << " ";
return 0;
}

Output:
1 1 2 5 14 42 132 429 1430 4862

Time Complexity: Time complexity of above implementation is O(n2)

Using Binomial Coefficient

We can also use the below formula to find nth catalan number in O(n) time.
We have discussed a O(n) approach to find binomial coefficient nCr.
#include<iostream>
using namespace std;
// Returns value of Binomial Coefficient C(n, k)
unsigned long int binomialCoeff(unsigned int n, unsigned int k)
{
unsigned long int res = 1;
// Since C(n, k) = C(n, n-k)
if (k > n - k)
k = n - k;
// Calculate value of [n*(n-1)*---*(n-k+1)] / [k*(k-1)*---*1]
for (int i = 0; i < k; ++i)
{
res *= (n - i);
res /= (i + 1);
}
return res;
}
// A Binomial coefficient based function to find nth catalan
// number in O(n) time
unsigned long int catalan(unsigned int n)
{
// Calculate value of 2nCn
unsigned long int c = binomialCoeff(2*n, n);
// return 2nCn/(n+1)
return c/(n+1);
}
// Driver program to test above functions
int main()
{
for (int i = 0; i < 10; i++)
cout << catalan(i) << " ";
return 0;
}

Output:
1 1 2 5 14 42 132 429 1430 4862

Time Complexity: Time complexity of above implementation is O(n).

We can also use below formula to find nth catalan number in O(n) time.

Write a function that generates one of 3 numbers according to given probabilities

You are given a function rand(a, b) which generates equiprobable random numbers between [a, b] inclusive. Generate 3 numbers x, y, z with
probability P(x), P(y), P(z) such that P(x) + P(y) + P(z) = 1 using the given rand(a,b) function.
The idea is to utilize the equiprobable feature of the rand(a,b) provided. Let the given probabilities be in percentage form, for example
P(x)=40%, P(y)=25%, P(z)=35%..
Following are the detailed steps.
1) Generate a random number between 1 and 100. Since they are equiprobable, the probability of each number appearing is 1/100.
2) Following are some important points to note about generated random number r.
a) r is smaller than or equal to P(x) with probability P(x)/100.
b) r is greater than P(x) and smaller than or equal P(x) + P(y) with P(y)/100.
c) r is greater than P(x) + P(y) and smaller than or equal 100 (or P(x) + P(y) + P(z)) with probability P(z)/100.
// This function generates 'x' with probability px/100, 'y' with
// probability py/100 and 'z' with probability pz/100:
// Assumption: px + py + pz = 100 where px, py and pz lie
// between 0 to 100
int random(int x, int y, int z, int px, int py, int pz)
{
// Generate a number from 1 to 100
int r = rand(1, 100);
// r is smaller than px with probability px/100
if (r <= px)
return x;
// r is greater than px and smaller than or equal to px+py
// with probability py/100
if (r <= (px+py))
return y;
// r is greater than px+py and smaller than or equal to 100
// with probability pz/100
else
return z;
}

This function will solve the purpose of generating 3 numbers with given three probabilities.

Find Excel column name from a given column number

MS Excel columns has a pattern like A, B, C, ,Z, AA, AB, AC,. ,AZ, BA, BB, ZZ, AAA, AAB .. etc. In other words, column 1 is named as A,
column 2 as B, column 27 as AA.
Given a column number, find its corresponding Excel column name. Following are more examples.
Input
26
51
52
80
676
702
705

Output
Z
AY
AZ
CB
YZ
ZZ
AAC

Thanks to Mrigank Dembla for suggesting the below solution in a comment.

Suppose we have a number n, lets say 28. so corresponding to it we need to print the column name. We need to take remainder with 26.
If remainder with 26 comes out to be 0 (meaning 26, 52 and so on) then we put Z in the output string and new n becomes n/26 -1 because here
we are considering 26 to be Z while in actual its 25th with respect to A.
Similarly if the remainder comes out to be non zero. (like 1, 2, 3 and so on) then we need to just insert the char accordingly in the string and do n =
n/26.
Finally we reverse the string and print.
Example:
n = 700
Remainder (n%26) is 24. So we put X in output string and n becomes n/26 which is 26.
Remainder (26%26) is 0. So we put Z in output string and n becomes n/26 -1 which is 0.
Following is C++ implementation of above approach.
#include<iostream>
#include<cstring>
#define MAX 50
using namespace std;
// Function to print Excel column name for a given column number
void printString(int n)
{
char str[MAX]; // To store result (Excel column name)
int i = 0; // To store current index in str which is result
while (n>0)
{
// Find remainder
int rem = n%26;
// If remainder is 0, then a 'Z' must be there in output
if (rem==0)
{
str[i++] = 'Z';
n = (n/26)-1;
}
else // If remainder is non-zero
{
str[i++] = (rem-1) + 'A';
n = n/26;
}
}
str[i] = '\0';
// Reverse the string and print result
cout << strrev(str) << endl;
return;
}
// Driver program to test above function
int main()
{
printString(26);

printString(51);
printString(52);
printString(80);
printString(676);
printString(702);
printString(705);
return 0;
}

Output
Z
AY
AZ
CB
YZ
ZZ
AAC

Find next greater number with same set of digits

Given a number n, find the smallest number that has same set of digits as n and is greater than n. If x is the greatest possible number with its set of
digits, then print not possible.
Examples:
For simplicity of implementation, we have considered input number as a string.
Input: n = "218765"
Output: "251678"
Input: n = "1234"
Output: "1243"
Input: n = "4321"
Output: "Not Possible"
Input: n = "534976"
Output: "536479"

Following are few observations about the next greater number.

1) If all digits sorted in descending order, then output is always Not Possible. For example, 4321.
2) If all digits are sorted in ascending order, then we need to swap last two digits. For example, 1234.
3) For other cases, we need to process the number from rightmost side (why? because we need to find the smallest of all greater numbers)
You can now try developing an algorithm yourself.
Following is the algorithm for finding the next greater number.
I) Traverse the given number from rightmost digit, keep traversing till you find a digit which is smaller than the previously traversed digit. For
example, if the input number is 534976, we stop at 4 because 4 is smaller than next digit 9. If we do not find such a digit, then output is Not
Possible.
II) Now search the right side of above found digit d for the smallest digit greater than d. For 534976?, the right side of 4 contains 976. The
smallest digit greater than 4 is 6.
III) Swap the above found two digits, we get 536974 in above example.
IV) Now sort all digits from position next to d to the end of number. The number that we get after sorting is the output. For above example, we
sort digits in bold 536974. We get 536479 which is the next greater number for input 534976.
Following is C++ implementation of above approach.
// C++ program to find the smallest number which greater than a given number
// and has same set of digits as given number
#include <iostream>
#include <cstring>
#include <algorithm>
using namespace std;
// Utility function to swap two digits
void swap(char *a, char *b)
{
char temp = *a;
*a = *b;
*b = temp;
}
// Given a number as a char array number[], this function finds the
// next greater number. It modifies the same array to store the result
void findNext(char number[], int n)
{
int i, j;
// I) Start from the right most digit and find the first digit that is
// smaller than the digit next to it.
for (i = n-1; i > 0; i--)
if (number[i] > number[i-1])
break;
// If no such digit is found, then all digits are in descending order
// means there cannot be a greater number with same set of digits
if (i==0)
{
cout << "Next number is not possible";
return;
}

// II) Find the smallest digit on right side of (i-1)'th digit that is
// greater than number[i-1]
int x = number[i-1], smallest = i;
for (j = i+1; j < n; j++)
if (number[j] > x && number[j] < number[smallest])
smallest = j;
// III) Swap the above found smallest digit with number[i-1]
swap(&number[smallest], &number[i-1]);
// IV) Sort the digits after (i-1) in ascending order
sort(number + i, number + n);
cout << "Next number with same set of digits is " << number;
return;
}
// Driver program to test above function
int main()
{
char digits[] = "534976";
int n = strlen(digits);
findNext(digits, n);
return 0;
}

Output:
Next number with same set of digits is 536479

The above implementation can be optimized in following ways.

1) We can use binary search in step II instead of linear search.
2) In step IV, instead of doing simple sort, we can apply some clever technique to do it in linear time. Hint: We know that all digits are linearly
sorted in reverse order except one digit which was swapped.
With above optimizations, we can say that the time complexity of this method is O(n).

Count Possible Decodings of a given Digit Sequence

Let 1 represent A, 2 represents B, etc. Given a digit sequence, count the number of possible decodings of the given digit sequence.
Examples:
Input: digits[] = "121"
Output: 3
// The possible decodings are "ABA", "AU", "LA"
Input: digits[] = "1234"
Output: 3
// The possible decodings are "ABCD", "LCD", "AWD"

An empty digit sequence is considered to have one decoding. It may be assumed that the input contains valid digits from 0 to 9 and there are no
leading 0s, no extra trailing 0s and no two or more consecutive 0s.
This problem is recursive and can be broken in sub-problems. We start from end of the given digit sequence. We initialize the total count of
decodings as 0. We recur for two subproblems.
1) If the last digit is non-zero, recur for remaining (n-1) digits and add the result to total count.
2) If the last two digits form a valid character (or smaller than 27), recur for remaining (n-2) digits and add the result to total count.
Following is C++ implementation of the above approach.
// A naive recursive C++ implementation to count number of decodings
// that can be formed from a given digit sequence
#include <iostream>
#include <cstring>
using namespace std;
// Given a digit sequence of length n, returns count of possible
// decodings by replacing 1 with A, 2 woth B, ... 26 with Z
int countDecoding(char *digits, int n)
{
// base cases
if (n == 0 || n == 1)
return 1;
int count = 0; // Initialize count
// If the last digit is not 0, then last digit must add to
// the number of words
if (digits[n-1] > '0')
count = countDecoding(digits, n-1);
// If the last two digits form a number smaller than or equal to 26,
// then consider last two digits and recur
if (digits[n-2] < '2' || (digits[n-2] == '2' && digits[n-1] < '7') )
count += countDecoding(digits, n-2);
return count;
}
// Driver program to test above function
int main()
{
char digits[] = "1234";
int n = strlen(digits);
cout << "Count is " << countDecoding(digits, n);
return 0;
}

Output:
Count is 3

The time complexity of above the code is exponential. If we take a closer look at the above program, we can observe that the recursive solution is
similar to Fibonacci Numbers. Therefore, we can optimize the above solution to work in O(n) time using Dynamic Programming. Following is C++
implementation for the same.
// A Dynamic Programming based C++ implementation to count decodings
#include <iostream>
#include <cstring>
using namespace std;
// A Dynamic Programming based function to count decodings
int countDecodingDP(char *digits, int n)

{
int count[n+1]; // A table to store results of subproblems
count[0] = 1;
count[1] = 1;
for (int i = 2; i <= n; i++)
{
count[i] = 0;
// If the last digit is not 0, then last digit must add to
// the number of words
if (digits[i-1] > '0')
count[i] = count[i-1];
// If second last digit is smaller than 2 and last digit is
// smaller than 7, then last two digits form a valid character
if (digits[i-2] < '2' || (digits[i-2] == '2' && digits[i-1] < '7') )
count[i] += count[i-2];
}
return count[n];
}
// Driver program to test above function
int main()
{
char digits[] = "1234";
int n = strlen(digits);
cout << "Count is " << countDecodingDP(digits, n);
return 0;
}

Output:
Count is 3

Time Complexity of the above solution is O(n) and it requires O(n) auxiliary space. We can reduce auxiliary space to O(1) by using space
optimized version discussed in the Fibonacci Number Post.

Calculate the angle between hour hand and minute hand

This problem is know as Clock angle problem where we need to find angle between hands of an analog clock at .
Examples:
Input: h = 12:00, m = 30.00
Output: 165 degree
Input: h = 3.00, m = 30.00
Output: 75 degree

The idea is to take 12:00 (h = 12, m = 0) as a reference. Following are detailed steps.
1) Calculate the angle made by hour hand with respect to 12:00 in h hours and m minutes.
2) Calculate the angle made by minute hand with respect to 12:00 in h hours and m minutes.
3) The difference between two angles is the angle between two hands.
How to calculate the two angles with respect to 12:00?
The minute hand moves 360 degree in 60 minute(or 6 degree in one minute) and hour hand moves 360 degree in 12 hours(or 0.5 degree in 1
minute). In h hours and m minutes, the minute hand would move (h*60 + m)*6 and hour hand would move (h*60 + m)*0.5.
// C program to find angle between hour and minute hands
#include <stdio.h>
#include <stdlib.h>
// Utility function to find minimum of two integers
int min(int x, int y) { return (x < y)? x: y; }
int calcAngle(double h, double m)
{
// validate the input
if (h <0 || m < 0 || h >12 || m > 60)
printf("Wrong input");
if (h == 12) h = 0;
if (m == 60) m = 0;
// Calculate the angles moved by hour and minute hands
// with reference to 12:00
int hour_angle = 0.5 * (h*60 + m);
int minute_angle = 6*m;
// Find the difference between two angles
int angle = abs(hour_angle - minute_angle);
// Return the smaller angle of two possible angles
angle = min(360-angle, angle);
return angle;
}
// Driver program to test above function
int main()
{
printf("%d \n", calcAngle(9, 60));
printf("%d \n", calcAngle(3, 30));
return 0;
}

Output:
90
75

Exercise: Find all times when hour and minute hands get superimposed.

Count number of binary strings without consecutive 1s

Given a positive integer N, count all possible distinct binary strings of length N such that there are no consecutive 1s.
Examples:
Input: N = 2
Output: 3
// The 3 strings are 00, 01, 10
Input: N = 3
Output: 5
// The 5 strings are 000, 001, 010, 100, 101

This problem can be solved using Dynamic Programming. Let a[i] be the number of binary strings of length i which do not contain any two
consecutive 1s and which end in 0. Similarly, let b[i] be the number of such strings which end in 1. We can append either 0 or 1 to a string ending
in 0, but we can only append 0 to a string ending in 1. This yields the recurrence relation:
a[i] = a[i - 1] + b[i - 1]
b[i] = a[i - 1]

The base cases of above recurrence are a[1] = b[1] = 1. The total number of strings of length i is just a[i] + b[i].
Following is C++ implementation of above solution. In the following implementation, indexes start from 0. So a[i] represents the number of binary
strings for input length i+1. Similarly, b[i] represents binary strings for input length i+1.
// C++ program to count all distinct binary strings
// without two consecutive 1's
#include <iostream>
using namespace std;
int countStrings(int n)
{
int a[n], b[n];
a[0] = b[0] = 1;
for (int i = 1; i < n; i++)
{
a[i] = a[i-1] + b[i-1];
b[i] = a[i-1];
}
return a[n-1] + b[n-1];
}
// Driver program to test above functions
int main()
{
cout << countStrings(3) << endl;
return 0;
}

Output:
5

Source:
courses.csail.mit.edu/6.006/oldquizzes/solutions/q2-f2009-sol.pdf

Find the smallest number whose digits multiply to a given number n

Given a number n, find the smallest number p such that if we multiply all digits of p, we get n. The result p should have minimum two digits.
Examples:
Input: n = 36
Output: p = 49
// Note that 4*9 = 36 and 49 is the smallest such number
Input: n = 100
Output: p = 455
// Note that 4*5*5 = 100 and 455 is the smallest such number
Input: n = 1
Output:p = 11
// Note that 1*1 = 1
Input: n = 13
Output: Not Possible

For a given n, following are the two cases to be considered.

Case 1: n < 10 When n is smaller than n, the output is always n+10. For example for n = 7, output is 17. For n = 9, output is 19.
Case 2: n >= 10 Find all factors of n which are between 2 and 9 (both inclusive). The idea is to start searching from 9 so that the number of digits
in result are minimized. For example 9 is preferred over 33 and 8 is preferred over 24.
Store all found factors in an array. The array would contain digits in non-increasing order, so finally print the array in reverse order.
Following is C implementation of above concept.
#include<stdio.h>
// Maximum number of digits in output
#define MAX 50
// prints the smallest number whose digits multiply to n
void findSmallest(int n)
{
int i, j=0;
int res[MAX]; // To sore digits of result in reverse order
// Case 1: If number is smaller than 10
if (n < 10)
{
printf("%d", n+10);
return;
}
// Case 2: Start with 9 and try every possible digit
for (i=9; i>1; i--)
{
// If current digit divides n, then store all
// occurrences of current digit in res
while (n%i == 0)
{
n = n/i;
res[j] = i;
j++;
}
}
// If n could not be broken in form of digits (prime factors of n
// are greater than 9)
if (n > 10)
{
printf("Not possible");
return;
}
// Print the result array in reverse order
for (i=j-1; i>=0; i--)
printf("%d", res[i]);
}
// Driver program to test above function
int main()
{
findSmallest(7);

printf("\n");
findSmallest(36);
printf("\n");
findSmallest(13);
printf("\n");
findSmallest(100);
return 0;
}

Output:
17
49
Not possible
455

Draw a circle without floating point arithmetic

Given a radius of a circle, draw the circle without using floating point arithmetic.
Following program uses a simple concept. Let the radius of the circle be r. Consider a square of size (2r+1)*(2r+1) around the circle to be drawn.
Now walk through every point inside the square. For every every point (x,y), if (x, y) lies inside the circle (or x^2+ y^2 < r^2), then print it,
otherwise print space.
#include <stdio.h>
void drawCircle(int r)
{
// Consider a rectangle of size N*N
int N = 2*r+1;
int x, y; // Coordinates inside the rectangle
// Draw a square of size N*N.
for (int i = 0; i < N; i++)
{
for (int j = 0; j < N; j++)
{
// Start from the left most corner point
x = i-r;
y = j-r;
// If this point is inside the circle, print it
if (x*x + y*y <= r*r+1 )
printf(".");
else // If outside the circle, print space
printf(" ");
printf(" ");
}
printf("\n");
}
}
// Driver Program to test above function
int main()
{
drawCircle(8);
return 0;
}

Output:

How to check if an instance of 8 puzzle is solvable?

What is 8 puzzle?
Given a 33 board with 8 tiles (every tile has one number from 1 to 8) and one empty space. The objective is to place the numbers on tiles in order
using the empty space. We can slide four adjacent (left, right, above and below) tiles into the empty space.

How to find if given state is solvable?

Following are two examples, the first example can reach goal state by a series of slides. The second example cannot.

Following is simple rule to check if a 8 puzzle is solvable.

It is not possible to solve an instance of 8 puzzle if number of inversions is odd in the input state. In the examples given in above figure, the
first example has 10 inversions, therefore solvable. The second example has 11 inversions, therefore unsolvable.
What is inversion?
A pair of tiles form an inversion if the the values on tiles are in reverse order of their appearance in goal state. For example, the following instance
of 8 puzzle has two inversions, (8, 6) and (8, 7).
1
4
8

2
_
6

3
5
7

Following is a simple C++ program to check whether a given instance of 8 puzzle is solvable or not. The idea is simple, we count inversions in the
given 8 puzzle.
// C++ program to check if a given instance of 8 puzzle is solvable or not
#include <iostream>
using namespace std;
// A utility function to count inversions in given array 'arr[]'
int getInvCount(int arr[])
{
int inv_count = 0;
for (int i = 0; i < 9 - 1; i++)
for (int j = i+1; j < 9; j++)
// Value 0 is used for empty space
if (arr[j] && arr[i] && arr[i] > arr[j])
inv_count++;
return inv_count;
}
// This function returns true if given 8 puzzle is solvable.
bool isSolvable(int puzzle[3][3])
{
// Count inversions in given 8 puzzle
int invCount = getInvCount((int *)puzzle);
// return true if inversion count is even.
return (invCount%2 == 0);
}

/* Driver progra to test above functions */

int main(int argv, int** args)
{
int puzzle[3][3] = {{1, 8, 2},
{0, 4, 3}, // Value 0 is used for empty space
{7, 6, 5}};
isSolvable(puzzle)? cout << "Solvable":
cout << "Not Solvable";
return 0;
}

Output:
Solvable

Note that the above implementation uses simple algorithm for inversion count. It is done this way for simplicity. The code can be optimized to
O(nLogn) using the merge sort based algorithm for inversion count.
How does this work?
The idea is based on the fact the parity of inversions remains same after a set of moves, i.e., if the inversion count is odd in initial stage, then it
remain odd after any sequence of moves and if the inversion count is even, then it remains even after any sequence of moves. In the goal state,
there are 0 inversions. So we can reach goal state only from a state which has even inversion count.
How parity of inversion count is invariant?
When we slide a tile, we either make a row move (moving a left or right tile into the blank space), or make a column move (moving a up or down
tile to the blank space).
a) A row move doesnt change the inversion count. See following example
1 2 3
Row Move
4 _ 5 ---------->
8 6 7
Inversion count remains 2

1 2 3
_ 4 5
8 6 7
after the move

1 2 3
Row Move
4 _ 5 ---------->
8 6 7
Inversion count remains 2

1 2 3
4 5 _
8 6 7
after the move

b) A column move does one of the following three.

..(i) Increases inversion count by 2. See following example.
1 2
4 _
8 6
Inversion

3 Column Move
5 ----------->
7
count increases by 2

1 _ 3
4 2 5
8 6 7
(changes from 2 to 4)

..(ii) Decreases inversion count by 2

1 3
5 _
7 2
Inversion

4
Column Move
6 ------------>
8
count decreases by 2

1 3
5 2
7 _
(changes

4
6
8
from 5 to 3)

..(iii) Keeps the inversion count same.

1 2 3
Column Move
1 2 3
4 _ 5 ------------>
4 6 5
7 6 8
7 _ 8
Inversion count remains 1 after the move

So if a move either increases/decreases inversion count by 2, or keeps the inversion count same, then it is not possible to change parity of a state
by any sequence of row/column moves.
Exercise: How to check if a given instance of 15 puzzle is solvable or not. In a 15 puzzle, we have 44 board where 15 tiles have a number and
one empty space. Note that the above simple rules of inversion count dont directly work for 15 puzzle, the rules need to be modified for 15 puzzle.

Birthday Paradox
How many people must be there in a room to make the probability 100% that two people in the room have same birthday?
Answer: 367 (since there are 366 possible birthdays, including February 29).
The above question was simple. Try the below question yourself.
How many people must be there in a room to make the probability 50% that two people in the room have same birthday?
Answer: 23
The number is surprisingly very low. In fact, we need only 70 people to make the probability 99.9 %.
Let us discuss the generalized formula.
What is the probability that two persons among n have same birthday?
Let the probability that two people in a room with n have same birthday be P(same). P(Same) can be easily evaluated in terms of P(different)
where P(different) is the probability that all of them have different birthday.
P(same) = 1 P(different)
P(different) can be written as 1 x (364/365) x (363/365) x (362/365) x . x (1 (n-1)/365)
How did we get the above expression?
Persons from first to last can get birthdays in following order for all birthdays to be distinct:
The first person can have any birthday among 365
The second person should have a birthday which is not same as first person
The third person should have a birthday which is not same as first two persons.
.
The nth person should have a birthday which is not same as any of the earlier considered (n-1) persons.
Approximation of above expression
The above expression can be approximated using Taylors Series.
provides a first-order approximation for ex for x << 1:
To apply this approximation to the first expression derived for p(different), set x = -a / 365. Thus,

The above expression derived for p(different) can be written as

1 x (1 1/365) x (1 2/365) x (1 3/365) x . x (1 (n-1)/365)
By putting the value of 1 a/365 as e-a/365, we get following.

Therefore,
p(same) = 1- p(different)
An even coarser approximation is given by
p(same)
By taking Log on both sides, we get the reverse formula.

Using the above approximate formula, we can approximate number of people for a given probability. For example the following C++ function
find() returns the smallest n for which the probability is greater than the given p.
C++ Implementation of approximate formula.
The following is C++ program to approximate number of people for a given probability.

// C++ program to approximate number of people in Birthday Paradox

// problem
#include <cmath>
#include <iostream>
using namespace std;
// Returns approximate number of people for a given probability
int find(double p)
{
return ceil(sqrt(2*365*log(1/(1-p))));
}
int main()
{
cout << find(0.70);
}

Output:
30

Source:
http://en.wikipedia.org/wiki/Birthday_problem
Applications:
1) Birthday Paradox is generally discussed with hashing to show importance of collision handling even for a small set of keys.
2) Birthday Attack

Multiply two polynomials

Given two polynomials represented by two arrays, write a function that multiplies given two polynomials.
Example:
Input: A[] = {5, 0, 10, 6}
B[] = {1, 2, 4}
Output: prod[] = {5, 10, 30, 26, 52, 24}
The first input array represents "5 + 0x^1 + 10x^2 + 6x^3"
The second array represents "1 + 2x^1 + 4x^2"
And Output is "5 + 10x^1 + 30x^2 + 26x^3 + 52x^4 + 24x^5"

A simple solution is to one by one consider every term of first polynomial and multiply it with every term of second polynomial. Following is
algorithm of this simple method.
multiply(A[0..m-1], B[0..n01])
1) Create a product array prod[] of size m+n-1.
2) Initialize all entries in prod[] as 0.
3) Travers array A[] and do following for every element A[i]
...(3.a) Traverse array B[] and do following for every element B[j]
prod[i+j] = prod[i+j] + A[i] * B[j]
4) Return prod[].

The following is C++ implementation of above algorithm.

// Simple C++ program to multiply two polynomials
#include <iostream>
using namespace std;
// A[] represents coefficients of first polynomial
// B[] represents coefficients of second polynomial
// m and n are sizes of A[] and B[] respectively
int *multiply(int A[], int B[], int m, int n)
{
int *prod = new int[m+n-1];
// Initialize the porduct polynomial
for (int i = 0; i<m+n-1; i++)
prod[i] = 0;
// Multiply two polynomials term by term
// Take ever term of first polynomial
for (int i=0; i<m; i++)
{
// Multiply the current term of first polynomial
// with every term of second polynomial.
for (int j=0; j<n; j++)
prod[i+j] += A[i]*B[j];
}
return prod;
}
// A utility function to print a polynomial
void printPoly(int poly[], int n)
{
for (int i=0; i<n; i++)
{
cout << poly[i];
if (i != 0)
cout << "x^" << i ;
if (i != n-1)
cout << " + ";
}
}
// Driver program to test above functions
int main()
{
// The following array represents polynomial 5 + 10x^2 + 6x^3
int A[] = {5, 0, 10, 6};
// The following array represents polynomial 1 + 2x + 4x^2
int B[] = {1, 2, 4};
int m = sizeof(A)/sizeof(A[0]);
int n = sizeof(B)/sizeof(B[0]);

cout << "First polynomial is \n";

printPoly(A, m);
cout << "\nSecond polynomial is \n";
printPoly(B, n);
int *prod = multiply(A, B, m, n);
cout << "\nProduct polynomial is \n";
printPoly(prod, m+n-1);
return 0;
}

Output
First polynomial is
5 + 0x^1 + 10x^2 + 6x^3
Second polynomial is
1 + 2x^1 + 4x^2
Product polynomial is
5 + 10x^1 + 30x^2 + 26x^3 + 52x^4 + 24x^5

Time complexity of the above solution is O(mn). If size of two polynomials same, then time complexity is O(n2).
Can we do better?
There are methods to do multiplication faster than O(n2) time. These methods are mainly based on divide and conquer. Following is one simple
method that divides the given polynomial (of degree n) into two polynomials one containing lower degree terms(lower than n/2) and other
containing higher degree terns (higher than or equal to n/2)
Let the two given polynomials be A and B.
For simplicity, Let us assume that the given two polynomials are of
same degree and have degree in powers of 2, i.e., n = 2i
The polynomial 'A' can be written as A0 + A1*xn/2
The polynomial 'B' can be written as B0 + B1*xn/2
For example 1 + 10x + 6x2 - 4x3 + 5x4 can be
written as (1 + 10x) + (6 - 4x + 5x2)*x2
A * B = (A0 + A1*xn/2) * (B0 + B1*xn/2)
= A0*B0 + A0*B1*xn/2 + A1*B0*xn/2 + A1*B1*xn
= A0*B0 + (A0*B1 + A1*B0)xn/2 + A1*B1*xn

So the above divide and conquer approach requires 4 multiplications and O(n) time to add all 4 results. Therefore the time complexity is T(n) =
4T(n/2) + O(n). The solution of the recurrence is O(n2) which is same as the above simple solution.
The idea is to reduce number of multiplications to 3 and make the recurrence as T(n) = 3T(n/2) + O(n)
How to reduce number of multiplications?
This requires a little trick similar to Strassens Matrix Multiplication. We do following 3 multiplications.
X = (A0 + A1)*(B0 + B1) // First Multiplication
Y = A0B0 // Second
Z = A1B1 // Third
The missing middle term in above multiplication equation A0*B0 + (A0*B1 +
A1*B0)xn/2 + A1*B1*xn can obtained using below.
A0B1 + A1B0 = X - Y - Z

So the time taken by this algorithm is T(n) = 3T(n/2) + O(n)

The solution of above recurrence is O(nLg3) which is better than O(n2).
We will soon be discussing implementation of above approach.
There is a O(nLogn) algorithm also that uses Fast Fourier Transform to multiply two polynomials (Refer this and this for details)
Sources:
http://www.cse.ust.hk/~dekai/271/notes/L03/L03.pdf

Count Distinct Non-Negative Integer Pairs (x, y) that Satisfy the Inequality x*x + y*y
<n
Given a positive number n, count all distinct Non-Negative Integer pairs (x, y) that satisfy the inequality x*x + y*y < n.
Examples:
Input: n = 5
Output: 6
The pairs are (0, 0), (0, 1), (1, 0), (1, 1), (2, 0), (0, 2)
Input: n = 6
Output: 8
The pairs are (0, 0), (0, 1), (1, 0), (1, 1), (2, 0), (0, 2),
(1, 2), (2, 1)

A Simple Solution is to run two loops. The outer loop goes for all possible values of x (from 0 to ?n). The inner loops picks all possible values of
y for current value of x (picked by outer loop). Following is C++ implementation of simple solution.
#include <iostream>
using namespace std;
// This function counts number of pairs (x, y) that satisfy
// the inequality x*x + y*y < n.
int countSolutions(int n)
{
int res = 0;
for (int x = 0; x*x < n; x++)
for (int y = 0; x*x + y*y < n; y++)
res++;
return res;
}
// Driver program to test above function
int main()
{
cout << "Total Number of distinct Non-Negative pairs is "
<< countSolutions(6) << endl;
return 0;
}

Output:
Total Number of distinct Non-Negative pairs is 8

An upper bound for time complexity of the above solution is O(n). The outer loop runs ?n times. The inner loop runs at most ?n times.
Using an Efficient Solution, we can find the count in O(?n) time. The idea is to first find the count of all y values corresponding the 0 value of x.
Let count of distinct y values be yCount. We can find yCount by running a loop and comparing yCount*yCount with n.
After we have initial yCount, we can one by one increase value of x and find the next value of yCount by reducing yCount.
// An efficient C program to find different (x, y) pairs that
// satisfy x*x + y*y < n.
#include <iostream>
using namespace std;
// This function counts number of pairs (x, y) that satisfy
// the inequality x*x + y*y < n.
int countSolutions(int n)
{
int x = 0, yCount, res = 0;
// Find the count of different y values for x = 0.
for (yCount = 0; yCount*yCount < n; yCount++) ;
// One by one increase value of x, and find yCount for
// current x. If yCount becomes 0, then we have reached
// maximum possible value of x.
while (yCount != 0)
{
// Add yCount (count of different possible values of y
// for current x) to result
res += yCount;
// Increment x
x++;

// Update yCount for current x. Keep reducing yCount while

// the inequality is not satisfied.
while (yCount != 0 && (x*x + (yCount-1)*(yCount-1) >= n))
yCount--;
}
return res;
}
// Driver program to test above function
int main()
{
cout << "Total Number of distinct Non-Negative pairs is "
<< countSolutions(6) << endl;
return 0;
}

Output:
Total Number of distinct Non-Negative pairs is 8

Time Complexity of the above solution seems more but if we take a closer look, we can see that it is O(?n). In every step inside the inner loop,
value of yCount is decremented by 1. The value yCount can decrement at most O(?n) times as yCount is count y values for x = 0. In the outer
loop, the value of x is incremented. The value of x can also increment at most O(?n) times as the last x is for yCount equals to 1.

Count ways to reach the nth stair

There are n stairs, a person standing at the bottom wants to reach the top. The person can climb either 1 stair or 2 stairs at a time. Count the
number of ways, the person can reach the top.

Consider the example shown in diagram. The value of n is 3. There are 3 ways to reach the top. The diagram is taken from Easier Fibonacci
puzzles

More Examples:
Input: n = 1
Output: 1
There is only one way to climb 1 stair
Input: n = 2
Output: 2
There are two ways: (1, 1) and (2)
Input: n = 4
Output: 5
(1, 1, 1, 1), (1, 1, 2), (2, 1, 1), (1, 2, 1), (2, 2)

We can easily find recursive nature in above problem. The person can reach nth stair from either (n-1)th stair or from (n-2)th stair. Let the total
number of ways to reach nt stair be ways(n). The value of ways(n) can be written as following.
ways(n) = ways(n-1) + ways(n-2)

The above expression is actually the expression for Fibonacci numbers, but there is one thing to notice, the value of ways(n) is equal to
fibonacci(n+1).
ways(1) = fib(2) = 1
ways(2) = fib(3) = 2
ways(3) = fib(4) = 3
So we can use function for fibonacci numbers to find the value of ways(n). Following is C++ implementation of the above idea.
// A C program to count number of ways to reach n't stair when
// a person can climb 1, 2, ..m stairs at a time.
#include<stdio.h>
// A simple recursive program to find n'th fibonacci number
int fib(int n)
{
if (n <= 1)
return n;
return fib(n-1) + fib(n-2);
}
// Returns number of ways to reach s'th stair
int countWays(int s)
{
return fib(s + 1);
}
// Driver program to test above functions

int main ()
{
int s = 4;
printf("Number of ways = %d", countWays(s));
getchar();
return 0;
}

Output:
Number of ways = 5

The time complexity of the above implementation is exponential (golden ratio raised to power n). It can be optimized to work in O(Logn) time
using the previously discussed Fibonacci function optimizations.
Generalization of the above problem
How to count number of ways if the person can climb up to m stairs for a given value m? For example if m is 4, the person can climb 1 stair or 2
stairs or 3 stairs or 4 stairs at a time.
We can write the recurrence as following.
ways(n, m) = ways(n-1, m) + ways(n-2, m) + ... ways(n-m, m)

Following is C++ implementation of above recurrence.

// A C program to count number of ways to reach n't stair when
// a person can climb either 1 or 2 stairs at a time
#include<stdio.h>
// A recursive function used by countWays
int countWaysUtil(int n, int m)
{
if (n <= 1)
return n;
int res = 0;
for (int i = 1; i<=m && i<=n; i++)
res += countWaysUtil(n-i, m);
return res;
}
// Returns number of ways to reach s'th stair
int countWays(int s, int m)
{
return countWaysUtil(s+1, m);
}
// Driver program to test above functions
int main ()
{
int s = 4, m = 2;
printf("Nuber of ways = %d", countWays(s, m));
return 0;
}

Output:
Number of ways = 5

The time complexity of above solution is exponential. It can be optimized to O(mn) by using dynamic programming. Following is dynamic
programming based solution. We build a table res[] in bottom up manner.
// A C program to count number of ways to reach n't stair when
// a person can climb 1, 2, ..m stairs at a time
#include<stdio.h>
// A recursive function used by countWays
int countWaysUtil(int n, int m)
{
int res[n];
res[0] = 1; res[1] = 1;
for (int i=2; i<n; i++)
{
res[i] = 0;
for (int j=1; j<=m && j<=i; j++)
res[i] += res[i-j];
}
return res[n-1];
}

// Returns number of ways to reach s'th stair

int countWays(int s, int m)
{
return countWaysUtil(s+1, m);
}
// Driver program to test above functions
int main ()
{
int s = 4, m = 2;
printf("Nuber of ways = %d", countWays(s, m));
return 0;
}

Output:
Number of ways = 5

Replace all 0 with 5 in an input Integer

Given a integer as a input and replace all the 0 with 5 in the integer.
Examples:
102 - 152
1020 - 1525

Use of array to store all digits is not allowed.

Source: Amazon interview Experience | Set 136 (For SDE-T)
Following is C implementation of the above idea.
// C program to replace all 0 with 5 in an input Integer
#include<stdio.h>
// A recursive function to replace all 0s with 5s in an input number
// It doesn't work if input number itself is 0.
int convert0To5Rec(int num)
{
// Base case for recursion termination
if (num == 0)
return 0;
// Extraxt the last digit and change it if needed
int digit = num % 10;
if (digit == 0)
digit = 5;
// Convert remaining digits and append the last digit
return convert0To5Rec(num/10) * 10 + digit;
}
// It handles 0 and calls convert0To5Rec() for other numbers
int convert0To5(int num)
{
if (num == 0)
return 5;
else return convert0To5Rec(num);
}
// Driver program to test above function
int main()
{
int num = 10120;
printf("%d", convert0To5(num));
return 0;
}

Output:
15125

Program to add two polynomials

Given two polynomials represented by two arrays, write a function that adds given two polynomials.
Example:
Input: A[] = {5, 0, 10, 6}
B[] = {1, 2, 4}
Output: sum[] = {5, 10, 30, 26, 52, 24}
The first input array represents "5 + 0x^1 + 10x^2 + 6x^3"
The second array represents "1 + 2x^1 + 4x^2"
And Output is "6 + 2x^1 + 14x^2 + 6x^3"
add(A[0..m-1], B[0..n01])
1) Create a sum array sum[] of size equal to maximum of 'm' and 'n'
2) Copy A[] to sum[].
3) Travers array B[] and do following for every element B[i]
sum[i] = sum[i] + B[i]
4) Return sum[].

The following is C++ implementation of above algorithm.

// Simple C++ program to add two polynomials
#include <iostream>
using namespace std;
// A utility function to return maximum of two integers
int max(int m, int n) { return (m > n)? m: n; }
// A[] represents coefficients of first polynomial
// B[] represents coefficients of second polynomial
// m and n are sizes of A[] and B[] respectively
int *add(int A[], int B[], int m, int n)
{
int size = max(m, n);
int *sum = new int[size];
// Initialize the porduct polynomial
for (int i = 0; i<m; i++)
sum[i] = A[i];
// Take ever term of first polynomial
for (int i=0; i<n; i++)
sum[i] += B[i];
return sum;
}
// A utility function to print a polynomial
void printPoly(int poly[], int n)
{
for (int i=0; i<n; i++)
{
cout << poly[i];
if (i != 0)
cout << "x^" << i ;
if (i != n-1)
cout << " + ";
}
}
// Driver program to test above functions
int main()
{
// The following array represents polynomial 5 + 10x^2 + 6x^3
int A[] = {5, 0, 10, 6};
// The following array represents polynomial 1 + 2x + 4x^2
int B[] = {1, 2, 4};
int m = sizeof(A)/sizeof(A[0]);
int n = sizeof(B)/sizeof(B[0]);
cout << "First polynomial is \n";
printPoly(A, m);
cout << "\nSecond polynomial is \n";
printPoly(B, n);
int *sum = add(A, B, m, n);
int size = max(m, n);

cout << "\nsumuct polynomial is \n";

printPoly(sum, size);
return 0;
}

Output:
First polynomial is
5 + 0x^1 + 10x^2 + 6x^3
Second polynomial is
1 + 2x^1 + 4x^2
Sum polynomial is
6 + 2x^1 + 14x^2 + 6x^3

Time complexity of the above algorithm and program is O(m+n) where m and n are orders of two given polynomials.

Print first k digits of 1/n where n is a positive integer

Given a positive integer n, print first k digits after point in value of 1/n. Your program should avoid overflow and floating point arithmetic.
Examples:
Input: n = 3, k = 3
Output: 333
Input: n = 50, k = 4
Output: 0200

Let us consider an example n = 7, k = 3. The first digit of 1/7 is 1, it can be obtained by doing integer value of 10/7. Remainder of 10/7 is 3. Next
digit is 4 which can be obtained by taking integer value of 30/7. Remainder of 30/7 is 2. Next digits is 2 which can be obtained by taking integer
value of 20/7
#include <iostream>
using namespace std;
// Function to print first k digits after dot in value
// of 1/n. n is assumed to be a positive integer.
void print(int n, int k)
{
int rem = 1; // Initialize remainder
// Run a loop k times to print k digits
for (int i = 0; i < k; i++)
{
// The next digit can always be obtained as
// doing (10*rem)/10
cout << (10 * rem) / n;
// Update remainder
rem = (10*rem) % n;
}
}
// Driver program to test above function
int main()
{
int n = 7, k = 3;
print(n, k);
cout << endl;
n = 21, k = 4;
print(n, k);
return 0;
}

Output:
142
0476

Reference:
Algorithms And Programming: Problems And Solutions by Alexander Shen

Given a number as a string, find the number of contiguous subsequences which

recursively add up to 9
Given a number as a string, write a function to find the number of substrings (or contiguous subsequences) of the given string which recursively add
up to 9.
For example digits of 729 recursively add to 9,
7 + 2 + 9 = 18
Recur for 18
1+8=9
Examples:
Input: 4189
Output: 3
There are three substrings which recursively add to 9.
The substrings are 18, 9 and 189.
Input: 999
Output: 6
There are 6 substrings which recursively add to 9.
9, 99, 999, 9, 99, 9

All digits of a number recursively add up to 9, if only if the number is multiple of 9. We basically need to check for s%9 for all substrings s. One
trick used in below program is to do modular arithmetic to avoid overflow for big strings.
Following is a simple implementation based on this approach. The implementation assumes that there are no leading 0s in input number.
// C++ program to count substrings with recursive sum equal to 9
#include <iostream>
#include <cstring>
using namespace std;
int count9s(char number[])
{
int count = 0; // To store result
int n = strlen(number);
// Consider every character as beginning of substring
for (int i = 0; i < n; i++)
{
int sum = number[i] - '0'; //sum of digits in current substring
if (number[i] == '9') count++;
// One by one choose every character as an ending character
for (int j = i+1; j < n; j++)
{
// Add current digit to sum, if sum becomes multiple of 5
// then increment count. Let us do modular arithmetic to
// avoid overflow for big strings
sum = (sum + number[j] - '0')%9;
if (sum == 0)
count++;
}
}
return count;
}
// driver program to test above function
int main()
{
cout << count9s("4189") << endl;
cout << count9s("1809");
return 0;
}

Output:
3
5

Time complexity of the above program is O(n2). Please let me know if there is a better solution.

Find the element that appears once

Given an array where every element occurs three times, except one element which occurs only once. Find the element that occurs once. Expected
time complexity is O(n) and O(1) extra space.
Examples:
Input: arr[] = {12, 1, 12, 3, 12, 1, 1, 2, 3, 3}
Output: 2

We can use sorting to do it in O(nLogn) time. We can also use hashing, but the worst case time complexity of hashing may be more than O(n) and
hashing requires extra space.
The idea is to use bitwise operators for a solution that is O(n) time and uses O(1) extra space. The solution is not easy like other XOR based
solutions, because all elements appear odd number of times here. The idea is taken from here.
Run a loop for all elements in array. At the end of every iteration, maintain following two values.
ones: The bits that have appeared 1st time or 4th time or 7th time .. etc.
twos: The bits that have appeared 2nd time or 5th time or 8th time .. etc.
Finally, we return the value of ones
How to maintain the values of ones and twos?
ones and twos are initialized as 0. For every new element in array, find out the common set bits in the new element and previous value of ones.
These common set bits are actually the bits that should be added to twos. So do bitwise OR of the common set bits with twos. twos also gets
some extra bits that appear third time. These extra bits are removed later.
Update ones by doing XOR of new element with previous value of ones. There may be some bits which appear 3rd time. These extra bits are also
removed later.
Both ones and twos contain those extra bits which appear 3rd time. Remove these extra bits by finding out common set bits in ones and twos.
#include <stdio.h>
int getSingle(int arr[], int n)
{
int ones = 0, twos = 0 ;
int common_bit_mask;
// Let us take the example of {3, 3, 2, 3} to understand this
for( int i=0; i< n; i++ )
{
/* The expression "one & arr[i]" gives the bits that are
there in both 'ones' and new element from arr[]. We
add these bits to 'twos' using bitwise OR
Value of 'twos' will be set as 0, 3, 3 and 1 after 1st,
2nd, 3rd and 4th iterations respectively */
twos = twos | (ones & arr[i]);
/* XOR the new bits with previous 'ones' to get all bits
appearing odd number of times
Value of 'ones' will be set as 3, 0, 2 and 3 after 1st,
2nd, 3rd and 4th iterations respectively */
ones = ones ^ arr[i];
/* The common bits are those bits which appear third time
So these bits should not be there in both 'ones' and 'twos'.
common_bit_mask contains all these bits as 0, so that the bits can
be removed from 'ones' and 'twos'
Value of 'common_bit_mask' will be set as 00, 00, 01 and 10
after 1st, 2nd, 3rd and 4th iterations respectively */
common_bit_mask = ~(ones & twos);
/* Remove common bits (the bits that appear third time) from 'ones'
Value of 'ones' will be set as 3, 0, 0 and 2 after 1st,
2nd, 3rd and 4th iterations respectively */
ones &= common_bit_mask;

/* Remove common bits (the bits that appear third time) from 'twos'
Value of 'twos' will be set as 0, 3, 1 and 0 after 1st,
2nd, 3rd and 4th itearations respectively */
twos &= common_bit_mask;
// uncomment this code to see intermediate values
//printf (" %d %d \n", ones, twos);
}
return ones;
}
int main()
{
int arr[] = {3, 3, 2, 3};
int n = sizeof(arr) / sizeof(arr[0]);
printf("The element with single occurrence is %d ",
getSingle(arr, n));
return 0;
}

Output:
2

Time Complexity: O(n)

Auxiliary Space: O(1)
Following is another O(n) time complexity and O(1) extra space method suggested by aj. We can sum the bits in same positions for all the
numbers and take modulo with 3. The bits for which sum is not multiple of 3, are the bits of number with single occurrence.
Let us consider the example array {5, 5, 5, 8}. The 101, 101, 101, 1000
Sum of first bits%3 = (1 + 1 + 1 + 0)%3 = 0;
Sum of second bits%3 = (0 + 0 + 0 + 0)%0 = 0;
Sum of third bits%3 = (1 + 1 + 1 + 0)%3 = 0;
Sum of fourth bits%3 = (1)%3 = 1;
Hence number which appears once is 1000
#include <stdio.h>
#define INT_SIZE 32
int getSingle(int arr[], int n)
{
// Initialize result
int result = 0;
int x, sum;
// Iterate through every bit
for (int i = 0; i < INT_SIZE; i++)
{
// Find sum of set bits at ith position in all
// array elements
sum = 0;
x = (1 << i);
for (int j=0; j< n; j++ )
{
if (arr[j] & x)
sum++;
}
// The bits with sum not multiple of 3, are the
// bits of element with single occurrence.
if (sum % 3)
result |= x;
}
return result;
}
// Driver program to test above function
int main()
{
int arr[] = {12, 1, 12, 3, 12, 1, 1, 2, 3, 2, 2, 3, 7};
int n = sizeof(arr) / sizeof(arr[0]);
printf("The element with single occurrence is %d ",
getSingle(arr, n));
return 0;

}
7

Detect if two integers have opposite signs

Given two signed integers, write a function that returns true if the signs of given integers are different, otherwise false. For example, the function
should return true -1 and +100, and should return false for -100 and -200. The function should not use any of the arithmetic operators.
Let the given integers be x and y. The sign bit is 1 in negative numbers, and 0 in positive numbers. The XOR of x and y will have the sign bit as 1 iff
they have opposite sign. In other words, XOR of x and y will be negative number number iff x and y have opposite signs. The following code use
this logic.
#include<stdbool.h>
#include<stdio.h>
bool oppositeSigns(int x, int y)
{
return ((x ^ y) < 0);
}
int main()
{
int x = 100, y = -100;
if (oppositeSigns(x, y) == true)
printf ("Signs are opposite");
else
printf ("Signs are not opposite");
return 0;
}

Output:
Signs are opposite

Source: Detect if two integers have opposite signs

We can also solve this by using two comparison operators. See the following code.
bool oppositeSigns(int x, int y)
{
return (x < 0)? (y >= 0): (y < 0);
}

The first method is more efficient. The first method uses a bitwise XOR and a comparison operator. The second method uses two comparison
operators and a bitwise XOR operation is more efficient compared to a comparison operation.
We can also use following method. It doesnt use any comparison operator. The method is suggested by Hongliang and improved by gaurav.
bool oppositeSigns(int x, int y)
{
return ((x ^ y) >> 31);
}

The function is written only for compilers where size of an integer is 32 bit. The expression basically checks sign of (x^y) using bitwise operator
>>. As mentioned above, the sign bit for negative numbers is always 1. The sign bit is the leftmost bit in binary representation. So we need to
checks whether the 32th bit (or leftmost bit) of x^y is 1 or not. We do it by right shifting the value of x^y by 31, so that the sign bit becomes the
least significant bit. If sign bit is 1, then the value of (x^y)>>31 will be 1, otherwise 0.

Count total set bits in all numbers from 1 to n

Given a positive integer n, count the total number of set bits in binary representation of all numbers from 1 to n.
Examples:
Input: n = 3
Output: 4
Input: n = 6
Output: 9
Input: n = 7
Output: 12
Input: n = 8
Output: 13

Source: Amazon Interview Question

Method 1 (Simple)
A simple solution is to run a loop from 1 to n and sum the count of set bits in all numbers from 1 to n.
// A simple program to count set bits in all numbers from 1 to n.
#include <stdio.h>
// A utility function to count set bits in a number x
unsigned int countSetBitsUtil(unsigned int x);
// Returns count of set bits present in all numbers from 1 to n
unsigned int countSetBits(unsigned int n)
{
int bitCount = 0; // initialize the result
for(int i = 1; i <= n; i++)
bitCount += countSetBitsUtil(i);
return bitCount;
}
// A utility function to count set bits in a number x
unsigned int countSetBitsUtil(unsigned int x)
{
if (x <= 0)
return 0;
return (x %2 == 0? 0: 1) + countSetBitsUtil (x/2);
}
// Driver program to test above functions
int main()
{
int n = 4;
printf ("Total set bit count is %d", countSetBits(n));
return 0;
}

Output:
Total set bit count is 6

Time Complexity: O(nLogn)

Method 2 (Tricky)
If the input number is of the form 2^b -1 e.g., 1,3,7,15.. etc, the number of set bits is b * 2^(b-1). This is because for all the numbers 0 to (2^b)1, if you complement and flip the list you end up with the same list (half the bits are on, half off).
If the number does not have all set bits, then some position m is the position of leftmost set bit. The number of set bits in that position is n (1 << m)
+ 1. The remaining set bits are in two parts: 1) The bits in the (m-1) positions down to the point where the leftmost bit becomes 0, and 2) The
2^(m-1) numbers below that point, which is the closed form above. An easy way to look at it is to consider the number 6:
0|0
0|0
0|1
0|1
-|

0
1
0
1

1|0 0
1|0 1
1|1 0

The leftmost set bit is in position 2 (positions are considered starting from 0). If we mask that off what remains is 2 (the 1 0? in the right part of the
last row.) So the number of bits in the 2nd position (the lower left box) is 3 (that is, 2 + 1). The set bits from 0-3 (the upper right box above) is
2*2^(2-1) = 4. The box in the lower right is the remaining bits we havent yet counted, and is the number of set bits for all the numbers up to 2 (the
value of the last entry in the lower right box) which can be figured recursively.
// A O(Logn) complexity program to count set bits in all numbers from 1 to n
#include <stdio.h>
/* Returns position of leftmost set bit. The rightmost
position is considered as 0 */
unsigned int getLeftmostBit (int n)
{
int m = 0;
while (n > 1)
{
n = n >> 1;
m++;
}
return m;
}
/* Given the position of previous leftmost set bit in n (or an upper
bound on leftmost position) returns the new position of leftmost
set bit in n */
unsigned int getNextLeftmostBit (int n, int m)
{
unsigned int temp = 1 << m;
while (n < temp)
{
temp = temp >> 1;
m--;
}
return m;
}
// The main recursive function used by countSetBits()
unsigned int _countSetBits(unsigned int n, int m);
// Returns count of set bits present in all numbers from 1 to n
unsigned int countSetBits(unsigned int n)
{
// Get the position of leftmost set bit in n. This will be
// used as an upper bound for next set bit function
int m = getLeftmostBit (n);
// Use the position
return _countSetBits (n, m);
}
unsigned int _countSetBits(unsigned int n, int m)
{
// Base Case: if n is 0, then set bit count is 0
if (n == 0)
return 0;
/* get position of next leftmost set bit */
m = getNextLeftmostBit(n, m);
//
//
//
if

If n is of the form 2^x-1, i.e., if n is like 1, 3, 7, 15, 31,.. etc,

then we are done.
Since positions are considered starting from 0, 1 is added to m
(n == ((unsigned int)1<<(m+1))-1)
return (unsigned int)(m+1)*(1<<m);

// update n for next recursive call

n = n - (1<<m);
return (n+1) + countSetBits(n) + m*(1<<(m-1));
}
// Driver program to test above functions
int main()
{
int n = 17;
printf ("Total set bit count is %d", countSetBits(n));
return 0;
}

Total set bit count is 35

Time Complexity: O(Logn). From the first look at the implementation, time complexity looks more. But if we take a closer look, statements inside
while loop of getNextLeftmostBit() are executed for all 0 bits in n. And the number of times recursion is executed is less than or equal to set bits in
n. In other words, if the control goes inside while loop of getNextLeftmostBit(), then it skips those many bits in recursion.
Thanks to agatsu and IC for suggesting this solution.
See this for another solution suggested by Piyush Kapoor.

Swap bits in a given number

Given a number x and two positions (from right side) in binary representation of x, write a function that swaps n bits at given two positions and
returns the result. It is also given that the two sets of bits do not overlap.
Examples:
Let p1 and p2 be the two given positions.
Example 1
Input:
x = 47 (00101111)
p1 = 1 (Start from second bit from right side)
p2 = 5 (Start from 6th bit from right side)
n = 3 (No of bits to be swapped)
Output:
227 (11100011)
The 3 bits starting from the second bit (from right side) are
swapped with 3 bits starting from 6th position (from right side)
Example 2
Input:
x = 28 (11100)
p1 = 0 (Start from first bit from right side)
p2 = 3 (Start from 4th bit from right side)
n = 2 (No of bits to be swapped)
Output:
7 (00111)
The 2 bits starting from 0th postion (from right side) are
swapped with 2 bits starting from 4th position (from right side)

Solution
We need to swap two sets of bits. XOR can be used in a similar way as it is used to swap 2 numbers. Following is the algorithm.
1) Move all bits of first set to rightmost side
set1 = (x >> p1) & ((1U << n) - 1)
Here the expression (1U << n) - 1 gives a number that
contains last n bits set and other bits as 0. We do &
with this expression so that bits other than the last
n bits become 0.
2) Move all bits of second set to rightmost side
set2 = (x >> p2) & ((1U << n) - 1)
3) XOR the two sets of bits
xor = (set1 ^ set2)
4) Put the xor bits back to their original positions.
xor = (xor << p1) | (xor << p2)
5) Finally, XOR the xor with original number so
that the two sets are swapped.
result = x ^ xor

Implementation:
#include<stdio.h>
int swapBits(unsigned int x, unsigned int p1, unsigned int p2, unsigned int n)
{
/* Move all bits of first set to rightmost side */
unsigned int set1 = (x >> p1) & ((1U << n) - 1);
/* Moce all bits of second set to rightmost side */
unsigned int set2 = (x >> p2) & ((1U << n) - 1);
/* XOR the two sets */
unsigned int xor = (set1 ^ set2);
/* Put the xor bits back to their original positions */
xor = (xor << p1) | (xor << p2);
/* XOR the 'xor' with the original number so that the
two sets are swapped */
unsigned int result = x ^ xor;
return result;
}
/* Drier program to test above function*/
int main()
{

int res = swapBits(28, 0, 3, 2);

printf("\nResult = %d ", res);
return 0;
}

Output:
Result = 7

Following is a shorter implementation of the same logic

int swapBits(unsigned int x, unsigned int p1, unsigned int p2, unsigned int n)
{
/* xor contains xor of two sets */
unsigned int xor = ((x >> p1) ^ (x >> p2)) & ((1U << n) - 1);
/* To swap two sets, we need to again XOR the xor with original sets */
return x ^ ((xor << p1) | (xor << p2));
}

References:
Swapping individual bits with XOR

Add two numbers without using arithmetic operators

Write a function Add() that returns sum of two integers. The function should not use any of the arithmetic operators (+, ++, , -, .. etc).
Sum of two bits can be obtained by performing XOR (^) of the two bits. Carry bit can be obtained by performing AND (&) of two bits.
Above is simple Half Adder logic that can be used to add 2 single bits. We can extend this logic for integers. If x and y dont have set bits at same
position(s), then bitwise XOR (^) of x and y gives the sum of x and y. To incorporate common set bits also, bitwise AND (&) is used. Bitwise
AND of x and y gives all carry bits. We calculate (x & y) << 1 and add it to x ^ y to get the required result.
#include<stdio.h>
int Add(int x, int y)
{
// Iterate till there is no carry
while (y != 0)
{
// carry now contains common set bits of x and y
int carry = x & y;
// Sum of bits of x and y where at least one of the bits is not set
x = x ^ y;
// Carry is shifted by one so that adding it to x gives the required sum
y = carry << 1;
}
return x;
}
int main()
{
printf("%d", Add(15, 32));
return 0;
}

Following is recursive implementation for the same approach.

int Add(int x, int y)
{
if (y == 0)
return x;
else
return Add( x ^ y, (x & y) << 1);
}

Smallest of three integers without comparison operators

Write a C program to find the smallest of three integers, without using any of the comparison operators.
Let 3 input numbers be x, y and z.
Method 1 (Repeated Subtraction)
Take a counter variable c and initialize it with 0. In a loop, repeatedly subtract x, y and z by 1 and increment c. The number which becomes 0 first
is the smallest. After the loop terminates, c will hold the minimum of 3.
#include<stdio.h>
int smallest(int x, int y, int z)
{
int c = 0;
while ( x && y && z )
{
x--; y--; z--; c++;
}
return c;
}
int main()
{
int x = 12, y = 15, z = 5;
printf("Minimum of 3 numbers is %d", smallest(x, y, z));
return 0;
}

This methid doesnt work for negative numbers. Method 2 works for negative nnumbers also.

Method 2 (Use Bit Operations)

Use method 2 of this post to find minimum of two numbers (We cant use Method 1 as Method 1 uses comparison operator). Once we have
functionality to find minimum of 2 numbers, we can use this to find minimum of 3 numbers.
// See mthod 2 of http://www.geeksforgeeks.org/archives/2643
#include<stdio.h>
#define CHAR_BIT 8
/*Function to find minimum of x and y*/
int min(int x, int y)
{
return y + ((x - y) & ((x - y) >>
(sizeof(int) * CHAR_BIT - 1)));
}
/* Function to find minimum of 3 numbers x, y and z*/
int smallest(int x, int y, int z)
{
return min(x, min(y, z));
}
int main()
{
int x = 12, y = 15, z = 5;
printf("Minimum of 3 numbers is %d", smallest(x, y, z));
return 0;
}

Method 3 (Use Division operator)

We can also use division operator to find minimum of two numbers. If value of (a/b) is zero, then b is greater than a, else a is greater. Thanks to
gopinath and Vignesh for suggesting this method.
#include <stdio.h>
// Using division operator to find minimum of three numbers
int smallest(int x, int y, int z)
{
if (!(y/x)) // Same as "if (y < x)"
return (!(y/z))? y : z;
return (!(x/z))? x : z;
}
int main()
{

int x = 78, y = 88, z = 68;

printf("Minimum of 3 numbers is %d", smallest(x, y, z));
return 0;
}

A Boolean Array Puzzle

Input: A array arr[] of two elements having value 0 and 1
Output: Make both elements 0.
Specifications: Following are the specifications to follow.
1) It is guaranteed that one element is 0 but we do not know its position.
2) We cant say about another element it can be 0 or 1.
3) We can only complement array elements, no other operation like and, or, multi, division, . etc.
4) We cant use if, else and loop constructs.
5) Obviously, we cant directly assign 0 to array elements.
There are several ways we can do it as we are sure that always one Zero is there. Thanks to devendraiiit for suggesting following 3 methods.
Method 1
void changeToZero(int a[2])
{
a[ a[1] ] = a[ !a[1] ];
}
int main()
{
int a[] = {1, 0};
changeToZero(a);
printf(" arr[0] = %d \n", a[0]);
printf(" arr[1] = %d ", a[1]);
getchar();
return 0;
}

Method 2
void changeToZero(int a[2])
{
a[ !a[0] ] = a[ !a[1] ]
}

Method 3
This method doesnt even need complement.
void changeToZero(int a[2])
{
a[ a[1] ] = a[ a[0] ]
}

Method 4
Thanks to purvi for suggesting this method.
void changeToZero(int a[2])
{
a[0] = a[a[0]];
a[1] = a[0];
}

There may be many more methods.

Source: http://geeksforgeeks.org/forum/topic/google-challenge

Program to count number of set bits in an (big) array

Given an integer array of length N (anarbitrarilylarge number). How to count number of set bits in the array?
The simple approach would be, create an efficient method to count set bits in a word (most prominent size, usually equal to bit length of
processor), and add bits from individual elements of array.
Various methods of counting set bits of an integer exists, see this for example. These methods run at best O(logN) where N is number of bits.
Note that on a processor N is fixed, countcan be done in O(1) time on 32 bit machine irrespective of total set bits. Overall, the bits in array can be
computed in O(n) time, where n is array size.
However, a table look up will be more efficient method when array size is large. Storing table look up that can handle 232 integers will be
impractical.
The following code illustrates simple program to count set bits in a randomly generated 64 K integer array. The idea is to generate a look up for
first 256 numbers (one byte), and break every element of array at byte boundary. A meta program using C/C++ preprocessor generates the look
up table for counting set bits in a byte.
Themathematicalderivation behind meta program is evident from the following table (Add the column and row indices to get the number, then look
into the table to get set bits in that number. For example, to get set bits in 10, it can be extracted from row named as 8 and column named as 2),
0, 1, 2, 3
0 - 0, 1, 1,
4 - 1, 2, 2,
8 - 1, 2, 2,
12 - 2, 3, 3,
16 - 1, 2, 2,
20 - 2, 3, 3,
24 - 2, 3, 3,
28 - 3, 4, 4,

2
3
3
4
3
4
4
5

---------------------------------------------------------

GROUP_A(0)
GROUP_A(1)
GROUP_A(1)
GROUP_A(2)
GROUP_A(1)
GROUP_A(2)
GROUP_A(2)
GROUP_A(3) ... so on

From the table, there is a patten emerging in multiples of 4, both in the table as well as in the groupparameter. The sequence can be generalized as
shown in the code.
Complexity:
All the operations takes O(1) except iterating over the array. The time complexity is O(n) where n is size of array. Space complexity depends on
the meta program that generates look up.
Code:
#include <stdio.h>
#include <stdlib.h>
#include <time.h>
/* Size of array 64 K */
#define SIZE (1 << 16)
/* Meta program that generates set bit count
array of first 256 integers */
/* GROUP_A - When combined with META_LOOK_UP
generates count for 4x4 elements */
#define GROUP_A(x) x, x + 1, x + 1, x + 2
/* GROUP_B - When combined with META_LOOK_UP
generates count for 4x4x4 elements */
#define GROUP_B(x) GROUP_A(x), GROUP_A(x+1), GROUP_A(x+1), GROUP_A(x+2)
/* GROUP_C - When combined with META_LOOK_UP
generates count for 4x4x4x4 elements */
#define GROUP_C(x) GROUP_B(x), GROUP_B(x+1), GROUP_B(x+1), GROUP_B(x+2)
/* Provide appropriate letter to generate the table */
#define META_LOOK_UP(PARAMETER) \
GROUP_##PARAMETER(0), \
GROUP_##PARAMETER(1), \
GROUP_##PARAMETER(1), \
GROUP_##PARAMETER(2) \
int countSetBits(int array[], size_t array_size)

{
int count = 0;
/* META_LOOK_UP(C) - generates a table of 256 integers whose
sequence will be number of bits in i-th position
where 0 <= i < 256
*/
/* A static table will be much faster to access */
static unsigned char const look_up[] = { META_LOOK_UP(C) };
/* No shifting funda (for better readability) */
unsigned char *pData = NULL;
for(size_t index = 0; index < array_size; index++)
{
/* It is fine, bypass the type system */
pData = (unsigned char *)&array[index];
/* Count
count +=
count +=
count +=
count +=

set bits in individual bytes */

look_up[pData[0]];
look_up[pData[1]];
look_up[pData[2]];
look_up[pData[3]];

}
return count;
}
/* Driver program, generates table of random 64 K numbers */
int main()
{
int index;
int random[SIZE];
/* Seed to the random-number generator */
srand((unsigned)time(0));
/* Generate random numbers. */
for( index = 0; index < SIZE; index++ )
{
random[index] = rand();
}
printf("Total number of bits = %d\n", countSetBits(random, SIZE));
return 0;
}

Next higher number with same number of set bits

Given a number x, find next number with same number of 1 bits in its binary representation.
For example, consider x = 12, whose binary representation is 1100 (excluding leading zeros on 32 bit machine). It contains two logic 1 bits. The
next higher number with two logic 1 bits is 17 (100012).
Algorithm:
When we observe the binary sequence from 0 to 2n 1 (n is # of bits), right most bits (least significant) vary rapidly than left most bits.The idea is to
find right most string of 1s in x, and shift the pattern to right extreme, except the left most bit in the pattern. Shift the left most bit in the pattern
(omitted bit) to left part of x by one position. An example makes it more clear,
x = 156
10

x = 10011100
(2)

10011100
00011100- right most string of 1's in x
00000011 - right shifted pattern except left most bit ------> [A]
00010000 - isolated left most bit of right most 1's pattern
00100000 - shiftleft-ed the isolated bit by one position ------> [B]
10000000 - left part of x, excluding right most 1's pattern ------> [C]
10100000 - add B and C (OR operation) ------> [D]
10100011 - add A and D which is required number 163
(10)

After practicing with few examples, it easy to understand. Use the below given program for generating more sets.
Program Design:
We need to note few facts of binary numbers.The expression x & -x will isolate right most set bit in x (ensuring x will use 2s complement form for
negative numbers). If we add the result to x, right most string of 1s in x will be reset, and the immediate 0 left to this pattern of 1s will be set, which
is part [B] of above explanation. For example if x = 156, x & -x will result in 00000100, adding this result to x yields 10100000 (see part D). We
left with the right shifting part of pattern of 1s (part A of above explanation).
There are different ways to achieve part A. Right shifting is essentiallya division operation. What should be our divisor? Clearly, it should be
multiple of 2 (avoids 0.5 error in right shifting), and it should shift the right most 1s pattern to right extreme. The expression (x & -x) will serve the
purpose of divisor. An EX-OR operation between the number X and expression which is used to reset right most bits, will isolate the rightmost 1s
pattern.
A Correction Factor:
Note that we are adding right most set bit to the bit pattern. The addition operation causes a shift in the bit positions. The weight of binary system is
2, one shift causes an increase by a factor of 2. Since the increased number (rightOnesPattern in the code) being used twice, the error
propagates twice. The error needs to be corrected. A right shift by 2 positions will correct the result.
The popular name for this program is same number of one bits.
#include<iostream>
using namespace std;
typedef unsigned int uint_t;
// this function returns next higher number with same number of set bits as x.
uint_t snoob(uint_t x)
{
uint_t rightOne;
uint_t nextHigherOneBit;
uint_t rightOnesPattern;
uint_t next = 0;
if(x)
{
// right most set bit
rightOne = x & -(signed)x;

// reset the pattern and set next higher bit

// left part of x will be here
nextHigherOneBit = x + rightOne;
// nextHigherOneBit is now part [D] of the above explanation.
// isolate the pattern
rightOnesPattern = x ^ nextHigherOneBit;
// right adjust pattern
rightOnesPattern = (rightOnesPattern)/rightOne;
// correction factor
rightOnesPattern >>= 2;
// rightOnesPattern is now part [A] of the above explanation.
// integrate new pattern (Add [D] and [A])
next = nextHigherOneBit | rightOnesPattern;
}
return next;
}
int main()
{
int x = 156;
cout<<"Next higher number with same number of set bits is "<<snoob(x);
getchar();
return 0;
}

Usage: Finding/Generating subsets.

Variations:
1. Write a program to find a number immediately smaller than given, with same number of logic 1 bits? (Pretty simple)
2. How to count or generate the subsets available in the given set?
References:
1. A nice presentationhere.
2. Hackers Delight by Warren (An excellent and short book on various bit magic algorithms, a must forenthusiasts)
3. C A Reference Manual by Harbison and Steele (A good book on standard C, you can access code part of this post here).

Optimization Techniques | Set 1 (Modulus)

Modulus operator is costly.
The modulus operator (%) in various languages is costly operation. Ultimately every operator/operation must result in processor instructions. Some
processors wont have modulus instruction at hardware level, in such case the compilers will insert stubs (predefined functions) to perform modulus.
It impacts performance.
There is simple technique to extract remainder when a number is divided by another number (divisor)that is power of 2? A number that is an exact
power of 2 will have only one bit set in its binary representation. Consider the following powers of 2 and their binary representations
2 10
4 100
8 1000
16 10000
Note those zeros in red color, they contribute to remainder in division operation. We can get mask for those zeros by decrementing
thedivisorby 1.
Generalizing the above pattern, a number that can be written in 2n form will have only one bit set followed by n zeros on the right side of 1. When
a number (N) divided by (2n), the bit positions corresponding to the above mentioned zeros will contribute to the remainder of division operation.
An example can make it clear,
N = 87 (1010111 binary form)
N%2 = N & (2-1) = 1010111 & 1 = 1 = 1
N%4 = N & (4-1) = 1010111 & 11 = 11 = 3
N%8 = N & (8-1) = 1010111 & 111 = 111 = 7
N%16 = N & (16-1) = 1010111 & 1111 = 111 = 7
N%32 = N & (32-1) = 1010111 & 11111 = 10111 = 23

Modulus operation over exact powers of 2 is simple and faster bitwise ANDing. This is the reason, programmers usually make buffer length as
powers of 2.
Note that the technique will work only for divisors that are powers of 2.
An Example:
Implementation of circular queue (ring buffer) using an array. Omitting one position in the circular bufferimplementationcan make it easy to
distinguish between full and empty conditions. When the buffer reaches SIZE-1, it needs to wrap back to initial position. The wrap back
operation can be simple AND operation if the buffer size is power of 2. If we use any other size, we would need to use modulus operation.
Note:
Per experts comments, premature optimization is an evil. The optimization techniques provided are to fine tune your code
afterfinalizingdesignstrategy, algorithm, data structures and implementation. Werecommendto avoid them at the start of code development. Code
readability is key formaintenance.

Add 1 to a given number

Write a program to add one to a given number. You are not allowed to use operators like +, -, *, /, ++, etc.
Examples:
Input: 12
Output: 13
Input: 6
Output: 7
Yes, you guessed it right, we can use bitwise operators to achieve this. Following are different methods to achieve same using bitwise operators.
Method 1
To add 1 to a number x (say 0011000111), we need to flip all the bits after the rightmost 0 bit (we get 0011000000). Finally, flip the rightmost 0
bit also (we get 0011001000) and we are done.
#include<stdio.h>
int addOne(int x)
{
int m = 1;
/* Flip all the set bits until we find a 0 */
while( x & m )
{
x = x^m;
m <<= 1;
}
/* flip the rightmost 0 bit */
x = x^m;
return x;
}
/* Driver program to test above functions*/
int main()
{
printf("%d", addOne(13));
getchar();
return 0;
}

Method 2
We know that the negative number is represented in 2s complement form on most of the architectures. We have the following lemma hold for 2s
complement representation of signed numbers.
Say, x is numerical value of a number, then
~x = -(x+1) [ ~ is for bitwise complement ]
(x + 1) is due to addition of 1 in 2s complement conversion
To get (x + 1) apply negation once again. So, the final expression becomes (-(~x)).
int addOne(int x)
{
return (-(~x));
}
/* Driver program to test above functions*/
int main()
{
printf("%d", addOne(13));
getchar();
return 0;
}

Example, assume the machine word length is one *nibble* for simplicity.
And x = 2 (0010),
~x = ~2 = 1101 (13 numerical)
-~x = -1101
Interpreting bits 1101 in 2s complement form yields numerical value as -(2^4 13) = -3. Applying - on the result leaves 3. Same analogy holds for
decrement. See this comment for implementation of decrement.

Note that this method works only if the numbers are stored in 2s complement form.
Thanks to Venki for suggesting this method.

Multiply a given Integer with 3.5

Given a integer x, write a function that multiplies x with 3.5 and returns the integer result. You are not allowed to use %, /, *.
Examples:
Input: 2
Output: 7
Input: 5
Output: 17 (Ignore the digits after decimal point)
Solution:
1. We can get x*3.5 by adding 2*x, x and x/2. To calculate 2*x, left shift x by 1 and to calculate x/2, right shift x by 2.
#include <stdio.h>
int multiplyWith3Point5(int x)
{
return (x<<1) + x + (x>>1);
}
/* Driver program to test above functions*/
int main()
{
int x = 4;
printf("%d", multiplyWith3Point5(x));
getchar();
return 0;
}

2. Another way of doing this could be (8*x x)/2 (See below code). Thanks to ajaym for suggesting this.
#include <stdio.h>
int multiplyWith3Point5(int x)
{
return ((x<<3) - x)>>1;
}

Turn off the rightmost set bit

Write a C function that unsets the rightmost set bit of an integer.
Examples:
Input: 12 (00...01100)
Output: 8 (00...01000)
Input: 7 (00...00111)
Output: 6 (00...00110)

Let the input number be n. n-1 would have all the bits flipped after the rightmost set bit (including the set bit). So, doing n&(n-1) would give us the
required result.
#include<stdio.h>
/* unsets the rightmost set bit of n and returns the result */
int fun(unsigned int n)
{
return n&(n-1);
}
/* Driver program to test above function */
int main()
{
int n = 7;
printf("The number after unsetting the rightmost set bit %d", fun(n));
getchar();
return 0;
}

Find whether a given number is a power of 4 or not

Asked by Ajay
1. A simple method is to take log of the given number on base 4, and if we get an integer then number is power of 4.
2. Another solution is to keep dividing the number by 4, i.e, do n = n/4 iteratively. In any iteration, if n%4 becomes non-zero and n is not 1 then n
is not a power of 4, otherwise n is a power of 4.
#include<stdio.h>
#define bool int
/* Function to check if x is power of 4*/
bool isPowerOfFour(int n)
{
if(n == 0)
return 0;
while(n != 1)
{
if(n%4 != 0)
return 0;
n = n/4;
}
return 1;
}
/*Driver program to test above function*/
int main()
{
int test_no = 64;
if(isPowerOfFour(test_no))
printf("%d is a power of 4", test_no);
else
printf("%d is not a power of 4", test_no);
getchar();
}

3. A number n is a power of 4 if following conditions are met.

a) There is only one bit set in the binary representation of n (or n is a power of 2)
b) The count of zero bits before the (only) set bit is even.
For example: 16 (10000) is power of 4 because there is only one bit set and count of 0s before the set bit is 4 which is even.
Thanks to Geek4u for suggesting the approach and providing the code.
#include<stdio.h>
#define bool int
bool isPowerOfFour(unsigned int n)
{
int count = 0;
/*Check if there is only one bit set in n*/
if ( n && !(n&(n-1)) )
{
/* count 0 bits before set bit */
while(n > 1)
{
n >>= 1;
count += 1;
}
/*If count is even then return true else false*/
return (count%2 == 0)? 1 :0;
}
/* If there are more than 1 bit set
then n is not a power of 4*/
return 0;
}
/*Driver program to test above function*/
int main()
{
int test_no = 64;
if(isPowerOfFour(test_no))
printf("%d is a power of 4", test_no);
else

printf("%d is not a power of 4", test_no);

getchar();
}

Compute the integer absolute value (abs) without branching

We need not to do anything if a number is positive. We want to change only negative numbers. Since negative numbers are stored in 2s
complement form, to get the absolute value of a negative number we have to toggle bits of the number and add 1 to the result.
For example -2 in a 8 bit system is stored as follows 1 1 1 1 1 1 1 0 where leftmost bit is the sign bit. To get the absolute value of a negative
number, we have to toggle all bits and add 1 to the toggled number i.e, 0 0 0 0 0 0 0 1 + 1 will give the absolute value of 1 1 1 1 1 1 1 0. Also
remember, we need to do these operations only if the number is negative (sign bit is set).
Method 1
1) Set the mask as right shift of integer by 31 (assuming integers are stored using 32 bits).
mask = n>>31

2) For negative numbers, above step sets mask as 1 1 1 1 1 1 1 1 and 0 0 0 0 0 0 0 0 for positive numbers. Add the mask to the given number.
mask + n

3) XOR of mask +n and mask gives the absolute value.

(mask + n)^mask

Implementation:
#include <stdio.h>
#define CHAR_BIT 8
/* This function will return absoulte value of n*/
unsigned int getAbs(int n)
{
int const mask = n >> (sizeof(int) * CHAR_BIT - 1);
return ((n + mask) ^ mask);
}
/* Driver program to test above function */
int main()
{
int n = -6;
printf("Absoute value of %d is %u", n, getAbs(n));
getchar();
return 0;
}

Method 2:
1) Set the mask as right shift of integer by 31 (assuming integers are stored using 32 bits).
mask = n>>31

2) XOR the mask with number

mask ^ n

3) Subtract mask from result of step 2 and return the result.

(mask^n) - mask

Implementation:
/* This function will return absoulte value of n*/
unsigned int getAbs(int n)
{
int const mask = n >> (sizeof(int) * CHAR_BIT - 1);
return ((n ^ mask) - mask);
}

On machines where branching is expensive, the above expression can be faster than the obvious approach, r = (v < 0) ? -(unsigned)v : v, even
though the number of operations is the same. Please see this for more details about the above two methods.
References:
http://graphics.stanford.edu/~seander/bithacks.html#IntegerAbs

Compute modulus division by a power-of-2-number

Compute n modulo d without division(/) and modulo(%) operators, where d is a power of 2 number.
Let ith bit from right is set in d. For getting n modulus d, we just need to return 0 to i-1 (from right) bits of n as they are and other bits as 0.
For example if n = 6 (00..110) and d = 4(00..100). Last set bit in d is at position 3 (from right side). So we need to return last two bits of n as
they are and other bits as 0, i.e., 00..010.
Now doing it is so easy, guess it.
Yes, you have guessing it right. See the below program.
#include<stdio.h>
/* This function will return n % d.
d must be one of: 1, 2, 4, 8, 16, 32, */
unsigned int getModulo(unsigned int n, unsigned int d)
{
return ( n & (d-1) );
}
/* Driver program to test above function */
int main()
{
unsigned int n = 6;
unsigned int d = 4; /*d must be a power of 2*/
printf("%u moduo %u is %u", n, d, getModulo(n, d));
getchar();
return 0;
}

References:
http://graphics.stanford.edu/~seander/bithacks.html#ModulusDivisionEasy

Compute the minimum or maximum of two integers without branching

On some rare machines where branching is expensive, the below obvious approach to find minimum can be slow as it uses branching.
/* The obvious approach to find minimum (involves branching) */
int min(int x, int y)
{
return (x < y) ? x : y
}

Below are the methods to get minimum(or maximum) without using branching. Typically, the obvious approach is best, though.
Method 1(Use XOR and comparison operator)
Minimum of x and y will be
y ^ ((x ^ y) & -(x < y))

It works because if x < y, then -(x < y) will be all ones, so r = y ^ (x ^ y) & ~0 = y ^ x ^ y = x. Otherwise, if x >= y, then -(x < y) will be all
zeros, so r = y ^ ((x ^ y) & 0) = y. On some machines, evaluating (x < y) as 0 or 1 requires a branch instruction, so there may be no advantage.
To find the maximum, use
x ^ ((x ^ y) & -(x < y));
#include<stdio.h>
/*Function to find minimum of x and y*/
int min(int x, int y)
{
return y ^ ((x ^ y) & -(x < y));
}
/*Function to find maximum of x and y*/
int max(int x, int y)
{
return x ^ ((x ^ y) & -(x < y));
}
/* Driver program to test above functions */
int main()
{
int x = 15;
int y = 6;
printf("Minimum of %d and %d is ", x, y);
printf("%d", min(x, y));
printf("\nMaximum of %d and %d is ", x, y);
printf("%d", max(x, y));
getchar();
}

Method 2(Use subtraction and shift)

If we know that
INT_MIN <= (x - y) <= INT_MAX

, then we can use the following, which are faster because (x - y) only needs to be evaluated once.
Minimum of x and y will be
y + ((x - y) & ((x - y) >>(sizeof(int) * CHAR_BIT - 1)))

This method shifts the subtraction of x and y by 31 (if size of integer is 32). If (x-y) is smaller than 0, then (x -y)>>31 will be 1. If (x-y) is greater
than or equal to 0, then (x -y)>>31 will be 0.
So if x >= y, we get minimum as y + (x-y)&0 which is y.
If x < y, we get minimum as y + (x-y)&1 which is x. Similarly, to find the maximum use
x - ((x - y) & ((x - y) >> (sizeof(int) * CHAR_BIT - 1)))
#include<stdio.h>
#define CHAR_BIT 8
/*Function to find minimum of x and y*/
int min(int x, int y)
{
return y + ((x - y) & ((x - y) >>
(sizeof(int) * CHAR_BIT - 1)));
}

/*Function to find maximum of x and y*/

int max(int x, int y)
{
return x - ((x - y) & ((x - y) >>
(sizeof(int) * CHAR_BIT - 1)));
}
/* Driver program to test above functions */
int main()
{
int x = 15;
int y = 6;
printf("Minimum of %d and %d is ", x, y);
printf("%d", min(x, y));
printf("\nMaximum of %d and %d is ", x, y);
printf("%d", max(x, y));
getchar();
}

Note that the 1989 ANSI C specification doesn't specify the result of signed right-shift, so above method is not portable. If exceptions are thrown
on overflows, then the values of x and y should be unsigned or cast to unsigned for the subtractions to avoid unnecessarily throwing an exception,
however the right-shift needs a signed operand to produce all one bits when negative, so cast to signed there.
Source:
http://graphics.stanford.edu/~seander/bithacks.html#IntegerMinOrMax

Rotate bits of a number

Bit Rotation: A rotation (or circular shift) is an operation similar to shift except that the bits that fall off at one end are put back to the other end.
In left rotation, the bits that fall off at left end are put back at right end.
In right rotation, the bits that fall off at right end are put back at left end.
Example:
Let n is stored using 8 bits. Left rotation of n = 11100101 by 3 makes n = 00101111 (Left shifted by 3 and first 3 bits are put back in last ). If n is
stored using 16 bits or 32 bits then left rotation of n (00011100101) becomes 00..0011100101000.
Right rotation of n = 11100101 by 3 makes n = 10111100 (Right shifted by 3 and last 3 bits are put back in first ) if n is stored using 8 bits. If n is
stored using 16 bits or 32 bits then right rotation of n (00011100101) by 3 becomes 101000..0011100.
#include<stdio.h>
#define INT_BITS 32
/*Function to left rotate n by d bits*/
int leftRotate(int n, unsigned int d)
{
/* In n<<d, last d bits are 0. To put first 3 bits of n at
last, do bitwise or of n<<d with n >>(INT_BITS - d) */
return (n << d)|(n >> (INT_BITS - d));
}
/*Function to right rotate n by d bits*/
int rightRotate(int n, unsigned int d)
{
/* In n>>d, first d bits are 0. To put last 3 bits of at
first, do bitwise or of n>>d with n <<(INT_BITS - d) */
return (n >> d)|(n << (INT_BITS - d));
}
/* Driver program to test above functions */
int main()
{
int n = 16;
int d = 2;
printf("Left Rotation of %d by %d is ", n, d);
printf("%d", leftRotate(n, d));
printf("\nRight Rotation of %d by %d is ", n, d);
printf("%d", rightRotate(n, d));
getchar();
}

Find the two non-repeating elements in an array of repeating elements

Asked by SG
Given an array in which all numbers except two are repeated once. (i.e. we have 2n+2 numbers and n numbers are occurring twice and remaining
two have occurred once). Find those two numbers in the most efficient way.
Method 1(Use Sorting)
First sort all the elements. In the sorted array, by comparing adjacent elements we can easily get the non-repeating elements. Time complexity of
this method is O(nLogn)
Method 2(Use XOR)
Let x and y be the non-repeating elements we are looking for and arr[] be the input array. First calculate the XOR of all the array elements.
xor = arr[0]^arr[1]^arr[2].....arr[n-1]

All the bits that are set in xor will be set in one non-repeating element (x or y) and not in other. So if we take any set bit of xor and divide the
elements of the array in two sets one set of elements with same bit set and other set with same bit not set. By doing so, we will get x in one set and
y in another set. Now if we do XOR of all the elements in first set, we will get first non-repeating element, and by doing same in other set we will
get the second non-repeating element.
Let us see an example.
arr[] = {2, 4, 7, 9, 2, 4}
1) Get the XOR of all the elements.
xor = 2^4^7^9^2^4 = 14 (1110)
2) Get a number which has only one set bit of the xor.
Since we can easily get the rightmost set bit, let us use it.
set_bit_no = xor & ~(xor-1) = (1110) & ~(1101) = 0010
Now set_bit_no will have only set as rightmost set bit of xor.
3) Now divide the elements in two sets and do xor of
elements in each set, and we get the non-repeating
elements 7 and 9. Please see implementation for this
step.

Implementation:
#include <stdio.h>
#include <stdlib.h>
/* This finction sets the values of *x and *y to nonr-epeating
elements in an array arr[] of size n*/
void get2NonRepeatingNos(int arr[], int n, int *x, int *y)
{
int xor = arr[0]; /* Will hold xor of all elements */
int set_bit_no; /* Will have only single set bit of xor */
int i;
*x = 0;
*y = 0;
/* Get the xor of all elements */
for(i = 1; i < n; i++)
xor ^= arr[i];
/* Get the rightmost set bit in set_bit_no */
set_bit_no = xor & ~(xor-1);
/* Now divide elements in two sets by comparing rightmost set
bit of xor with bit at same position in each element. */
for(i = 0; i < n; i++)
{
if(arr[i] & set_bit_no)
*x = *x ^ arr[i]; /*XOR of first set */
else
*y = *y ^ arr[i]; /*XOR of second set*/
}
}
/* Driver program to test above function */
int main()
{
int arr[] = {2, 3, 7, 9, 11, 2, 3, 11};
int *x = (int *)malloc(sizeof(int));
int *y = (int *)malloc(sizeof(int));
get2NonRepeatingNos(arr, 8, x, y);
printf("The non-repeating elements are %d and %d", *x, *y);
getchar();
}

Time Complexity: O(n)

Auxiliary Space: O(1)

Find the Number Occurring Odd Number of Times

Given an array of positive integers. All numbers occur even number of times except one number which occurs odd number of times. Find the
number in O(n) time & constant space.
Example:
I/P = [1, 2, 3, 2, 3, 1, 3]
O/P = 3
A Simple Solution is to run two nested loops. The outer loop picks all elements one by one and inner loop counts number of occurrences of the
element picked by outer loop. Time complexity of this solution is O(n2).
A Better Solution is to use Hashing. Use array elements as key and their counts as value. Create an empty hash table. One by one traverse the
given array elements and store counts. Time complexity of this solution is O(n). But it requires extra space for hashing.
The Best Solution is to do bitwise XOR of all the elements. XOR of all elements gives us odd occurring element. Please note that XOR of two
elements is 0 if both elements are same and XOR of a number x with 0 is x.
Below are implementations of this best approach.
Program:

C/C++
#include <stdio.h>
int getOddOccurrence(int ar[], int ar_size)
{
int i;
int res = 0;
for (i=0; i < ar_size; i++)
res = res ^ ar[i];
return res;
}
/* Diver function to test above function */
int main()
{
int ar[] = {2, 3, 5, 4, 5, 2, 4, 3, 5, 2, 4, 4, 2};
int n = sizeof(ar)/sizeof(ar[0]);
printf("%d", getOddOccurrence(ar, n));
return 0;
}

Python
# Python program to find the element occurring odd number of times
def getOddOccurrence(arr):
# Initialize result
res = 0
# Traverse the array
for element in arr:
# XOR with the result
res = res ^ element
return res
# Test array
arr = [ 2, 3, 5, 4, 5, 2, 4, 3, 5, 2, 4, 4, 2]
print "%d" % getOddOccurrence(arr)

Output:
5

Time Complexity: O(n)

Check for Integer Overflow

Write a C function, int addOvf(int* result, int a, int b) If there is no overflow, the function places the resultant = sum a+b in result and returns 0.
Otherwise it returns -1. The solution of casting to long and adding to find detecting the overflow is not allowed.
Method 1
There can be overflow only if signs of two numbers are same, and sign of sum is opposite to the signs of numbers.
1) Calculate sum
2) If both numbers are positive and sum is negative then return -1
Else
If both numbers are negative and sum is positive then return -1
Else return 0
#include<stdio.h>
#include<stdlib.h>
/* Takes pointer to result and two numbers as
arguments. If there is no overflow, the function
places the resultant = sum a+b in result and
returns 0, otherwise it returns -1 */
int addOvf(int* result, int a, int b)
{
*result = a + b;
if(a > 0 && b > 0 && *result < 0)
return -1;
if(a < 0 && b < 0 && *result > 0)
return -1;
return 0;
}
int main()
{
int *res = (int *)malloc(sizeof(int));
int x = 2147483640;
int y = 10;
printf("%d", addOvf(res, x, y));
printf("\n %d", *res);
getchar();
return 0;
}

Time Complexity : O(1)

Space Complexity: O(1)

Method 2
Thanks to Himanshu Aggarwal for adding this method. This method doesnt modify *result if there us an overflow.
#include<stdio.h>
#include<limits.h>
#include<stdlib.h>
int addOvf(int* result, int a, int b)
{
if( a > INT_MAX - b)
return -1;
else
{
*result = a + b;
return 0;
}
}
int main()
{
int *res = (int *)malloc(sizeof(int));
int x = 2147483640;
int y = 10;
printf("%d", addOvf(res, x, y));
printf("\n %d", *res);
getchar();
return 0;
}

Time Complexity : O(1)

Space Complexity: O(1)

Little and Big Endian Mystery

What are these?
Little and big endian are two ways of storing multibyte data-types ( int, float, etc). In little endian machines, last byte of binary representation of the
multibyte data-type is stored first. On the other hand, in big endian machines, first byte of binary representation of the multibyte data-type is stored
first.
Suppose integer is stored as 4 bytes (For those who are using DOS based compilers such as C++ 3.0 , integer is 2 bytes) then a variable x with
value 0x01234567 will be stored as following.

Memory representation of integer ox01234567 inside Big and little endian machines
How to see memory representation of multibyte data types on your machine?
Here is a sample C code that shows the byte representation of int, float and pointer.
#include <stdio.h>
/* function to show bytes in memory, from location start to start+n*/
void show_mem_rep(char *start, int n)
{
int i;
for (i = 0; i < n; i++)
printf(" %.2x", start[i]);
printf("\n");
}
/*Main function to call above function for 0x01234567*/
int main()
{
int i = 0x01234567;
show_mem_rep((char *)&i, sizeof(i));
getchar();
return 0;
}

When above program is run on little endian machine, gives 67 45 23 01? as output , while if it is run on endian machine, gives 01 23 45 67? as
output.
Is there a quick way to determine endianness of your machine?
There are n no. of ways for determining endianness of your machine. Here is one quick way of doing the same.
#include <stdio.h>
int main()
{
unsigned int i = 1;
char *c = (char*)&i;
if (*c)
printf("Little endian");
else
printf("Big endian");
getchar();
return 0;
}

In the above program, a character pointer c is pointing to an integer i. Since size of character is 1 byte when the character pointer is de-referenced
it will contain only first byte of integer. If machine is little endian then *c will be 1 (because last byte is stored first) and if machine is big endian then
*c will be 0.
Does endianness matter for programmers?
Most of the times compiler takes care of endianness, however, endianness becomes an issue in following cases.
It matters in network programming: Suppose you write integers to file on a little endian machine and you transfer this file to a big endian machine.

Unless there is little andian to big endian transformation, big endian machine will read the file in reverse order. You can find such a practical
example here.
Standard byte order for networks is big endian, also known as network byte order. Before transferring data on network, data is first converted to
network byte order (big endian).
Sometimes it matters when you are using type casting, below program is an example.
#include <stdio.h>
int main()
{
unsigned char arr[2] = {0x01, 0x00};
unsigned short int x = *(unsigned short int *) arr;
printf("%d", x);
getchar();
return 0;
}

In the above program, a char array is typecasted to an unsigned short integer type. When I run above program on little endian machine, I get 1 as
output, while if I run it on a big endian machine I get 256. To make programs endianness independent, above programming style should be
avoided.
What are bi-endians?
Bi-endian processors can run in both modes little and big endian.
What are the examples of little, big endian and bi-endian machines ?
Intel based processors are little endians. ARM processors were little endians. Current generation ARM processors are bi-endian.
Motorola 68K processors are big endians. PowerPC (by Motorola) and SPARK (by Sun) processors were big endian. Current version of these
processors are bi-endians.
Does endianness effects file formats?
File formats which have 1 byte as a basic unit are independent of endianness e..g., ASCII files . Other file formats use some fixed endianness
forrmat e.g, JPEG files are stored in big endian format.
Which one is better little endian or big endian
The term little and big endian came from Gullivers Travels by Jonathan Swift. Two groups could not agree by which end a egg should be opened a-the little or the big. Just like the egg issue, there is no technological reason to choose one byte ordering convention over the other, hence the
arguments degenerate into bickering about sociopolitical issues. As long as one of the conventions is selected and adhered to consistently, the
choice is arbitrary.

Write an Efficient C Program to Reverse Bits of a Number

Method1 Simple
Loop through all the bits of an integer. If a bit at ith position is set in the i/p no. then set the bit at (NO_OF_BITS 1) i in o/p. Where
NO_OF_BITS is number of bits present in the given number.
/* Function to reverse bits of num */
unsigned int reverseBits(unsigned int num)
{
unsigned int NO_OF_BITS = sizeof(num) * 8;
unsigned int reverse_num = 0, i, temp;
for (i = 0; i < NO_OF_BITS; i++)
{
temp = (num & (1 << i));
if(temp)
reverse_num |= (1 << ((NO_OF_BITS - 1) - i));
}
return reverse_num;
}
/* Driver function to test above function */
int main()
{
unsigned int x = 2;
printf("%u", reverseBits(x));
getchar();
}

Above program can be optimized by removing the use of variable temp. See below the modified code.
unsigned int reverseBits(unsigned int num)
{
unsigned int NO_OF_BITS = sizeof(num) * 8;
unsigned int reverse_num = 0;
int i;
for (i = 0; i < NO_OF_BITS; i++)
{
if((num & (1 << i)))
reverse_num |= 1 << ((NO_OF_BITS - 1) - i);
}
return reverse_num;
}

Time Complexity: O(log n)

Space Complexity: O(1)
Method 2 Standard
The idea is to keep putting set bits of the num in reverse_num until num becomes zero. After num becomes zero, shift the remaining bits of
reverse_num.
Let num is stored using 8 bits and num be 00000110. After the loop you will get reverse_num as 00000011. Now you need to left shift
reverse_num 5 more times and you get the exact reverse 01100000.
unsigned int reverseBits(unsigned int num)
{
unsigned int count = sizeof(num) * 8 - 1;
unsigned int reverse_num = num;
num >>= 1;
while(num)
{
reverse_num <<= 1;
reverse_num |= num & 1;
num >>= 1;
count--;
}
reverse_num <<= count;
return reverse_num;
}
int main()
{
unsigned int x = 1;
printf("%u", reverseBits(x));
getchar();
}

Time Complexity: O(log n)

Space Complexity: O(1)
Method 3 Lookup Table:
We can reverse the bits of a number in O(1) if we know the size of the number. We can implement it using look up table. Go through the below
link for details. You will find some more interesting bit related stuff there.
http://www-graphics.stanford.edu/~seander/bithacks.html#BitReverseTable

Count set bits in an integer

Write an efficient program to count number of 1s in binary representation of an integer.
1. Simple Method Loop through all bits in an integer, check if a bit is set and if it is then increment the set bit count. See below program.
/* Function to get no of set bits in binary
representation of passed binary no. */
int countSetBits(unsigned int n)
{
unsigned int count = 0;
while(n)
{
count += n & 1;
n >>= 1;
}
return count;
}
/* Program to test function countSetBits */
int main()
{
int i = 9;
printf("%d", countSetBits(i));
getchar();
return 0;
}

Time Complexity: (-)(logn) (Theta of logn)

2. Brian Kernighans Algorithm:
Subtraction of 1 from a number toggles all the bits (from right to left) till the rightmost set bit(including the righmost set bit). So if we subtract a
number by 1 and do bitwise & with itself (n & (n-1)), we unset the righmost set bit. If we do n & (n-1) in a loop and count the no of times loop
executes we get the set bit count.
Beauty of the this solution is number of times it loops is equal to the number of set bits in a given integer.
1 Initialize count: = 0
2 If integer n is not zero
(a) Do bitwise & with (n-1) and assign the value back to n
n: = n&(n-1)
(b) Increment count by 1
(c) go to step 2
3 Else return count

Implementation of Brian Kernighans Algorithm:

#include<stdio.h>
/* Function to get no of set bits in binary
representation of passed binary no. */
int countSetBits(int n)
{
unsigned int count = 0;
while (n)
{
n &= (n-1) ;
count++;
}
return count;
}
/* Program to test function countSetBits */
int main()
{
int i = 9;
printf("%d", countSetBits(i));
getchar();
return 0;
}

Example for Brian Kernighans Algorithm:

n = 9 (1001)
count = 0
Since 9 > 0, subtract by 1 and do bitwise & with (9-1)

n = 9&8 (1001 & 1000)

n = 8
count = 1
Since 8 > 0, subtract by 1 and do bitwise & with (8-1)
n = 8&7 (1000 & 0111)
n = 0
count = 2
Since n = 0, return count which is 2 now.

Time Complexity: O(logn)

3. Using Lookup table: We can count bits in O(1) time using lookup table. Please see
http://graphics.stanford.edu/~seander/bithacks.html#CountBitsSetTable for details.
You can find one use of counting set bits at http://geeksforgeeks.org/?p=1465
References:
http://graphics.stanford.edu/~seander/bithacks.html#CountBitsSetNaive

Count number of bits to be flipped to convert A to B

Suggested by Dheeraj
Question: You are given two numbers A and B. Write a program to count number of bits needed to be flipped to convert A to B.
Solution:
1. Calculate XOR of A and B.
a_xor_b = A ^ B
2. Count the set bits in the above calculated XOR result.
countSetBits(a_xor_b)

XOR of two number will have set bits only at those places where A differs from B.
Example:
A = 1001001
B = 0010101
a_xor_b = 1011100
No of bits need to flipped = set bit count in a_xor_b i.e. 4

To get the set bit count please see another post on this portal http://geeksforgeeks.org/?p=1176

Next Power of 2
Write a function that, for a given no n, finds a number p which is greater than or equal to n and is a power of 2.
IP 5
OP 8
IP 17
OP 32
IP 32
OP 32

There are plenty of solutions for this. Let us take the example of 17 to explain some of them.
Method 1(Using Log of the number)
1. Calculate Position of set bit in p(next power of 2):
pos = ceil(lgn) (ceiling of log n with base 2)
2. Now calculate p:
p = pow(2, pos)

Example
Let us try for 17
pos = 5
p = 32

Method 2 (By getting the position of only set bit in result )

/* If n is a power of 2 then return n */
1 If (n & !(n&(n-1))) then return n
2 Else keep right shifting n until it becomes zero
and count no of shifts
a. Initialize: count = 0
b. While n ! = 0
n = n>>1
count = count + 1
/* Now count has the position of set bit in result */
3 Return (1 << count)

Example:
Let us try for 17
count = 5
p
= 32
unsigned int nextPowerOf2(unsigned int n)
{
unsigned count = 0;
/* First n in the below condition is for the case where n is 0*/
if (n && !(n&(n-1)))
return n;
while( n != 0)
{
n >>= 1;
count += 1;
}
return 1<<count;
}
/* Driver program to test above function */
int main()
{
unsigned int n = 0;
printf("%d", nextPowerOf2(n));
getchar();
return 0;
}

Method 3(Shift result one by one)

Thanks to coderyogi for suggesting this method . This method is a variation of method 2 where instead of getting count, we shift the result one by
one in a loop.
unsigned int nextPowerOf2(unsigned int n)
{
unsigned int p = 1;
if (n && !(n & (n - 1)))
return n;
while (p < n) {
p <<= 1;
}
return p;
}
/* Driver program to test above function */
int main()
{
unsigned int n = 5;
printf("%d", nextPowerOf2(n));
getchar();
return 0;
}

Time Complexity: O(lgn)

Method 4(Customized and Fast)
1. Subtract n by 1
n = n -1
2. Set all bits after the leftmost set bit.
/* Below solution
n = n
n = n
n = n
n = n
n = n
3. Return n + 1

works only if integer is 32 bits */

| (n >> 1);
| (n >> 2);
| (n >> 4);
| (n >> 8);
| (n >> 16);

Example:
Steps 1 & 3 of above algorithm are to handle cases
of power of 2 numbers e.g., 1, 2, 4, 8, 16,
Let us
step 1
n =
step 2
n =
n =
n =
n =
n =
n =
n =
n =
n =
n =
n =
n =
n =
n =
n =

try for 17(10001)

n - 1 = 16 (10000)
n | n
10000
11000
n | n
11000
11110
n | n
11110
11111
n | n
11111
11111
n | n
11110
11111

>> 1
| 01000
>> 2
| 00110
>> 4
| 00001
>> 8
| 00000
>> 16
| 00000

step 3: Return n+1

We get n + 1 as 100000 (32)

Program:
# include <stdio.h>
/* Finds next power of two for n. If n itself
is a power of two then returns n*/

unsigned int nextPowerOf2(unsigned int n)

{
n--;
n |= n >> 1;
n |= n >> 2;
n |= n >> 4;
n |= n >> 8;
n |= n >> 16;
n++;
return n;
}
/* Driver program to test above function */
int main()
{
unsigned int n = 5;
printf("%d", nextPowerOf2(n));
getchar();
return 0;
}

Time Complexity: O(lgn)

References:
http://en.wikipedia.org/wiki/Power_of_2

Write an Efficient Method to Check if a Number is Multiple of 3

The very first solution that comes to our mind is the one that we learned in school. If sum of digits in a number is multiple of 3 then number is
multiple of 3 e.g., for 612 sum of digits is 9 so its a multiple of 3. But this solution is not efficient. You have to get all decimal digits one by one, add
them and then check if sum is multiple of 3.
There is a pattern in binary representation of the number that can be used to find if number is a multiple of 3. If difference between count of odd set
bits (Bits set at odd positions) and even set bits is multiple of 3 then is the number.
Example: 23 (00..10111)
1) Get count of all set bits at odd positions (For 23 its 3).
2) Get count of all set bits at even positions (For 23 its 1).
3) If difference of above two counts is a multiple of 3 then number is also a multiple of 3.
(For 23 its 2 so 23 is not a multiple of 3)
Take some more examples like 21, 15, etc
Algorithm: isMutlipleOf3(n)
1) Make n positive if n is negative.
2) If number is 0 then return 1
3) If number is 1 then return 0
4) Initialize: odd_count = 0, even_count = 0
5) Loop while n != 0
a) If rightmost bit is set then increment odd count.
b) Right-shift n by 1 bit
c) If rightmost bit is set then increment even count.
d) Right-shift n by 1 bit
6) return isMutlipleOf3(odd_count - even_count)

Proof:
Above can be proved by taking the example of 11 in decimal numbers. (In this context 11 in decimal numbers is same as 3 in binary numbers)
If difference between sum of odd digits and even digits is multiple of 11 then decimal number is multiple of 11. Lets see how.
Lets take the example of 2 digit numbers in decimal
AB = 11A -A + B = 11A + (B A)
So if (B A) is a multiple of 11 then is AB.
Let us take 3 digit numbers.
ABC = 99A + A + 11B B + C = (99A + 11B) + (A + C B)
So if (A + C B) is a multiple of 11 then is (A+C-B)
Let us take 4 digit numbers now.
ABCD = 1001A + D + 11C C + 999B + B A
= (1001A 999B + 11C) + (D + B A -C )
So, if (B + D A C) is a multiple of 11 then is ABCD.
This can be continued for all decimal numbers.
Above concept can be proved for 3 in binary numbers in the same way.
Time Complexity: O(logn)
Program:
#include<stdio.h>
/* Fnction to check if n is a multiple of 3*/
int isMultipleOf3(int n)
{
int odd_count = 0;
int even_count = 0;
/* Make no positive if +n is multiple of 3
then is -n. We are doing this to avoid
stack overflow in recursion*/
if(n < 0) n = -n;
if(n == 0) return 1;
if(n == 1) return 0;
while(n)
{
/* If odd bit is set then

increment odd counter */

if(n & 1)
odd_count++;
n = n>>1;
/* If even bit is set then
increment even counter */
if(n & 1)
even_count++;
n = n>>1;
}
return isMultipleOf3(abs(odd_count - even_count));
}
/* Program to test function isMultipleOf3 */
int main()
{
int num = 23;
if (isMultipleOf3(num))
printf("num is multiple of 3");
else
printf("num is not a multiple of 3");
getchar();
return 0;
}

Write a C program to find the parity of an unsigned integer

Parity: Parity of a number refers to whether it contains an odd or even number of 1-bits. The number has odd parity, if it contains odd number of
1-bits and is even parity if it contains even number of 1-bits.
Main idea of the below solution is Loop while n is not 0 and in loop unset one of the set bits and invert parity.
Algorithm: getParity(n)
1. Initialize parity = 0
2. Loop while n != 0
a. Invert parity
parity = !parity
b. Unset rightmost set bit
n = n & (n-1)
3. return parity
Example:
Initialize: n = 13 (1101)

parity = 0

n = 13 & 12 = 12 (1100) parity = 1

n = 12 & 11 = 8 (1000) parity = 0
n = 8 & 7 = 0 (0000)
parity = 1

Program:
# include <stdio.h>
# define bool int
/* Function to get parity of number n. It returns 1
if n has odd parity, and returns 0 if n has even
parity */
bool getParity(unsigned int n)
{
bool parity = 0;
while (n)
{
parity = !parity;
n
= n & (n - 1);
}
return parity;
}
/* Driver program to test getParity() */
int main()
{
unsigned int n = 7;
printf("Parity of no %d = %s", n,
(getParity(n)? "odd": "even"));
getchar();
return 0;
}

Above solution can be optimized by using lookup table. Please refer to Bit Twiddle Hacks[1st reference] for details.
Time Complexity: The time taken by above algorithm is proportional to the number of bits set. Worst case complexity is O(Logn).
Uses: Parity is used in error detection and cryptography.
References:
http://graphics.stanford.edu/~seander/bithacks.html#ParityNaive last checked on 30 May 2009.

Efficient way to multiply with 7

We can multiply a number by 7 using bitwise operator. First left shift the number by 3 bits (you will get 8n) then subtract the original numberfrom
the shifted number and return the difference (8n n).
Program:
# include<stdio.h>
int multiplyBySeven(unsigned int n)
{
/* Note the inner bracket here. This is needed
because precedence of '-' operator is higher
than '<<' */
return ((n<<3) - n);
}
/* Driver program to test above function */
int main()
{
unsigned int n = 4;
printf("%u", multiplyBySeven(n));
getchar();
return 0;
}

Time Complexity: O(1)

Space Complexity: O(1)
Note: Works only for positive integers.
Same concept can be used for fast multiplication by 9 or other numbers.

Write one line C function to find whether a no is power of two

1. A simple method for this is to simply take the log of the number on base 2 and if you get an integer then number is power of 2.
2. Another solution is to keep dividing the number by two, i.e, do n = n/2 iteratively. In any iteration, if n%2 becomes non-zero and n is not 1 then
n is not a power of 2. If n becomes 1 then it is a power of 2.
#include<stdio.h>
#define bool int
/* Function to check if x is power of 2*/
bool isPowerOfTwo(int n)
{
if (n == 0)
return 0;
while (n != 1)
{
if (n%2 != 0)
return 0;
n = n/2;
}
return 1;
}
/*Driver program to test above function*/
int main()
{
isPowerOfTwo(31)? printf("Yes\n"): printf("No\n");
isPowerOfTwo(17)? printf("Yes\n"): printf("No\n");
isPowerOfTwo(16)? printf("Yes\n"): printf("No\n");
isPowerOfTwo(2)? printf("Yes\n"): printf("No\n");
isPowerOfTwo(18)? printf("Yes\n"): printf("No\n");
isPowerOfTwo(1)? printf("Yes\n"): printf("No\n");
return 0;
}

Output:
No
No
Yes
Yes
No
Yes

3. All power of two numbers have only one bit set. So count the no. of set bits and if you get 1 then number is a power of 2. Please see
http://geeksforgeeks.org/?p=1176 for counting set bits.
4. If we subtract a power of 2 numbers by 1 then all unset bits after the only set bit become set; and the set bit become unset.
For example for 4 ( 100) and 16(10000), we get following after subtracting 1
3 > 011
15 > 01111
So, if a number n is a power of 2 then bitwise & of n and n-1 will be zero. We can say n is a power of 2 or not based on value of n&(n-1). The
expression n&(n-1) will not work when n is 0. To handle this case also, our expression will become n& (!n&(n-1)) (thanks to Mohammad for
adding this case).
Below is the implementation of this method.
#include<stdio.h>
#define bool int
/* Function to check if x is power of 2*/
bool isPowerOfTwo (int x)
{
/* First x in the below expression is for the case when x is 0 */
return x && (!(x&(x-1)));
}
/*Driver program to test above function*/
int main()
{
isPowerOfTwo(31)? printf("Yes\n"): printf("No\n");
isPowerOfTwo(17)? printf("Yes\n"): printf("No\n");
isPowerOfTwo(16)? printf("Yes\n"): printf("No\n");
isPowerOfTwo(2)? printf("Yes\n"): printf("No\n");
isPowerOfTwo(18)? printf("Yes\n"): printf("No\n");

isPowerOfTwo(1)? printf("Yes\n"): printf("No\n");

return 0;
}

Output:
No
No
Yes
Yes
No
Yes

Position of rightmost set bit

Write a one line C function to return position of first 1 from right to left, in binary representation of an Integer.
I/P
O/P
I/P
O/P

18,
2
19,
1

Binary Representation 010010

Binary Representation 010011

Let I/P be 12 (1100)

Algorithm: (Example 18(010010))
1. Take two's complement of the given no as all bits are reverted
except the first '1' from right to left (10111)
2 Do an bit-wise & with original no, this will return no with the
required one only (00010)
3 Take the log2 of the no, you will get position -1 (1)
4 Add 1 (2)

Program:
#include<stdio.h>
#include<math.h>
unsigned int getFirstSetBitPos(int n)
{
return log2(n&-n)+1;
}
int main()
{
int n = 12;
printf("%u", getFirstSetBitPos(n));
getchar();
return 0;
}

Binary representation of a given number

Write a program to print Binary representation of a given number.
Source: Microsoft Interview Set-3
Method 1: Iterative
For any number, we can check whether its ith bit is 0(OFF) or 1(ON) by bitwise ANDing it with 2^i (2 raise to i).
1) Let us take number 'NUM' and we want to check whether it's 0th bit is ON or OFF
bit = 2 ^ 0 (0th bit)
if NUM & bit == 1 means 0th bit is ON else 0th bit is OFF
2) Similarly if we want to check whether 5th bit is ON or OFF
bit = 2 ^ 5 (5th bit)
if NUM & bit == 1 means its 5th bit is ON else 5th bit is OFF.

Let us take unsigned integer (32 bit), which consist of 0-31 bits. To print binary representation of unsigned integer, start from 31th bit, check
whether 31th bit is ON or OFF, if it is ON print 1 else print 0. Now check whether 30th bit is ON or OFF, if it is ON print 1 else print 0, do this
for all bits from 31 to 0, finally we will get binary representation of number.
void bin(unsigned n)
{
unsigned i;
for (i = 1 << 31; i > 0; i = i / 2)
(n & i)? printf("1"): printf("0");
}
int main(void)
{
bin(7);
printf("\n");
bin(4);
}

Method 2: Recursive
Following is recursive method to print binary representation of NUM.
step 1) if NUM > 1
a) push NUM on stack
b) recursively call function with 'NUM / 2'
step 2)
a) pop NUM from stack, divide it by 2 and print it's remainder.
void bin(unsigned n)
{
/* step 1 */
if (n > 1)
bin(n/2);
/* step 2 */
printf("%d", n % 2);
}
int main(void)
{
bin(7);
printf("\n");
bin(4);
}

Swap all odd and even bits

Given an unsigned integer, swap all odd bits with even bits. For example, if the given number is 23 (00010111), it should be converted to 43
(00101011). Every even position bit is swapped with adjacent bit on right side (even position bits are highlighted in binary representation of 23),
and every odd position bit is swapped with adjacent on left side.
If we take a closer look at the example, we can observe that we basically need to right shift (>>) all even bits (In the above example, even bits of
23 are highlighted) by 1 so that they become odd bits (highlighted in 43), and left shift (<<) all odd bits by 1 so that they become even bits. The
following solution is based on this observation. The solution assumes that input number is stored using 32 bits. Let the input number be x 1) Get all
even bits of x by doing bitwise and of x with 0xAAAAAAAA. The number 0xAAAAAAAA is a 32 bit number with all even bits set as 1 and all
odd bits as 0. 2) Get all odd bits of x by doing bitwise and of x with 0x55555555. The number 0x55555555 is a 32 bit number with all odd bits
set as 1 and all even bits as 0. 3) Right shift all even bits. 4) Left shift all odd bits. 5) Combine new even and odd bits and return.
// C program to swap even and odd bits of a given number
#include <stdio.h>
unsigned int swapBits(unsigned int x)
{
// Get all even bits of x
unsigned int even_bits = x & 0xAAAAAAAA;
// Get all odd bits of x
unsigned int odd_bits = x & 0x55555555;
even_bits >>= 1; // Right shift even bits
odd_bits <<= 1; // Left shift odd bits
return (even_bits | odd_bits); // Combine even and odd bits
}
// Driver program to test above function
int main()
{
unsigned int x = 23; // 00010111
// Output is 43 (00101011)
printf("%u ", swapBits(x));
return 0;
}

Output:
43

Find position of the only set bit

Given a number having only one 1 and all other 0s in its binary representation, find position of the only set bit. Source: Microsoft Interview | 18
The idea is to start from rightmost bit and one by one check value of every bit. Following is detailed algorithm.
1) If number is power of two then and then only its binary representation contains only one 1. Thats why check whether given number is power of
2 or not. If given number is not power of 2, then print error message and exit.
2) Initialize two variables; i = 1 (for looping) and pos = 1 (to find position of set bit)
3) Inside loop, do bitwise AND of i and number N. If value of this operation is true, then pos bit is set, so break the loop and return position.
Otherwise, increment pos by 1 and left shift i by 1 and repeat the procedure.
// C program to find position of only set bit in a given number
#include <stdio.h>
// A utility function to check whether n is power of 2 or not. See http://goo.gl/17Arj
int isPowerOfTwo(unsigned n)
{ return n && (! (n & (n-1)) ); }
// Returns position of the only set bit in 'n'
int findPosition(unsigned n)
{
if (!isPowerOfTwo(n))
return -1;
unsigned i = 1, pos = 1;
// Iterate through bits of n till we find a set bit
// i&n will be non-zero only when 'i' and 'n' have a set bit
// at same position
while (!(i & n))
{
// Unset current bit and set the next bit in 'i'
i = i << 1;
// increment position
++pos;
}
return pos;
}
// Driver program to test above function
int main(void)
{
int n = 16;
int pos = findPosition(n);
(pos == -1)? printf("n = %d, Invalid number\n", n):
printf("n = %d, Position %d \n", n, pos);
n = 12;
pos = findPosition(n);
(pos == -1)? printf("n = %d, Invalid number\n", n):
printf("n = %d, Position %d \n", n, pos);
n = 128;
pos = findPosition(n);
(pos == -1)? printf("n = %d, Invalid number\n", n):
printf("n = %d, Position %d \n", n, pos);
return 0;
}

Output:
n = 16, Position 5
n = 12, Invalid number
n = 128, Position 8

Following is another method for this problem. The idea is to one by one right shift the set bit of given number n until n becomes 0. Count how
many times we shifted to make n zero. The final count is position of the set bit.
// C program to find position of only set bit in a given number
#include <stdio.h>
// A utility function to check whether n is power of 2 or not

int isPowerOfTwo(unsigned n)
{ return n && (! (n & (n-1)) ); }
// Returns position of the only set bit in 'n'
int findPosition(unsigned n)
{
if (!isPowerOfTwo(n))
return -1;
unsigned count = 0;
// One by one move the only set bit to right till it reaches end
while (n)
{
n = n >> 1;
// increment count of shifts
++count;
}
return count;
}
// Driver program to test above function
int main(void)
{
int n = 0;
int pos = findPosition(n);
(pos == -1)? printf("n = %d, Invalid number\n", n):
printf("n = %d, Position %d \n", n, pos);
n = 12;
pos = findPosition(n);
(pos == -1)? printf("n = %d, Invalid number\n", n):
printf("n = %d, Position %d \n", n, pos);
n = 128;
pos = findPosition(n);
(pos == -1)? printf("n = %d, Invalid number\n", n):
printf("n = %d, Position %d \n", n, pos);
return 0;
}

Output:
n = 0, Invalid number
n = 12, Invalid number
n = 128, Position 8

We can also use log base 2 to find the position. Thanks to Arunkumar for suggesting this solution.
#include <stdio.h>
unsigned int Log2n(unsigned int n)
{
return (n > 1)? 1 + Log2n(n/2): 0;
}
int isPowerOfTwo(unsigned n)
{
return n && (! (n & (n-1)) );
}
int findPosition(unsigned n)
{
if (!isPowerOfTwo(n))
return -1;
return Log2n(n) + 1;
}
// Driver program to test above function
int main(void)
{
int n = 0;
int pos = findPosition(n);
(pos == -1)? printf("n = %d, Invalid number\n", n):
printf("n = %d, Position %d \n", n, pos);
n = 12;
pos = findPosition(n);

(pos == -1)? printf("n = %d, Invalid number\n", n):

printf("n = %d, Position %d \n", n, pos);
n = 128;
pos = findPosition(n);
(pos == -1)? printf("n = %d, Invalid number\n", n):
printf("n = %d, Position %d \n", n, pos);
return 0;
}

Output:
n = 0, Invalid number
n = 12, Invalid number
n = 128, Position 8

Divide and Conquer | Set 4 (Karatsuba algorithm for fast multiplication)

Given two binary strings that represent value of two integers, find the product of two strings. For example, if the first bit string is 1100 and second
bit string is 1010, output should be 120.
For simplicity, let the length of two strings be same and be n.
A Naive Approach is to follow the process we study in school. One by one take all bits of second number and multiply it with all bits of first
number. Finally add all multiplications. This algorithm takes O(n^2) time.

Using Divide and Conquer, we can multiply two integers in less time complexity. We divide the given numbers in two halves. Let the given
numbers be X and Y.
For simplicity let us assume that n is even
X = Xl*2n/2 + Xr
Y = Yl*2n/2 + Yr

[Xl and Xr contain leftmost and rightmost n/2 bits of X]

[Yl and Yr contain leftmost and rightmost n/2 bits of Y]

The product XY can be written as following.

XY = (Xl*2n/2 + Xr)(Yl*2n/2 + Yr)
= 2n XlYl + 2n/2(XlYr + XrYl) + XrYr

If we take a look at the above formula, there are four multiplications of size n/2, so we basically divided the problem of size n into for subproblems of size n/2. But that doesnt help because solution of recurrence T(n) = 4T(n/2) + O(n) is O(n^2). The tricky part of this algorithm is to
change the middle two terms to some other form so that only one extra multiplication would be sufficient. The following is tricky expression for
middle two terms.
XlYr + XrYl = (Xl + Xr)(Yl + Yr) - XlYl- XrYr

So the final value of XY becomes

XY = 2n XlYl + 2n/2 * [(Xl + Xr)(Yl + Yr) - XlYl - XrYr] + XrYr

With above trick, the recurrence becomes T(n) = 3T(n/2) + O(n) and solution of this recurrence is O(n1.59).
What if the lengths of input strings are different and are not even? To handle the different length case, we append 0s in the beginning. To
handle odd length, we put floor(n/2) bits in left half and ceil(n/2) bits in right half. So the expression for XY changes to following.
XY = 22ceil(n/2) XlYl + 2ceil(n/2) * [(Xl + Xr)(Yl + Yr) - XlYl - XrYr] + XrYr

The above algorithm is called Karatsuba algorithm and it can be used for any base.
Following is C++ implementation of above algorithm.
// C++ implementation of Karatsuba algorithm for bit string multiplication.
#include<iostream>
#include<stdio.h>
using namespace std;
// FOLLOWING TWO FUNCTIONS ARE COPIED FROM http://goo.gl/q0OhZ
// Helper method: given two unequal sized bit strings, converts them to
// same length by adding leading 0s in the smaller string. Returns the
// the new length
int makeEqualLength(string &str1, string &str2)
{
int len1 = str1.size();
int len2 = str2.size();
if (len1 < len2)
{
for (int i = 0 ; i < len2 - len1 ; i++)
str1 = '0' + str1;
return len2;

}
else if (len1 > len2)
{
for (int i = 0 ; i < len1 - len2 ; i++)
str2 = '0' + str2;
}
return len1; // If len1 >= len2
}
// The main function that adds two bit sequences and returns the addition
string addBitStrings( string first, string second )
{
string result; // To store the sum bits
// make the lengths same before adding
int length = makeEqualLength(first, second);
int carry = 0; // Initialize carry
// Add all bits one by one
for (int i = length-1 ; i >= 0 ; i--)
{
int firstBit = first.at(i) - '0';
int secondBit = second.at(i) - '0';
// boolean expression for sum of 3 bits
int sum = (firstBit ^ secondBit ^ carry)+'0';
result = (char)sum + result;
// boolean expression for 3-bit addition
carry = (firstBit&secondBit) | (secondBit&carry) | (firstBit&carry);
}
// if overflow, then add a leading 1
if (carry) result = '1' + result;
return result;
}
// A utility function to multiply single bits of strings a and b
int multiplyiSingleBit(string a, string b)
{ return (a[0] - '0')*(b[0] - '0'); }
// The main function that multiplies two bit strings X and Y and returns
// result as long integer
long int multiply(string X, string Y)
{
// Find the maximum of lengths of x and Y and make length
// of smaller string same as that of larger string
int n = makeEqualLength(X, Y);
// Base cases
if (n == 0) return 0;
if (n == 1) return multiplyiSingleBit(X, Y);
int fh = n/2; // First half of string, floor(n/2)
int sh = (n-fh); // Second half of string, ceil(n/2)
// Find the first half and second half of first string.
// Refer http://goo.gl/lLmgn for substr method
string Xl = X.substr(0, fh);
string Xr = X.substr(fh, sh);
// Find the first half and second half of second string
string Yl = Y.substr(0, fh);
string Yr = Y.substr(fh, sh);
// Recursively calculate the three products of inputs of size n/2
long int P1 = multiply(Xl, Yl);
long int P2 = multiply(Xr, Yr);
long int P3 = multiply(addBitStrings(Xl, Xr), addBitStrings(Yl, Yr));
// Combine the three products to get the final result.
return P1*(1<<(2*sh)) + (P3 - P1 - P2)*(1<<sh) + P2;
}
// Driver program to test aboev functions
int main()
{
printf ("%ld\n", multiply("1100", "1010"));
printf ("%ld\n", multiply("110", "1010"));

printf
printf
printf
printf
printf

("%ld\n",
("%ld\n",
("%ld\n",
("%ld\n",
("%ld\n",

multiply("11", "1010"));
multiply("1", "1010"));
multiply("0", "1010"));
multiply("111", "111"));
multiply("11", "11"));

Output:
120
60
30
10
0
49
9

Time Complexity: Time complexity of the above solution is O(n1.59).

Time complexity of multiplication can be further improved using another Divide and Conquer algorithm, fast Fourier transform. We will soon be
discussing fast Fourier transform as a separate post.
Exercise
The above program returns a long int value and will not work for big strings. Extend the above program to return a string instead of a long int value.
References:
Wikipedia page for Karatsuba algorithm
Algorithms 1st Edition by Sanjoy Dasgupta, Christos Papadimitriou and Umesh Vazirani
http://courses.csail.mit.edu/6.006/spring11/exams/notes3-karatsuba
http://www.cc.gatech.edu/~ninamf/Algos11/lectures/lect0131.pdf

How to swap two numbers without using a temporary variable?

Given two variables, x and y, swap two variables without using a third variable.
Method 1 (Using Arithmetic Operators)
The idea is to get sum in one of the two given numbers. The numbers can then be swapped using the sum and subtraction from sum.
#include <stdio.h>
int main()
{
int x = 10, y = 5;
// Code
x = x +
y = x x = x -

to
y;
y;
y;

swap 'x' and 'y'

// x now becomes 15
// y becomes 10
// x becomes 5

printf("After Swapping: x = %d, y = %d", x, y);

return 0;
}

Output:
After Swapping: x = 5, y = 10

Multiplication and division can also be used for swapping.

#include <stdio.h>
int main()
{
int x = 10, y = 5;
// Code
x = x *
y = x /
x = x /

to
y;
y;
y;

swap 'x' and 'y'

// x now becomes 50
// y becomes 10
// x becomes 5

printf("After Swapping: x = %d, y = %d", x, y);

return 0;
}

Output:
After Swapping: x = 5, y = 10

Method 2 (Using Bitwise XOR)

The bitwise XOR operator can be used to swap two variables. The XOR of two numbers x and y returns a number which has all the bits as 1
wherever bits of x and y differ. For example XOR of 10 (In Binary 1010) and 5 (In Binary 0101) is 1111 and XOR of 7 (0111) and 5 (0101) is
(0010).
#include <stdio.h>
int main()
{
int x = 10, y = 5;
// Code
x = x ^
y = x ^
x = x ^

to
y;
y;
y;

swap 'x' (1010) and 'y' (0101)

// x now becomes 15 (1111)
// y becomes 10 (1010)
// x becomes 5 (0101)

printf("After Swapping: x = %d, y = %d", x, y);

return 0;
}

Output:
After Swapping: x = 5, y = 10

Problems with above methods

1) The multiplication and division based approach doesn work if one of the numbers is 0 as the product becomes 0 irrespective of the other
number.
2) Both Arithmetic solutions may cause arithmetic overflow. If x and y are too large, addition and multiplication may go out of integer range.

3) When we use pointers to variable and make a function swap, all of the above methods fail when both pointers point to the same variable. Lets
take a look what will happen in this case if both are pointing to the same variable.
// Bitwise XOR based method
x = x ^ x; // x becomes 0
x = x ^ x; // x remains 0
x = x ^ x; // x remains 0
// Arithmetic based method
x = x + x; // x becomes 2x
x = x x; // x becomes 0
x = x x; // x remains 0
Let us see the following program.
#include <stdio.h>
void swap(int *xp, int *yp)
{
*xp = *xp ^ *yp;
*yp = *xp ^ *yp;
*xp = *xp ^ *yp;
}
int main()
{
int x = 10;
swap(&x, &x);
printf("After swap(&x, &x): x = %d", x);
return 0;
}

Output:
After swap(&x, &x): x = 0

Swapping a variable with itself may needed in many standard algorithms. For example see this implementation of QuickSort where we may swap a
variable with itself. The above problem can be avoided by putting a condition before the swapping.
#include <stdio.h>
void swap(int *xp, int *yp)
{
if (xp == yp) // Check if the two addresses are same
return;
*xp = *xp + *yp;
*yp = *xp - *yp;
*xp = *xp - *yp;
}
int main()
{
int x = 10;
swap(&x, &x);
printf("After swap(&x, &x): x = %d", x);
return 0;
}

Output:
After swap(&x, &x): x = 10

Check if a number is multiple of 9 using bitwise operators

Given a number n, write a function that returns true if n is divisible by 9, else false. The most simple way to check for ns divisibility by 9 is to do
n%9.
Another method is to sum the digits of n. If sum of digits is multiple of 9, then n is multiple of 9.
The above methods are not bitwise operators based methods and require use of % and /.
The bitwise operators are generally faster than modulo and division operators. Following is a bitwise operator based method to check divisibility
by 9.
#include<iostream>
using namespace std;
// Bitwise operator based function to check divisibility by 9
bool isDivBy9(int n)
{
// Base cases
if (n == 0 || n == 9)
return true;
if (n < 9)
return false;
// If n is greater than 9, then recur for [floor(n/9) - n%8]
return isDivBy9((int)(n>>3) - (int)(n&7));
}
// Driver program to test above function
int main()
{
// Let us print all multiples of 9 from 0 to 100
// using above method
for (int i = 0; i < 100; i++)
if (isDivBy9(i))
cout << i << " ";
return 0;
}

Output:
0 9 18 27 36 45 54 63 72 81 90 99

How does this work?

n/9 can be written in terms of n/8 using the following simple formula.
n/9 = n/8 - n/72

Since we need to use bitwise operators, we get the value of floor(n/8) using n>>3 and get value of n%8 using n&7. We need to write above
expression in terms of floor(n/8) and n%8.
n/8 is equal to floor(n/8) + (n%8)/8?. Let us write the above expression in terms of floor(n/8) and n%8
n/9 = floor(n/8) + (n%8)/8 - [floor(n/8) + (n%8)/8]/9
n/9 = floor(n/8) - [floor(n/8) - 9(n%8)/8 + (n%8)/8]/9
n/9 = floor(n/8) - [floor(n/8) - n%8]/9

From above equation, n is a multiple of 9 only if the expression floor(n/8) [floor(n/8) n%8]/9 is an integer. This expression can only be an integer
if the sub-expression [floor(n/8) n%8]/9 is an integer. The subexpression can only be an integer if [floor(n/8) n%8] is a multiple of 9. So the
problem reduces to a smaller value which can be written in terms of bitwise operators.

Swap two nibbles in a byte

A nibble is a four-bit aggregation, or half an octet. There are two nibbles in a byte.
Given a byte, swap the two nibbles in it. For example 100 is be represented as 01100100 in a byte (or 8 bits). The two nibbles are (0110) and
(0100). If we swap the two nibbles, we get 01000110 which is 70 in decimal.
To swap the nibbles, we can use bitwise &, bitwise <<' and '>> operators. A byte can be represented using a unsigned char in C as size of char is
1 byte in a typical C compiler. Following is C program to swap the two nibbles in a byte.
#include <stdio.h>
unsigned char swapNibbles(unsigned char x)
{
return ( (x & 0x0F)<<4 | (x & 0xF0)>>4 );
}
int main()
{
unsigned char x = 100;
printf("%u", swapNibbles(x));
return 0;
}

Output:
70

Explanation:
100 is 01100100 in binary. The operation can be split mainly in two parts
1) The expression x & 0x0F gives us last 4 bits of x. For x = 100, the result is 00000100. Using bitwise <<' operator, we shift the last four bits to
the left 4 times and make the new last four bits as 0. The result after shift is 01000000. 2) The expression x & 0xF0 gives us first four bits of x.
For x = 100, the result is 01100000. Using bitwise >> operator, we shift the digit to the right 4 times and make the first four bits as 0. The result
after shift is 00000110.
At the end we use the bitwise OR | operation of the two expressions explained above. The OR operator places first nibble to the end and last
nibble to first. For x = 100, the value of (01000000) OR (00000110) gives the result 01000110 which is equal to 70 in decimal.

How to turn off a particular bit in a number?

Difficulty Level: Rookie
Given a number n and a value k, turn of the kth bit in n.
Examples:
Input: n = 15, k = 1
Output: 14
Input: n = 15, k = 2
Output: 13
Input: n = 15, k = 3
Output: 11
Input: n = 15, k = 4
Output: 7
Input: n = 15, k >= 5
Output: 15

The idea is to use bitwise <<, & and ~ operators. Using expression "~(1 << (k - 1)), we get a number which has all bits set, except the kth bit. If
we do bitwise & of this expression with n, we get a number which has all bits same as n except the kth bit which is 0.
Following is C++ implementation of this.
#include <iostream>
using namespace std;
// Returns a number that has all bits same as n
// except the k'th bit which is made 0
int turnOffK(int n, int k)
{
// k must be greater than 0
if (k <= 0) return n;
// Do & of n with a number with all set bits except
// the k'th bit
return (n & ~(1 << (k - 1)));
}
// Driver program to test above function
int main()
{
int n = 15;
int k = 4;
cout << turnOffK(n, k);
return 0;
}

Output:
7

Exercise: Write a function turnOnK() that turns the kth bit on.

Check if binary representation of a number is palindrome

Given an integer x, write a C function that returns true if binary representation of x is palindrome else return false.
For example a numbers with binary representation as 10..01 is palindrome and number with binary representation as 10..00 is not palindrome.
The idea is similar to checking a string is palindrome or not. We start from leftmost and rightmost bits and compare bits one by one. If we find a
mismatch, then return false.
Algorithm:
isPalindrome(x)
1) Find number of bits in x using sizeof() operator.
2) Initialize left and right positions as 1 and n respectively.
3) Do following while left l is smaller than right r.
....a) If bit at position l is not same as bit at position r, then return false.
....b) Increment l and decrement r, i.e., do l++ and r-.
4) If we reach here, it means we didnt find a mismatching bit.
To find the bit at a given position, we can use the idea similar to this post. The expression x & (1 << (k-1)) gives us non-zero value if bit at kth
position from right is set and gives a zero value if if kth bit is not set.
Following is C++ implementation of the above algorithm.
#include<iostream>
using namespace std;
// This function returns true if k'th bit in x is set (or 1).
// For example if x (0010) is 2 and k is 2, then it returns true
bool isKthBitSet(unsigned int x, unsigned int k)
{
return (x & (1 << (k-1)))? true: false;
}
// This function returns true if binary representation of x is
// palindrome. For example (1000...001) is paldindrome
bool isPalindrome(unsigned int x)
{
int l = 1; // Initialize left position
int r = sizeof(unsigned int)*8; // initialize right position
// One by one compare bits
while (l < r)
{
if (isKthBitSet(x, l) != isKthBitSet(x, r))
return false;
l++;
r--;
}
return true;
}
// Driver program to test above function
int main()
{
unsigned int x = 1<<15 + 1<<16;
cout << isPalindrome(x) << endl;
x = 1<<31 + 1;
cout << isPalindrome(x) << endl;
return 0;
}

Output:
1
1

Graph and its representations

Graph is a data structure that consists of following two components:
1. A finite set of vertices also called as nodes.
2. A finite set of ordered pair of the form (u, v) called as edge. The pair is ordered because (u, v) is not same as (v, u) in case of directed graph(digraph). The pair of form (u, v) indicates that there is an edge from vertex u to vertex v. The edges may contain weight/value/cost.
Graphs are used to represent many real life applications: Graphs are used to represent networks. The networks may include paths in a city or
telephone network or circuit network. Graphs are also used in social networks like linkedIn, facebook. For example, in facebook, each person is
represented with a vertex(or node). Each node is a structure and contains information like person id, name, gender and locale. This can be easily
viewed by http://graph.facebook.com/barnwal.aashish where barnwal.aashish is the profile name. See this for more applications of graph.
Following is an example undirected graph with 5 vertices.

Following two are the most commonly used representations of graph.

1. Adjacency Matrix
2. Adjacency List
There are other representations also like, Incidence Matrix and Incidence List. The choice of the graph representation is situation specific. It totally
depends on the type of operations to be performed and ease of use.
Adjacency Matrix:
Adjacency Matrix is a 2D array of size V x V where V is the number of vertices in a graph. Let the 2D array be adj[][], a slot adj[i][j] = 1
indicates that there is an edge from vertex i to vertex j. Adjacency matrix for undirected graph is always symmetric. Adjacency Matrix is also used
to represent weighted graphs. If adj[i][j] = w, then there is an edge from vertex i to vertex j with weight w.
The adjacency matrix for the above example graph is:

Adjacency Matrix
Representation of the above
graph
Pros: Representation is easier to implement and follow. Removing an edge takes O(1) time. Queries like whether there is an edge from vertex u to
vertex v are efficient and can be done O(1).
Cons: Consumes more space O(V^2). Even if the graph is sparse(contains less number of edges), it consumes the same space. Adding a vertex is
O(V^2) time.

Adjacency List:
An array of linked lists is used. Size of the array is equal to number of vertices. Let the array be array[]. An entry array[i] represents the linked list
of vertices adjacent to the ith vertex. This representation can also be used to represent a weighted graph. The weights of edges can be stored in
nodes of linked lists. Following is adjacency list representation of the above graph.

Adjacency List Representation of the above Graph

Below is C code for adjacency list representation of an undirected graph:
// A C Program to demonstrate adjacency list representation of graphs
#include <stdio.h>
#include <stdlib.h>
// A structure to represent an adjacency list node
struct AdjListNode
{
int dest;
struct AdjListNode* next;
};
// A structure to represent an adjacency list
struct AdjList
{
struct AdjListNode *head; // pointer to head node of list
};
// A structure to represent a graph. A graph is an array of adjacency lists.
// Size of array will be V (number of vertices in graph)
struct Graph
{
int V;
struct AdjList* array;
};
// A utility function to create a new adjacency list node
struct AdjListNode* newAdjListNode(int dest)
{
struct AdjListNode* newNode =
(struct AdjListNode*) malloc(sizeof(struct AdjListNode));
newNode->dest = dest;
newNode->next = NULL;
return newNode;
}
// A utility function that creates a graph of V vertices
struct Graph* createGraph(int V)
{
struct Graph* graph = (struct Graph*) malloc(sizeof(struct Graph));
graph->V = V;
// Create an array of adjacency lists. Size of array will be V
graph->array = (struct AdjList*) malloc(V * sizeof(struct AdjList));
// Initialize each adjacency list as empty by making head as NULL
int i;
for (i = 0; i < V; ++i)
graph->array[i].head = NULL;
return graph;
}
// Adds an edge to an undirected graph
void addEdge(struct Graph* graph, int src, int dest)
{
// Add an edge from src to dest. A new node is added to the adjacency
// list of src. The node is added at the begining
struct AdjListNode* newNode = newAdjListNode(dest);
newNode->next = graph->array[src].head;
graph->array[src].head = newNode;
// Since graph is undirected, add an edge from dest to src also
newNode = newAdjListNode(src);
newNode->next = graph->array[dest].head;
graph->array[dest].head = newNode;

}
// A utility function to print the adjacenncy list representation of graph
void printGraph(struct Graph* graph)
{
int v;
for (v = 0; v < graph->V; ++v)
{
struct AdjListNode* pCrawl = graph->array[v].head;
printf("\n Adjacency list of vertex %d\n head ", v);
while (pCrawl)
{
printf("-> %d", pCrawl->dest);
pCrawl = pCrawl->next;
}
printf("\n");
}
}
// Driver program to test above functions
int main()
{
// create the graph given in above fugure
int V = 5;
struct Graph* graph = createGraph(V);
addEdge(graph, 0, 1);
addEdge(graph, 0, 4);
addEdge(graph, 1, 2);
addEdge(graph, 1, 3);
addEdge(graph, 1, 4);
addEdge(graph, 2, 3);
addEdge(graph, 3, 4);
// print the adjacency list representation of the above graph
printGraph(graph);
return 0;
}

Output:
Adjacency list of vertex 0
head -> 4-> 1
Adjacency list of vertex 1
head -> 4-> 3-> 2-> 0
Adjacency list of vertex 2
head -> 3-> 1
Adjacency list of vertex 3
head -> 4-> 2-> 1
Adjacency list of vertex 4
head -> 3-> 1-> 0

Pros: Saves space O(|V|+|E|) . In the worst case, there can be C(V, 2) number of edges in a graph thus consuming O(V^2) space. Adding a
vertex is easier.
Cons: Queries like whether there is an edge from vertex u to vertex v are not efficient and can be done O(V).
Reference:
http://en.wikipedia.org/wiki/Graph_%28abstract_data_type%29

Breadth First Traversal for a Graph

Breadth First Traversal (or Search) for a graph is similar to Breadth First Traversal of a tree (See mthod 2 of this post). The only catch here is,
unlike trees, graphs may contain cycles, so we may come to the same node again. To avoid processing a node more than once, we use a boolean
visited array. For simplicity, it is assumed that all vertices are reachable from the starting vertex.
For example, in the following graph, we start traversal from vertex 2. When we come to vertex 0, we look for all adjacent vertices of it. 2 is also
an adjacent vertex of 0. If we dont mark visited vertices, then 2 will be processed again and it will become a non-terminating process. A Breadth
First Traversal of the following graph is 2, 0, 3, 1.

Following are C++ and Java implementations of simple Breadth First Traversal from a given source.
The C++ implementation uses adjacency list representation of graphs. STLs list container is used to store lists of adjacent nodes and queue of
nodes needed for BFS traversal.

C++
// Program to print BFS traversal from a given source vertex. BFS(int s)
// traverses vertices reachable from s.
#include<iostream>
#include <list>
using namespace std;
// This class represents a directed graph using adjacency list representation
class Graph
{
int V;
// No. of vertices
list<int> *adj;
// Pointer to an array containing adjacency lists
public:
Graph(int V); // Constructor
void addEdge(int v, int w); // function to add an edge to graph
void BFS(int s); // prints BFS traversal from a given source s
};
Graph::Graph(int V)
{
this->V = V;
adj = new list<int>[V];
}
void Graph::addEdge(int v, int w)
{
adj[v].push_back(w); // Add w to vs list.
}
void Graph::BFS(int s)
{
// Mark all the vertices as not visited
bool *visited = new bool[V];
for(int i = 0; i < V; i++)
visited[i] = false;
// Create a queue for BFS
list<int> queue;
// Mark the current node as visited and enqueue it
visited[s] = true;
queue.push_back(s);
// 'i' will be used to get all adjacent vertices of a vertex
list<int>::iterator i;
while(!queue.empty())
{
// Dequeue a vertex from queue and print it

s = queue.front();
cout << s << " ";
queue.pop_front();
// Get all adjacent vertices of the dequeued vertex s
// If a adjacent has not been visited, then mark it visited
// and enqueue it
for(i = adj[s].begin(); i != adj[s].end(); ++i)
{
if(!visited[*i])
{
visited[*i] = true;
queue.push_back(*i);
}
}
}
}
// Driver program to test methods of graph class
int main()
{
// Create a graph given in the above diagram
Graph g(4);
g.addEdge(0, 1);
g.addEdge(0, 2);
g.addEdge(1, 2);
g.addEdge(2, 0);
g.addEdge(2, 3);
g.addEdge(3, 3);
cout << "Following is Breadth First Traversal (starting from vertex 2) \n";
g.BFS(2);
return 0;
}

Java
// Java program to print BFS traversal from a given source vertex.
// BFS(int s) traverses vertices reachable from s.
import java.io.*;
import java.util.*;
// This class represents a directed graph using adjacency list
// representation
class Graph
{
private int V; // No. of vertices
private LinkedList<Integer> adj[]; //Adjacency Lists
// Constructor
Graph(int v)
{
V = v;
adj = new LinkedList[v];
for (int i=0; i<v; ++i)
adj[i] = new LinkedList();
}
// Function to add an edge into the graph
void addEdge(int v,int w)
{
adj[v].add(w);
}
// prints BFS traversal from a given source s
void BFS(int s)
{
// Mark all the vertices as not visited(By default
// set as false)
boolean visited[] = new boolean[V];
// Create a queue for BFS
LinkedList<Integer> queue = new LinkedList<Integer>();
// Mark the current node as visited and enqueue it
visited[s]=true;
queue.add(s);
while (queue.size() != 0)

{
// Dequeue a vertex from queue and print it
s = queue.poll();
System.out.print(s+" ");
// Get all adjacent vertices of the dequeued vertex s
// If a adjacent has not been visited, then mark it
// visited and enqueue it
Iterator<Integer> i = adj[s].listIterator();
while (i.hasNext())
{
int n = i.next();
if (!visited[n])
{
visited[n] = true;
queue.add(n);
}
}
}
}
// Driver method to
public static void main(String args[])
{
Graph g = new Graph(4);
g.addEdge(0,
g.addEdge(0,
g.addEdge(1,
g.addEdge(2,
g.addEdge(2,
g.addEdge(3,

1);
2);
2);
0);
3);
3);

System.out.println("Following is Breadth First Traversal "+

"(starting from vertex 2)");
g.BFS(2);
}
}
// This code is contributed by Aakash Hasija

Following is Breadth First Traversal (starting from vertex 2)

2 0 3 1

Note that the above code traverses only the vertices reachable from a given source vertex. All the vertices may not be reachable from a given
vertex (example Disconnected graph). To print all the vertices, we can modify the BFS function to do traversal starting from all nodes one by one
(Like the DFS modified version) .
Time Complexity: O(V+E) where V is number of vertices in the graph and E is number of edges in the graph.
Also see Depth First Traversal

Depth First Traversal for a Graph

Depth First Traversal (or Search) for a graph is similar to Depth First Traversal of a tree. The only catch here is, unlike trees, graphs may contain
cycles, so we may come to the same node again. To avoid processing a node more than once, we use a boolean visited array.
For example, in the following graph, we start traversal from vertex 2. When we come to vertex 0, we look for all adjacent vertices of it. 2 is also
an adjacent vertex of 0. If we dont mark visited vertices, then 2 will be processed again and it will become a non-terminating process. A Depth
First Traversal of the following graph is 2, 0, 1, 3.

See this post for all applications of Depth First Traversal.

Following are implementations of simple Depth First Traversal. The C++ implementation uses adjacency list representation of graphs. STLs list
container is used to store lists of adjacent nodes.

C++
// C++ program to print DFS traversal from a given vertex in a given graph
#include<iostream>
#include <list>
using namespace std;
// Graph class represents a directed graph using adjacency list representation
class Graph
{
int V;
// No. of vertices
list<int> *adj;
// Pointer to an array containing adjacency lists
void DFSUtil(int v, bool visited[]); // A function used by DFS
public:
Graph(int V); // Constructor
void addEdge(int v, int w); // function to add an edge to graph
void DFS(int v);
// DFS traversal of the vertices reachable from v
};
Graph::Graph(int V)
{
this->V = V;
adj = new list<int>[V];
}
void Graph::addEdge(int v, int w)
{
adj[v].push_back(w); // Add w to vs list.
}
void Graph::DFSUtil(int v, bool visited[])
{
// Mark the current node as visited and print it
visited[v] = true;
cout << v << " ";
// Recur for all the vertices adjacent to this vertex
list<int>::iterator i;
for (i = adj[v].begin(); i != adj[v].end(); ++i)
if (!visited[*i])
DFSUtil(*i, visited);
}
// DFS traversal of the vertices reachable from v. It uses recursive DFSUtil()
void Graph::DFS(int v)
{
// Mark all the vertices as not visited
bool *visited = new bool[V];
for (int i = 0; i < V; i++)
visited[i] = false;
// Call the recursive helper function to print DFS traversal

DFSUtil(v, visited);
}
int main()
{
// Create a graph given in the above diagram
Graph g(4);
g.addEdge(0, 1);
g.addEdge(0, 2);
g.addEdge(1, 2);
g.addEdge(2, 0);
g.addEdge(2, 3);
g.addEdge(3, 3);
cout << "Following is Depth First Traversal (starting from vertex 2) \n";
g.DFS(2);
return 0;
}

Java
// Java program to print DFS traversal from a given given graph
import java.io.*;
import java.util.*;
// This class represents a directed graph using adjacency list
// representation
class Graph
{
private int V; // No. of vertices
// Array of lists for Adjacency List Representation
private LinkedList<Integer> adj[];
// Constructor
Graph(int v)
{
V = v;
adj = new LinkedList[v];
for (int i=0; i<v; ++i)
adj[i] = new LinkedList();
}
//Function to add an edge into the graph
void addEdge(int v, int w)
{
adj[v].add(w); // Add w to v's list.
}
// A function used by DFS
void DFSUtil(int v,boolean visited[])
{
// Mark the current node as visited and print it
visited[v] = true;
System.out.print(v+" ");
// Recur for all the vertices adjacent to this vertex
Iterator<Integer> i = adj[v].listIterator();
while (i.hasNext())
{
int n = i.next();
if (!visited[n])
DFSUtil(n, visited);
}
}
// The function to do DFS traversal. It uses recursive DFSUtil()
void DFS(int v)
{
// Mark all the vertices as not visited(set as
// false by default in java)
boolean visited[] = new boolean[V];
// Call the recursive helper function to print DFS traversal
DFSUtil(v, visited);
}
public static void main(String args[])
{

Graph g = new Graph(4);

g.addEdge(0,
g.addEdge(0,
g.addEdge(1,
g.addEdge(2,
g.addEdge(2,
g.addEdge(3,

1);
2);
2);
0);
3);
3);

System.out.println("Following is Depth First Traversal "+

"(starting from vertex 2)");
g.DFS(2);
}
}
// This code is contributed by Aakash Hasija

Following is Depth First Traversal (starting from vertex 2)

2 0 1 3

Note that the above code traverses only the vertices reachable from a given source vertex. All the vertices may not be reachable from a given
vertex (example Disconnected graph). To do complete DFS traversal of such graphs, we must call DFSUtil() for every vertex. Also, before calling
DFSUtil(), we should check if it is already printed by some other call of DFSUtil(). Following implementation does the complete graph traversal
even if the nodes are unreachable. The differences from the above code are highlighted in the below code.

C++
// C++ program to print DFS traversal for a given given graph
#include<iostream>
#include <list>
using namespace std;
class Graph
{
int V;
// No. of vertices
list<int> *adj;
// Pointer to an array containing adjacency lists
void DFSUtil(int v, bool visited[]); // A function used by DFS
public:
Graph(int V); // Constructor
void addEdge(int v, int w); // function to add an edge to graph
void DFS();
// prints DFS traversal of the complete graph
};
Graph::Graph(int V)
{
this->V = V;
adj = new list<int>[V];
}
void Graph::addEdge(int v, int w)
{
adj[v].push_back(w); // Add w to vs list.
}
void Graph::DFSUtil(int v, bool visited[])
{
// Mark the current node as visited and print it
visited[v] = true;
cout << v << " ";
// Recur for all the vertices adjacent to this vertex
list<int>::iterator i;
for(i = adj[v].begin(); i != adj[v].end(); ++i)
if(!visited[*i])
DFSUtil(*i, visited);
}
// The function to do DFS traversal. It uses recursive DFSUtil()
void Graph::DFS()
{
// Mark all the vertices as not visited
bool *visited = new bool[V];
for (int i = 0; i < V; i++)
visited[i] = false;
// Call the recursive helper function to print DFS traversal
// starting from all vertices one by one

for (int i = 0; i < V; i++)

if (visited[i] == false)
DFSUtil(i, visited);
}
int main()
{
// Create a graph given in the above diagram
Graph g(4);
g.addEdge(0, 1);
g.addEdge(0, 2);
g.addEdge(1, 2);
g.addEdge(2, 0);
g.addEdge(2, 3);
g.addEdge(3, 3);
cout << "Following is Depth First Traversal\n";
g.DFS();
return 0;
}

Java
// Java program to print DFS traversal from a given given graph
import java.io.*;
import java.util.*;
// This class represents a directed graph using adjacency list
// representation
class Graph
{
private int V; // No. of vertices
// Array of lists for Adjacency List Representation
private LinkedList<Integer> adj[];
// Constructor
Graph(int v)
{
V = v;
adj = new LinkedList[v];
for (int i=0; i<v; ++i)
adj[i] = new LinkedList();
}
//Function to add an edge into the graph
void addEdge(int v, int w)
{
adj[v].add(w); // Add w to v's list.
}
// A function used by DFS
void DFSUtil(int v,boolean visited[])
{
// Mark the current node as visited and print it
visited[v] = true;
System.out.print(v+" ");
// Recur for all the vertices adjacent to this vertex
Iterator<Integer> i = adj[v].listIterator();
while (i.hasNext())
{
int n = i.next();
if (!visited[n])
DFSUtil(n,visited);
}
}
// The function to do DFS traversal. It uses recursive DFSUtil()
void DFS()
{
// Mark all the vertices as not visited(set as
// false by default in java)
boolean visited[] = new boolean[V];
// Call the recursive helper function to print DFS traversal
// starting from all vertices one by one
for (int i=0; i<V; ++i)
if (visited[i] == false)

DFSUtil(i, visited);
}
public static void main(String args[])
{
Graph g = new Graph(4);
g.addEdge(0,
g.addEdge(0,
g.addEdge(1,
g.addEdge(2,
g.addEdge(2,
g.addEdge(3,

1);
2);
2);
0);
3);
3);

System.out.println("Following is Depth First Traversal");

g.DFS();
}
}
// This code is contributed by Aakash Hasija

Following is Depth First Traversal

0 1 2 3

Time Complexity: O(V+E) where V is number of vertices in the graph and E is number of edges in the graph.
Breadth First Traversal for a Graph

Applications of Depth First Search

Depth-first search (DFS) is an algorithm (or technique) for traversing a graph.
Following are the problems that use DFS as a bulding block.
1) For an unweighted graph, DFS traversal of the graph produces the minimum spanning tree and all pair shortest path tree.
2) Detecting cycle in a graph
A graph has cycle if and only if we see a back edge during DFS. So we can run DFS for the graph and check for back edges. (See this for details)
3) Path Finding
We can specialize the DFS algorithm to find a path between two given vertices u and z.
i) Call DFS(G, u) with u as the start vertex.
ii) Use a stack S to keep track of the path between the start vertex and the current vertex.
iii) As soon as destination vertex z is encountered, return the path as the
contents of the stack
See this for details.
4) Topological Sorting
See this for details.
5) To test if a graph is bipartite
We can augment either BFS or DFS when we first discover a new vertex, color it opposited its parents, and for each other edge, check it doesnt
link two vertices of the same color. The first vertex in any connected component can be red or black! See this for details.
6) Finding Strongly Connected Components of a graph A directed graph is called strongly connected if there is a path from each vertex in the
graph to every other vertex. (See this for DFS based algo for finding Strongly Connected Components)
7) Solving puzzles with only one solution, such as mazes. (DFS can be adapted to find all solutions to a maze by only including nodes on the
current path in the visited set.)
Sources:
http://www8.cs.umu.se/kurser/TDBAfl/VT06/algorithms/LEC/LECTUR16/NODE16.HTM
http://en.wikipedia.org/wiki/Depth-first_search
http://www.personal.kent.edu/~rmuhamma/Algorithms/MyAlgorithms/GraphAlgor/depthSearch.htm
http://ww3.algorithmdesign.net/handouts/DFS.pdf

Detect Cycle in a Directed Graph

Given a directed graph, check whether the graph contains a cycle or not. Your function should return true if the given graph contains at least one
cycle, else return false. For example, the following graph contains three cycles 0->2->0, 0->1->2->0 and 3->3, so your function must return true.
Solution
Depth First Traversal can be used to detect cycle in a Graph. DFS for a connected graph produces a tree. There is a cycle in a graph only if there
is a back edge present in the graph. A back edge is an edge that is from a node to itself (selfloop) or one of its ancestor in the tree produced by
DFS. In the following graph, there are 3 back edges, marked with cross sign. We can observe that these 3 back edges indicate 3 cycles present in
the graph.

For a disconnected graph, we get the DFS forrest as output. To detect cycle, we can check for cycle in individual trees by checking back edges.
To detect a back edge, we can keep track of vertices currently in recursion stack of function for DFS traversal. If we reach a vertex that is already
in the recursion stack, then there is a cycle in the tree. The edge that connects current vertex to the vertex in the recursion stack is back edge. We
have used recStack[] array to keep track of vertices in the recursion stack.
// A C++ Program to detect cycle in a graph
#include<iostream>
#include <list>
#include <limits.h>
using namespace std;
class Graph
{
int V;
// No. of vertices
list<int> *adj;
// Pointer to an array containing adjacency lists
bool isCyclicUtil(int v, bool visited[], bool *rs); // used by isCyclic()
public:
Graph(int V); // Constructor
void addEdge(int v, int w); // to add an edge to graph
bool isCyclic();
// returns true if there is a cycle in this graph
};
Graph::Graph(int V)
{
this->V = V;
adj = new list<int>[V];
}
void Graph::addEdge(int v, int w)
{
adj[v].push_back(w); // Add w to vs list.
}
// This function is a variation of DFSUytil() in http://www.geeksforgeeks.org/archives/18212
bool Graph::isCyclicUtil(int v, bool visited[], bool *recStack)
{
if(visited[v] == false)
{
// Mark the current node as visited and part of recursion stack
visited[v] = true;
recStack[v] = true;
// Recur for all the vertices adjacent to this vertex
list<int>::iterator i;
for(i = adj[v].begin(); i != adj[v].end(); ++i)
{
if ( !visited[*i] && isCyclicUtil(*i, visited, recStack) )
return true;
else if (recStack[*i])
return true;
}

}
recStack[v] = false; // remove the vertex from recursion stack
return false;
}
// Returns true if the graph contains a cycle, else false.
// This function is a variation of DFS() in http://www.geeksforgeeks.org/archives/18212
bool Graph::isCyclic()
{
// Mark all the vertices as not visited and not part of recursion
// stack
bool *visited = new bool[V];
bool *recStack = new bool[V];
for(int i = 0; i < V; i++)
{
visited[i] = false;
recStack[i] = false;
}
// Call the recursive helper function to detect cycle in different
// DFS trees
for(int i = 0; i < V; i++)
if (isCyclicUtil(i, visited, recStack))
return true;
return false;
}
int main()
{
// Create a graph given in the above diagram
Graph g(4);
g.addEdge(0, 1);
g.addEdge(0, 2);
g.addEdge(1, 2);
g.addEdge(2, 0);
g.addEdge(2, 3);
g.addEdge(3, 3);
if(g.isCyclic())
cout << "Graph contains cycle";
else
cout << "Graph doesn't contain cycle";
return 0;
}

Output:
Graph contains cycle

Time Complexity of this method is same as time complexity of DFS traversal which is O(V+E).

Union-Find Algorithm | Set 1 (Detect Cycle in a an Undirected Graph)

A disjoint-set data structure is a data structure that keeps track of a set of elements partitioned into a number of disjoint (non-overlapping)
subsets. A union-find algorithm is an algorithm that performs two useful operations on such a data structure:
Find: Determine which subset a particular element is in. This can be used for determining if two elements are in the same subset.
Union: Join two subsets into a single subset.
In this post, we will discuss an application of Disjoint Set Data Structure. The application is to check whether a given graph contains a cycle or not.
Union-Find Algorithm can be used to check whether an undirected graph contains cycle or not. Note that we have discussed an algorithm to
detect cycle. This is another method based on Union-Find. This method assumes that graph doesnt contain any self-loops.
We can keeps track of the subsets in a 1D array, lets call it parent[].
Let us consider the following graph:
0
| \
|
\
1-----2

For each edge, make subsets using both the vertices of the edge. If both the vertices are in the same subset, a cycle is found.
Initially, all slots of parent array are initialized to -1 (means there is only one item in every subset).
0 1 2
-1 -1 -1

Now process all edges one by one.

Edge 0-1: Find the subsets in which vertices 0 and 1 are. Since they are in different subsets, we take the union of them. For taking the union, either
make node 0 as parent of node 1 or vice-versa.
0 1 2
1 -1 -1

<----- 1 is made parent of 0 (1 is now representative of subset {0, 1})

Edge 1-2: 1 is in subset 1 and 2 is in subset 2. So, take union.

0
1

1 2
2 -1

<----- 2 is made parent of 1 (2 is now representative of subset {0, 1, 2})

Edge 0-2: 0 is in subset 2 and 2 is also in subset 2. Hence, including this edge forms a cycle.
How subset of 0 is same as 2?
0->1->2 // 1 is parent of 0 and 2 is parent of 1
Based on the above explanation, below are implementations:

C/C++
// A union-find algorithm to detect cycle in a graph
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
// a structure to represent an edge in graph
struct Edge
{
int src, dest;
};
// a structure to represent a graph
struct Graph
{
// V-> Number of vertices, E-> Number of edges
int V, E;
// graph is represented as an array of edges
struct Edge* edge;
};
// Creates a graph with V vertices and E edges
struct Graph* createGraph(int V, int E)
{

struct Graph* graph =

(struct Graph*) malloc( sizeof(struct Graph) );
graph->V = V;
graph->E = E;
graph->edge =
(struct Edge*) malloc( graph->E * sizeof( struct Edge ) );
return graph;
}
// A utility function to find the subset of an element i
int find(int parent[], int i)
{
if (parent[i] == -1)
return i;
return find(parent, parent[i]);
}
// A utility function to do union of two subsets
void Union(int parent[], int x, int y)
{
int xset = find(parent, x);
int yset = find(parent, y);
parent[xset] = yset;
}
// The main function to check whether a given graph contains
// cycle or not
int isCycle( struct Graph* graph )
{
// Allocate memory for creating V subsets
int *parent = (int*) malloc( graph->V * sizeof(int) );
// Initialize all subsets as single element sets
memset(parent, -1, sizeof(int) * graph->V);
// Iterate through all edges of graph, find subset of both
// vertices of every edge, if both subsets are same, then
// there is cycle in graph.
for(int i = 0; i < graph->E; ++i)
{
int x = find(parent, graph->edge[i].src);
int y = find(parent, graph->edge[i].dest);
if (x == y)
return 1;
Union(parent, x, y);
}
return 0;
}
// Driver program to test above functions
int main()
{
/* Let us create following graph
0
| \
|
\
1-----2 */
struct Graph* graph = createGraph(3, 3);
// add edge 0-1
graph->edge[0].src = 0;
graph->edge[0].dest = 1;
// add edge 1-2
graph->edge[1].src = 1;
graph->edge[1].dest = 2;
// add edge 0-2
graph->edge[2].src = 0;
graph->edge[2].dest = 2;
if (isCycle(graph))
printf( "Graph contains cycle" );
else
printf( "Graph doesn't contain cycle" );
return 0;

Java
// Java Program for union-find algorithm to detect cycle in a graph
import java.util.*;
import java.lang.*;
import java.io.*;
class Graph
{
int V, E;
// V-> no. of vertices & E->no.of edges
Edge edge[]; // /collection of all edges
class Edge
{
int src, dest;
};
// Creates a graph with V vertices and E edges
Graph(int v,int e)
{
V = v;
E = e;
edge = new Edge[E];
for (int i=0; i<e; ++i)
edge[i] = new Edge();
}
// A utility function to find the subset of an element i
int find(int parent[], int i)
{
if (parent[i] == -1)
return i;
return find(parent, parent[i]);
}
// A utility function to do union of two subsets
void Union(int parent[], int x, int y)
{
int xset = find(parent, x);
int yset = find(parent, y);
parent[xset] = yset;
}
// The main function to check whether a given graph
// contains cycle or not
int isCycle( Graph graph)
{
// Allocate memory for creating V subsets
int parent[] = new int[graph.V];
// Initialize all subsets as single element sets
for(int i=0; i<graph.V; ++i)
parent[i]=-1;
// Iterate through all edges of graph, find subset of both
// vertices of every edge, if both subsets are same, then
// there is cycle in graph.
for (int i = 0; i < graph.E; ++i)
{
int x = graph.find(parent, graph.edge[i].src);
int y = graph.find(parent, graph.edge[i].dest);
if (x == y)
return 1;
graph.Union(parent, x, y);
}
return 0;
}
// Driver Method
public static void main (String[] args)
{
/* Let us create following graph
0
| \
|
\

1-----2 */
Graph graph = new Graph(3,3);
// add edge 0-1
graph.edge[0].src = 0;
graph.edge[0].dest = 1;
// add edge 1-2
graph.edge[1].src = 1;
graph.edge[1].dest = 2;
// add edge 0-2
graph.edge[2].src = 0;
graph.edge[2].dest = 2;
if (graph.isCycle(graph)==1)
System.out.println( "Graph contains cycle" );
else
System.out.println( "Graph doesn't contain cycle" );
}
}

Graph contains cycle

Note that the implementation of union() and find() is naive and takes O(n) time in worst case. These methods can be improved to O(Logn) using
Union by Rank or Height. We will soon be discussing Union by Rank in a separate post.

Detect cycle in an undirected graph

Given an undirected graph, how to check if there is a cycle in the graph? For example, the following graph has a cycle 1-0-2-1.

We have discussed cycle detection for directed graph. We have also discussed a union-find algorithm for cycle detection in undirected graphs. The
time complexity of the union-find algorithm is O(ELogV). Like directed graphs, we can use DFS to detect cycle in an undirected graph in O(V+E)
time. We do a DFS traversal of the given graph. For every visited vertex v, if there is an adjacent u such that u is already visited and u is not parent
of v, then there is a cycle in graph. If we dont find such an adjacent for any vertex, we say that there is no cycle. The assumption of this approach
is that there are no parallel edges between any two vertices.

C++
// A C++ Program to detect cycle in an undirected graph
#include<iostream>
#include <list>
#include <limits.h>
using namespace std;
// Class for an undirected graph
class Graph
{
int V;
// No. of vertices
list<int> *adj;
// Pointer to an array containing adjacency lists
bool isCyclicUtil(int v, bool visited[], int parent);
public:
Graph(int V); // Constructor
void addEdge(int v, int w); // to add an edge to graph
bool isCyclic(); // returns true if there is a cycle
};
Graph::Graph(int V)
{
this->V = V;
adj = new list<int>[V];
}
void Graph::addEdge(int v, int w)
{
adj[v].push_back(w); // Add w to vs list.
adj[w].push_back(v); // Add v to ws list.
}
// A recursive function that uses visited[] and parent to detect
// cycle in subgraph reachable from vertex v.
bool Graph::isCyclicUtil(int v, bool visited[], int parent)
{
// Mark the current node as visited
visited[v] = true;
// Recur for all the vertices adjacent to this vertex
list<int>::iterator i;
for (i = adj[v].begin(); i != adj[v].end(); ++i)
{
// If an adjacent is not visited, then recur for that adjacent
if (!visited[*i])
{
if (isCyclicUtil(*i, visited, v))
return true;
}
// If an adjacent is visited and not parent of current vertex,
// then there is a cycle.
else if (*i != parent)
return true;
}
return false;
}
// Returns true if the graph contains a cycle, else false.

bool Graph::isCyclic()
{
// Mark all the vertices as not visited and not part of recursion
// stack
bool *visited = new bool[V];
for (int i = 0; i < V; i++)
visited[i] = false;
// Call the recursive helper function to detect cycle in different
// DFS trees
for (int u = 0; u < V; u++)
if (!visited[u]) // Don't recur for u if it is already visited
if (isCyclicUtil(u, visited, -1))
return true;
return false;
}
// Driver program to test above functions
int main()
{
Graph g1(5);
g1.addEdge(1, 0);
g1.addEdge(0, 2);
g1.addEdge(2, 0);
g1.addEdge(0, 3);
g1.addEdge(3, 4);
g1.isCyclic()? cout << "Graph contains cycle\n":
cout << "Graph doesn't contain cycle\n";
Graph g2(3);
g2.addEdge(0, 1);
g2.addEdge(1, 2);
g2.isCyclic()? cout << "Graph contains cycle\n":
cout << "Graph doesn't contain cycle\n";
return 0;
}

Java
// A Java Program to detect cycle in an undirected graph
import java.io.*;
import java.util.*;
// This class represents a directed graph using adjacency list
// representation
class Graph
{
private int V; // No. of vertices
private LinkedList<Integer> adj[]; // Adjacency List Represntation
// Constructor
Graph(int v) {
V = v;
adj = new LinkedList[v];
for(int i=0; i<v; ++i)
adj[i] = new LinkedList();
}
// Function to add an edge into the graph
void addEdge(int v,int w) {
adj[v].add(w);
adj[w].add(v);
}
// A recursive function that uses visited[] and parent to detect
// cycle in subgraph reachable from vertex v.
Boolean isCyclicUtil(int v, Boolean visited[], int parent)
{
// Mark the current node as visited
visited[v] = true;
Integer i;
// Recur for all the vertices adjacent to this vertex
Iterator<Integer> it = adj[v].iterator();
while (it.hasNext())
{
i = it.next();

// If an adjacent is not visited, then recur for that

// adjacent
if (!visited[i])
{
if (isCyclicUtil(i, visited, v))
return true;
}
// If an adjacent is visited and not parent of current
// vertex, then there is a cycle.
else if (i != parent)
return true;
}
return false;
}
// Returns true if the graph contains a cycle, else false.
Boolean isCyclic()
{
// Mark all the vertices as not visited and not part of
// recursion stack
Boolean visited[] = new Boolean[V];
for (int i = 0; i < V; i++)
visited[i] = false;
// Call the recursive helper function to detect cycle in
// different DFS trees
for (int u = 0; u < V; u++)
if (!visited[u]) // Don't recur for u if already visited
if (isCyclicUtil(u, visited, -1))
return true;
return false;
}
// Driver method to test above methods
public static void main(String args[])
{
// Create a graph given in the above diagram
Graph g1 = new Graph(5);
g1.addEdge(1, 0);
g1.addEdge(0, 2);
g1.addEdge(2, 0);
g1.addEdge(0, 3);
g1.addEdge(3, 4);
if (g1.isCyclic())
System.out.println("Graph contains cycle");
else
System.out.println("Graph doesn't contains cycle");
Graph g2 = new Graph(3);
g2.addEdge(0, 1);
g2.addEdge(1, 2);
if (g2.isCyclic())
System.out.println("Graph contains cycle");
else
System.out.println("Graph doesn't contains cycle");
}
}
// This code is contributed by Aakash Hasija

Output:
Graph contains cycle
Graph doesn't contain cycle

Time Complexity: The program does a simple DFS Traversal of graph and graph is represented using adjacency list. So the time complexity is
O(V+E)
Exercise: Can we use BFS to detect cycle in an undirected graph in O(V+E) time? What about directed graphs?

Longest Path in a Directed Acyclic Graph

Given a Weighted Directed Acyclic Graph (DAG) and a source vertex s in it, find the longest distances from s to all other vertices in the given
graph.
The longest path problem for a general graph is not as easy as the shortest path problem because the longest path problem doesnt have optimal
substructure property. In fact, the Longest Path problem is NP-Hard for a general graph. However, the longest path problem has a linear time
solution for directed acyclic graphs. The idea is similar to linear time solution for shortest path in a directed acyclic graph., we use Tological
Sorting.
We initialize distances to all vertices as minus infinite and distance to source as 0, then we find a topological sorting of the graph. Topological
Sorting of a graph represents a linear ordering of the graph (See below, figure (b) is a linear representation of figure (a) ). Once we have
topological order (or linear representation), we one by one process all vertices in topological order. For every vertex being processed, we update
distances of its adjacent using distance of current vertex.
Following figure shows step by step process of finding longest paths.

Following is complete algorithm for finding longest distances.

1) Initialize dist[] = {NINF, NINF, .} and dist[s] = 0 where s is the source vertex. Here NINF means negative infinite.
2) Create a toplogical order of all vertices.
3) Do following for every vertex u in topological order.
..Do following for every adjacent vertex v of u
if (dist[v] < dist[u] + weight(u, v)) dist[v] = dist[u] + weight(u, v) Following is C++ implementation of the above algorithm.
// A C++ program to find single source longest distances in a DAG
#include <iostream>
#include <list>
#include <stack>
#include <limits.h>
#define NINF INT_MIN
using namespace std;
// Graph is represented using adjacency list. Every node of adjacency list
// contains vertex number of the vertex to which edge connects. It also
// contains weight of the edge
class AdjListNode
{
int v;
int weight;

public:
AdjListNode(int _v, int _w) { v = _v; weight = _w;}
int getV()
{ return v; }
int getWeight() { return weight; }
};
// Class to represent a graph using adjacency list representation
class Graph
{
int V;
// No. of vertices
// Pointer to an array containing adjacency lists
list<AdjListNode> *adj;
// A function used by longestPath
void topologicalSortUtil(int v, bool visited[], stack<int> &Stack);
public:
Graph(int V); // Constructor
// function to add an edge to graph
void addEdge(int u, int v, int weight);
// Finds longest distances from given source vertex
void longestPath(int s);
};
Graph::Graph(int V) // Constructor
{
this->V = V;
adj = new list<AdjListNode>[V];
}
void Graph::addEdge(int u, int v, int weight)
{
AdjListNode node(v, weight);
adj[u].push_back(node); // Add v to us list
}
// A recursive function used by longestPath. See below link for details
// http://www.geeksforgeeks.org/topological-sorting/
void Graph::topologicalSortUtil(int v, bool visited[], stack<int> &Stack)
{
// Mark the current node as visited
visited[v] = true;
// Recur for all the vertices adjacent to this vertex
list<AdjListNode>::iterator i;
for (i = adj[v].begin(); i != adj[v].end(); ++i)
{
AdjListNode node = *i;
if (!visited[node.getV()])
topologicalSortUtil(node.getV(), visited, Stack);
}
// Push current vertex to stack which stores topological sort
Stack.push(v);
}
// The function to find longest distances from a given vertex. It uses
// recursive topologicalSortUtil() to get topological sorting.
void Graph::longestPath(int s)
{
stack<int> Stack;
int dist[V];
// Mark all the vertices as not visited
bool *visited = new bool[V];
for (int i = 0; i < V; i++)
visited[i] = false;
// Call the recursive helper function to store Topological Sort
// starting from all vertices one by one
for (int i = 0; i < V; i++)
if (visited[i] == false)
topologicalSortUtil(i, visited, Stack);
// Initialize distances to all vertices as infinite and distance
// to source as 0
for (int i = 0; i < V; i++)
dist[i] = NINF;
dist[s] = 0;

// Process vertices in topological order

while (Stack.empty() == false)
{
// Get the next vertex from topological order
int u = Stack.top();
Stack.pop();
// Update distances of all adjacent vertices
list<AdjListNode>::iterator i;
if (dist[u] != NINF)
{
for (i = adj[u].begin(); i != adj[u].end(); ++i)
if (dist[i->getV()] < dist[u] + i->getWeight())
dist[i->getV()] = dist[u] + i->getWeight();
}
}
// Print the calculated longest distances
for (int i = 0; i < V; i++)
(dist[i] == NINF)? cout << "INF ": cout << dist[i] << " ";
}
// Driver program to test above functions
int main()
{
// Create a graph given in the above diagram. Here vertex numbers are
// 0, 1, 2, 3, 4, 5 with following mappings:
// 0=r, 1=s, 2=t, 3=x, 4=y, 5=z
Graph g(6);
g.addEdge(0, 1, 5);
g.addEdge(0, 2, 3);
g.addEdge(1, 3, 6);
g.addEdge(1, 2, 2);
g.addEdge(2, 4, 4);
g.addEdge(2, 5, 2);
g.addEdge(2, 3, 7);
g.addEdge(3, 5, 1);
g.addEdge(3, 4, -1);
g.addEdge(4, 5, -2);
int s = 1;
cout << "Following are longest distances from source vertex " << s <<" \n";
g.longestPath(s);
return 0;
}

Output:
Following are longest distances from source vertex 1
INF 0 2 9 8 10

Time Complexity: Time complexity of topological sorting is O(V+E). After finding topological order, the algorithm process all vertices and for
every vertex, it runs a loop for all adjacent vertices. Total adjacent vertices in a graph is O(E). So the inner loop runs O(V+E) times. Therefore,
overall time complexity of this algorithm is O(V+E).
Exercise: The above solution print longest distances, extend the code to print paths also.

Topological Sorting
Topological sorting for Directed Acyclic Graph (DAG) is a linear ordering of vertices such that for every directed edge uv, vertex u comes before
v in the ordering. Topological Sorting for a graph is not possible if the graph is not a DAG.
For example, a topological sorting of the following graph is 5 4 2 3 1 0?. There can be more than one topological sorting for a graph. For example,
another topological sorting of the following graph is 4 5 2 3 1 0?. The first vertex in topological sorting is always a vertex with in-degree as 0 (a
vertex with no in-coming edges).

Topological Sorting vs Depth First Traversal (DFS):

In DFS, we print a vertex and then recursively call DFS for its adjacent vertices. In topological sorting, we need to print a vertex before its
adjacent vertices. For example, in the given graph, the vertex 5 should be printed before vertex 0, but unlike DFS, the vertex 4 should also be
printed before vertex 0. So Topological sorting is different from DFS. For example, a DFS of the above graph is 5 2 3 1 0 4?, but it is not a
topological sorting
Algorithm to find Topological Sorting:
We recommend to first see implementation of DFS here. We can modify DFS to find Topological Sorting of a graph. In DFS, we start from a
vertex, we first print it and then recursively call DFS for its adjacent vertices. In topological sorting, we use a temporary stack. We dont print the
vertex immediately, we first recursively call topological sorting for all its adjacent vertices, then push it to a stack. Finally, print contents of stack.
Note that a vertex is pushed to stack only when all of its adjacent vertices (and their adjacent vertices and so on) are already in stack.
Following are C++ and Java implementations of topological sorting. Please see the code for Depth First Traversal for a disconnected Graph and
note the differences between the second code given there and the below code.

C++
// A C++ program to print topological sorting of a DAG
#include<iostream>
#include <list>
#include <stack>
using namespace std;
// Class to represent a graph
class Graph
{
int V;
// No. of vertices'
// Pointer to an array containing adjacency listsList
list<int> *adj;
// A function used by topologicalSort
void topologicalSortUtil(int v, bool visited[], stack<int> &Stack);
public:
Graph(int V); // Constructor
// function to add an edge to graph
void addEdge(int v, int w);
// prints a Topological Sort of the complete graph
void topologicalSort();
};
Graph::Graph(int V)
{
this->V = V;
adj = new list<int>[V];
}
void Graph::addEdge(int v, int w)
{

adj[v].push_back(w); // Add w to vs list.

}
// A recursive function used by topologicalSort
void Graph::topologicalSortUtil(int v, bool visited[],
stack<int> &Stack)
{
// Mark the current node as visited.
visited[v] = true;
// Recur for all the vertices adjacent to this vertex
list<int>::iterator i;
for (i = adj[v].begin(); i != adj[v].end(); ++i)
if (!visited[*i])
topologicalSortUtil(*i, visited, Stack);
// Push current vertex to stack which stores result
Stack.push(v);
}
// The function to do Topological Sort. It uses recursive
// topologicalSortUtil()
void Graph::topologicalSort()
{
stack<int> Stack;
// Mark all the vertices as not visited
bool *visited = new bool[V];
for (int i = 0; i < V; i++)
visited[i] = false;
// Call the recursive helper function to store Topological
// Sort starting from all vertices one by one
for (int i = 0; i < V; i++)
if (visited[i] == false)
topologicalSortUtil(i, visited, Stack);
// Print contents of stack
while (Stack.empty() == false)
{
cout << Stack.top() << " ";
Stack.pop();
}
}
// Driver program to test above functions
int main()
{
// Create a graph given in the above diagram
Graph g(6);
g.addEdge(5, 2);
g.addEdge(5, 0);
g.addEdge(4, 0);
g.addEdge(4, 1);
g.addEdge(2, 3);
g.addEdge(3, 1);
cout << "Following is a Topological Sort of the given graph \n";
g.topologicalSort();
return 0;
}

Java
// A Java program to print topological sorting of a DAG
import java.io.*;
import java.util.*;
// This class represents a directed graph using adjacency
// list representation
class Graph
{
private int V; // No. of vertices
private LinkedList<Integer> adj[]; // Adjacency List
//Constructor
Graph(int v)
{
V = v;

adj = new LinkedList[v];

for (int i=0; i<v; ++i)
adj[i] = new LinkedList();
}
// Function to add an edge into the graph
void addEdge(int v,int w) { adj[v].add(w); }
// A recursive function used by topologicalSort
void topologicalSortUtil(int v, Boolean visited[],Stack stack)
{
// Mark the current node as visited.
visited[v] = true;
Integer i;
// Recur for all the vertices adjacent to this vertex
Iterator<Integer> it = adj[v].iterator();
while (it.hasNext())
{
i = it.next();
if (!visited[i])
topologicalSortUtil(i, visited, stack);
}
// Push current vertex to stack which stores result
stack.push(new Integer(v));
}
// The function to do Topological Sort. It uses recursive
// topologicalSortUtil()
void topologicalSort()
{
Stack stack = new Stack();
// Mark all the vertices as not visited
Boolean visited[] = new Boolean[V];
for (int i = 0; i < V; i++)
visited[i] = false;
// Call the recursive helper function to store Topological
// Sort starting from all vertices one by one
for (int i = 0; i < V; i++)
if (visited[i] == false)
topologicalSortUtil(i, visited, stack);
// Print contents of stack
while (stack.empty()==false)
System.out.print(stack.pop() + " ");
}
// Driver method
public static void main(String args[])
{
// Create a graph given in the above diagram
Graph g = new Graph(6);
g.addEdge(5, 2);
g.addEdge(5, 0);
g.addEdge(4, 0);
g.addEdge(4, 1);
g.addEdge(2, 3);
g.addEdge(3, 1);
System.out.println("Following is a Topological " +
"sort of the given graph");
g.topologicalSort();
}
}
// This code is contributed by Aakash Hasija

Following is a Topological Sort of the given graph

5 4 2 3 1 0

Time Complexity: The above algorithm is simply DFS with an extra stack. So time complexity is same as DFS which is O(V+E).
Applications:
Topological Sorting is mainly used for scheduling jobs from the given dependencies among jobs. In computer science, applications of this type
arise in instruction scheduling, ordering of formula cell evaluation when recomputing formula values in spreadsheets, logic synthesis, determining the
order of compilation tasks to perform in makefiles, data serialization, and resolving symbol dependencies in linkers [2].

References:
http://www.personal.kent.edu/~rmuhamma/Algorithms/MyAlgorithms/GraphAlgor/topoSort.htm
http://en.wikipedia.org/wiki/Topological_sorting

Check whether a given graph is Bipartite or not

A Bipartite Graph is a graph whose vertices can be divided into two independent sets, U and V such that every edge (u, v) either connects a
vertex from U to V or a vertex from V to U. In other words, for every edge (u, v), either u belongs to U and v to V, or u belongs to V and v to U.
We can also say that there is no edge that connects vertices of same set.

A bipartite graph is possible if the graph coloring is possible using two colors such that vertices in a set are colored with the same color. Note that
it is possible to color a cycle graph with even cycle using two colors. For example, see the following graph.

It is not possible to color a cycle graph with odd cycle using two colors.

Algorithm to check if a graph is Bipartite:

One approach is to check whether the graph is 2-colorable or not using backtracking algorithm m coloring problem.
Following is a simple algorithm to find out whether a given graph is Birpartite or not using Breadth First Search (BFS).
1. Assign RED color to the source vertex (putting into set U).
2. Color all the neighbors with BLUE color (putting into set V).
3. Color all neighbors neighbor with RED color (putting into set U).
4. This way, assign color to all vertices such that it satisfies all the constraints of m way coloring problem where m = 2.
5. While assigning colors, if we find a neighbor which is colored with same color as current vertex, then the graph cannot be colored with 2
vertices (or graph is not Bipartite)

C++
// C++ program to find out whether a given graph is Bipartite or not
#include <iostream>
#include <queue>
#define V 4
using namespace std;
// This function returns true if graph G[V][V] is Bipartite, else false
bool isBipartite(int G[][V], int src)
{
// Create a color array to store colors assigned to all veritces. Vertex
// number is used as index in this array. The value '-1' of colorArr[i]
// is used to indicate that no color is assigned to vertex 'i'. The value
// 1 is used to indicate first color is assigned and value 0 indicates
// second color is assigned.
int colorArr[V];
for (int i = 0; i < V; ++i)
colorArr[i] = -1;
// Assign first color to source
colorArr[src] = 1;
// Create a queue (FIFO) of vertex numbers and enqueue source vertex

// for BFS traversal

queue <int> q;
q.push(src);
// Run while there are vertices in queue (Similar to BFS)
while (!q.empty())
{
// Dequeue a vertex from queue ( Refer http://goo.gl/35oz8 )
int u = q.front();
q.pop();
// Find all non-colored adjacent vertices
for (int v = 0; v < V; ++v)
{
// An edge from u to v exists and destination v is not colored
if (G[u][v] && colorArr[v] == -1)
{
// Assign alternate color to this adjacent v of u
colorArr[v] = 1 - colorArr[u];
q.push(v);
}
// An edge from u to v exists and destination v is colored with
// same color as u
else if (G[u][v] && colorArr[v] == colorArr[u])
return false;
}
}
// If we reach here, then all adjacent vertices can be colored with
// alternate color
return true;
}
// Driver program to test above function
int main()
{
int G[][V] = {{0, 1, 0, 1},
{1, 0, 1, 0},
{0, 1, 0, 1},
{1, 0, 1, 0}
};
isBipartite(G, 0) ? cout << "Yes" : cout << "No";
return 0;
}

Java
// Java program to find out whether a given graph is Bipartite or not
import java.util.*;
import java.lang.*;
import java.io.*;
class Bipartite
{
final static int V = 4; // No. of Vertices
// This function returns true if graph G[V][V] is Bipartite, else false
boolean isBipartite(int G[][],int src)
{
// Create a color array to store colors assigned to all veritces.
// Vertex number is used as index in this array. The value '-1'
// of colorArr[i] is used to indicate that no color is assigned
// to vertex 'i'. The value 1 is used to indicate first color
// is assigned and value 0 indicates second color is assigned.
int colorArr[] = new int[V];
for (int i=0; i<V; ++i)
colorArr[i] = -1;
// Assign first color to source
colorArr[src] = 1;
// Create a queue (FIFO) of vertex numbers and enqueue
// source vertex for BFS traversal
LinkedList<Integer>q = new LinkedList<Integer>();
q.add(src);
// Run while there are vertices in queue (Similar to BFS)
while (q.size() != 0)

{
// Dequeue a vertex from queue
int u = q.poll();
// Find all non-colored adjacent vertices
for (int v=0; v<V; ++v)
{
// An edge from u to v exists and destination v is
// not colored
if (G[u][v]==1 && colorArr[v]==-1)
{
// Assign alternate color to this adjacent v of u
colorArr[v] = 1-colorArr[u];
q.add(v);
}
// An edge from u to v exists and destination v is
// colored with same color as u
else if (G[u][v]==1 && colorArr[v]==colorArr[u])
return false;
}
}
// If we reach here, then all adjacent vertices can
// be colored with alternate color
return true;
}
// Driver program to test above function
public static void main (String[] args)
{
int G[][] = {{0, 1, 0, 1},
{1, 0, 1, 0},
{0, 1, 0, 1},
{1, 0, 1, 0}
};
Bipartite b = new Bipartite();
if (b.isBipartite(G, 0))
System.out.println("Yes");
else
System.out.println("No");
}
}
// Contributed by Aakash Hasija

Yes

Refer this for C implementation of the same.

Time Complexity of the above approach is same as that Breadth First Search. In above implementation is O(V^2) where V is number of vertices.
If graph is represented using adjacency list, then the complexity becomes O(V+E).
Exercise:
1. Can DFS algorithm be used to check the bipartite-ness of a graph? If yes, how?
2. The above algorithm works if the graph is strongly connected. Extend above code to work for graph with more than one component.
References:
http://en.wikipedia.org/wiki/Graph_coloring
http://en.wikipedia.org/wiki/Bipartite_graph

Snake and Ladder Problem

Given a snake and ladder board, find the minimum number of dice throws required to reach the destination or last cell from source or 1st cell.
Basically, the player has total control over outcome of dice throw and wants to find out minimum number of throws required to reach last cell.
If the player reaches a cell which is base of a ladder, the player has to climb up that ladder and if reaches a cell is mouth of the snake, has to go
down to the tail of snake without a dice throw.

For example consider the board shown on right side (taken from here), the minimum number of dice throws required to reach cell 30 from cell 1 is
3. Following are steps.
a) First throw two on dice to reach cell number 3 and then ladder to reach 22
b) Then throw 6 to reach 28.
c) Finally through 2 to reach 30.
There can be other solutions as well like (2, 2, 6), (2, 4, 4), (2, 3, 5).. etc.
Following is C++ implementation of the above idea. The input is represented by two things, first is N which is number of cells in the given board,
second is an array move[0N-1] of size N. An entry move[i] is -1 if there is no snake and no ladder from i, otherwise move[i] contains index of
destination cell for the snake or the ladder at i.
// C++ program to find minimum number of dice throws required to
// reach last cell from first cell of a given snake and ladder
// board
#include<iostream>
#include <queue>
using namespace std;
// An entry in queue used in BFS
struct queueEntry
{
int v;
// Vertex number
int dist; // Distance of this vertex from source
};
// This function returns minimum number of dice throws required to
// Reach last cell from 0'th cell in a snake and ladder game.
// move[] is an array of size N where N is no. of cells on board
// If there is no snake or ladder from cell i, then move[i] is -1
// Otherwise move[i] contains cell to which snake or ladder at i
// takes to.
int getMinDiceThrows(int move[], int N)
{
// The graph has N vertices. Mark all the vertices as
// not visited
bool *visited = new bool[N];
for (int i = 0; i < N; i++)
visited[i] = false;
// Create a queue for BFS
queue<queueEntry> q;
// Mark the node 0 as visited and enqueue it.
visited[0] = true;
queueEntry s = {0, 0}; // distance of 0't vertex is also 0
q.push(s); // Enqueue 0'th vertex
// Do a BFS starting from vertex at index 0
queueEntry qe; // A queue entry (qe)
while (!q.empty())
{
qe = q.front();
int v = qe.v; // vertex no. of queue entry

// If front vertex is the destination vertex,

// we are done
if (v == N-1)
break;
// Otherwise dequeue the front vertex and enqueue
// its adjacent vertices (or cell numbers reachable
// through a dice throw)
q.pop();
for (int j=v+1; j<=(v+6) && j<N; ++j)
{
// If this cell is already visited, then ignore
if (!visited[j])
{
// Otherwise calculate its distance and mark it
// as visited
queueEntry a;
a.dist = (qe.dist + 1);
visited[j] = true;
//
//
//
if

Check if there a snake or ladder at 'j'

then tail of snake or top of ladder
become the adjacent of 'i'
(move[j] != -1)
a.v = move[j];
else
a.v = j;
q.push(a);
}
}
}
// We reach here when 'qe' has last vertex
// return the distance of vertex in 'qe'
return qe.dist;
}
// Driver program to test methods of graph class
int main()
{
// Let us construct the board given in above diagram
int N = 30;
int moves[N];
for (int i = 0; i<N; i++)
moves[i] = -1;
// Ladders
moves[2] = 21;
moves[4] = 7;
moves[10] = 25;
moves[19] = 28;
// Snakes
moves[26]
moves[20]
moves[16]
moves[18]

=
=
=
=

0;
8;
3;
6;

cout << "Min Dice throws required is " << getMinDiceThrows(moves, N);
return 0;
}

Output:
Min Dice throws required is 3

Time complexity of the above solution is O(N) as every cell is added and removed only once from queue. And a typical enqueue or dequeue
operation takes O(1) time.

Biconnected Components
A biconnected component is a maximal biconnected subgraph.
Biconnected Graph is already discussed here. In this article, we will see how to find biconnected component in a graph using algorithm by John
Hopcroft and Robert Tarjan.

In above graph, following are the biconnected components:

42 34 31 23 12
89
85 78 57
60 56 15 01
1011
Algorithm is based on Disc and Low Values discussed in Strongly Connected Components Article.
Idea is to store visited edges in a stack while DFS on a graph and keep looking for Articulation Points (highlighted in above figure). As soon as an
Articulation Point u is found, all edges visited while DFS from node u onwards will form one biconnected component. When DFS completes for
one connected component, all edges present in stack will form a biconnected component.
If there is no Articulation Point in graph, then graph is biconnected and so there will be one biconnected component which is the graph itself.

C++
// A C++ program to find biconnected components in a given undirected graph
#include<iostream>
#include <list>
#include <stack>
#define NIL -1
using namespace std;
int count = 0;
class Edge
{
public:
int u;
int v;
Edge(int u, int v);
};
Edge::Edge(int u, int v)
{
this->u = u;
this->v = v;
}
// A class that represents an directed graph
class Graph
{
int V;
// No. of vertices
int E;
// No. of edges
list<int> *adj;
// A dynamic array of adjacency lists
// A Recursive DFS based function used by BCC()
void BCCUtil(int u, int disc[], int low[],
list<Edge> *st, int parent[]);
public:
Graph(int V); // Constructor
void addEdge(int v, int w); // function to add an edge to graph
void BCC();
// prints strongly connected components
};
Graph::Graph(int V)
{
this->V = V;
this->E = 0;

adj = new list<int>[V];

}
void Graph::addEdge(int v, int w)
{
adj[v].push_back(w);
E++;
}
// A recursive function that finds and prints strongly connected
// components using DFS traversal
// u --> The vertex to be visited next
// disc[] --> Stores discovery times of visited vertices
// low[] -- >> earliest visited vertex (the vertex with minimum
//
discovery time) that can be reached from subtree
//
rooted with current vertex
// *st -- >> To store visited edges
void Graph::BCCUtil(int u, int disc[], int low[], list<Edge> *st,
int parent[])
{
// A static variable is used for simplicity, we can avoid use
// of static variable by passing a pointer.
static int time = 0;
// Initialize discovery time and low value
disc[u] = low[u] = ++time;
int children = 0;
// Go through all vertices adjacent to this
list<int>::iterator i;
for (i = adj[u].begin(); i != adj[u].end(); ++i)
{
int v = *i; // v is current adjacent of 'u'
// If v is not visited yet, then recur for it
if (disc[v] == -1)
{
children++;
parent[v] = u;
//store the edge in stack
st->push_back(Edge(u,v));
BCCUtil(v, disc, low, st, parent);
// Check if the subtree rooted with 'v' has a
// connection to one of the ancestors of 'u'
// Case 1 -- per Strongly Connected Components Article
low[u] = min(low[u], low[v]);
//If u is an articulation point,
//pop all edges from stack till u -- v
if( (disc[u] == 1 && children > 1) ||
(disc[u] > 1 && low[v] >= disc[u]) )
{
while(st->back().u != u || st->back().v != v)
{
cout << st->back().u << "--" << st->back().v << " ";
st->pop_back();
}
cout << st->back().u << "--" << st->back().v;
st->pop_back();
cout << endl;
count++;
}
}
// Update low value of 'u' only of 'v' is still in stack
// (i.e. it's a back edge, not cross edge).
// Case 2 -- per Strongly Connected Components Article
else if(v != parent[u] && disc[v] < low[u])
{
low[u] = min(low[u], disc[v]);
st->push_back(Edge(u,v));
}
}
}
// The function to do DFS traversal. It uses BCCUtil()
void Graph::BCC()
{
int *disc = new int[V];
int *low = new int[V];

int *parent = new int[V];

list<Edge> *st = new list<Edge>[E];
// Initialize disc and low, and parent arrays
for (int i = 0; i < V; i++)
{
disc[i] = NIL;
low[i] = NIL;
parent[i] = NIL;
}
for (int i = 0; i < V; i++)
{
if (disc[i] == NIL)
BCCUtil(i, disc, low, st, parent);
int j = 0;
//If stack is not empty, pop all edges from stack
while(st->size() > 0)
{
j = 1;
cout << st->back().u << "--" << st->back().v << " ";
st->pop_back();
}
if(j == 1)
{
cout << endl;
count++;
}
}
}
// Driver program to test above function
int main()
{
Graph g(12);
g.addEdge(0,1);g.addEdge(1,0);
g.addEdge(1,2);g.addEdge(2,1);
g.addEdge(1,3);g.addEdge(3,1);
g.addEdge(2,3);g.addEdge(3,2);
g.addEdge(2,4);g.addEdge(4,2);
g.addEdge(3,4);g.addEdge(4,3);
g.addEdge(1,5);g.addEdge(5,1);
g.addEdge(0,6);g.addEdge(6,0);
g.addEdge(5,6);g.addEdge(6,5);
g.addEdge(5,7);g.addEdge(7,5);
g.addEdge(5,8);g.addEdge(8,5);
g.addEdge(7,8);g.addEdge(8,7);
g.addEdge(8,9);g.addEdge(9,8);
g.addEdge(10,11);g.addEdge(11,10);
g.BCC();
cout << "Above are " << count << " biconnected components in graph";
return 0;
}

Java
// A Java program to find biconnected components in a given
// undirected graph
import java.io.*;
import java.util.*;
// This class represents a directed graph using adjacency
// list representation
class Graph
{
private int V, E; // No. of vertices & Edges respectively
private LinkedList<Integer> adj[]; // Adjacency List
// Count is number of biconnected components. time is
// used to find discovery times
static int count = 0, time = 0;
class Edge
{
int u;
int v;
Edge(int u, int v)
{
this.u = u;

this.v = v;
}
};
//Constructor
Graph(int v)
{
V = v;
E = 0;
adj = new LinkedList[v];
for (int i=0; i<v; ++i)
adj[i] = new LinkedList();
}
//Function to add an edge into the graph
void addEdge(int v,int w)
{
adj[v].add(w);
E++;
}
// A recursive function that finds and prints strongly connected
// components using DFS traversal
// u --> The vertex to be visited next
// disc[] --> Stores discovery times of visited vertices
// low[] -- >> earliest visited vertex (the vertex with minimum
//
discovery time) that can be reached from subtree
//
rooted with current vertex
// *st -- >> To store visited edges
void BCCUtil(int u, int disc[], int low[], LinkedList<Edge>st,
int parent[])
{
// Initialize discovery time and low value
disc[u] = low[u] = ++time;
int children = 0;
// Go through all vertices adjacent to this
Iterator<Integer> it = adj[u].iterator();
while (it.hasNext())
{
int v = it.next(); // v is current adjacent of 'u'
// If v is not visited yet, then recur for it
if (disc[v] == -1)
{
children++;
parent[v] = u;
// store the edge in stack
st.add(new Edge(u,v));
BCCUtil(v, disc, low, st, parent);
//
//
//
if

Check if the subtree rooted with 'v' has a

connection to one of the ancestors of 'u'
Case 1 -- per Strongly Connected Components Article
(low[u] > low[v])
low[u] = low[v];

// If u is an articulation point,
// pop all edges from stack till u -- v
if ( (disc[u] == 1 && children > 1) ||
(disc[u] > 1 && low[v] >= disc[u]) )
{
while (st.getLast().u != u || st.getLast().v != v)
{
System.out.print(st.getLast().u + "--" +
st.getLast().v + " ");
st.removeLast();
}
System.out.println(st.getLast().u + "--" +
st.getLast().v + " ");
st.removeLast();
count++;
}
}
// Update low value of 'u' only of 'v' is still in stack
// (i.e. it's a back edge, not cross edge).
// Case 2 -- per Strongly Connected Components Article

else if (v != parent[u] && disc[v] < low[u])

{
if (low[u]>disc[v])
low[u]=disc[v];
st.add(new Edge(u,v));
}
}
}
// The function to do DFS traversal. It uses BCCUtil()
void BCC()
{
int disc[] = new int[V];
int low[] = new int[V];
int parent[] = new int[V];
LinkedList<Edge> st = new LinkedList<Edge>();
// Initialize disc and low, and parent arrays
for (int i = 0; i < V; i++)
{
disc[i] = -1;
low[i] = -1;
parent[i] = -1;
}
for (int i = 0; i < V; i++)
{
if (disc[i] == -1)
BCCUtil(i, disc, low, st, parent);
int j = 0;
// If stack is not empty, pop all edges from stack
while (st.size() > 0)
{
j = 1;
System.out.print(st.getLast().u + "--" +
st.getLast().v + " ");
st.removeLast();
}
if (j == 1)
{
System.out.println();
count++;
}
}
}
public static void main(String args[])
{
Graph g = new Graph(12);
g.addEdge(0,1);
g.addEdge(1,0);
g.addEdge(1,2);
g.addEdge(2,1);
g.addEdge(1,3);
g.addEdge(3,1);
g.addEdge(2,3);
g.addEdge(3,2);
g.addEdge(2,4);
g.addEdge(4,2);
g.addEdge(3,4);
g.addEdge(4,3);
g.addEdge(1,5);
g.addEdge(5,1);
g.addEdge(0,6);
g.addEdge(6,0);
g.addEdge(5,6);
g.addEdge(6,5);
g.addEdge(5,7);
g.addEdge(7,5);
g.addEdge(5,8);
g.addEdge(8,5);
g.addEdge(7,8);
g.addEdge(8,7);
g.addEdge(8,9);
g.addEdge(9,8);
g.addEdge(10,11);
g.addEdge(11,10);
g.BCC();

System.out.println("Above are " + g.count +

" biconnected components in graph");
}
}
// This code is contributed by Aakash Hasija

Output:
4--2 3--4
8--9
8--5 7--8
6--0 5--6
10--11
Above are

3--1 2--3 1--2

5--7
1--5 0--1
5 biconnected components in graph

Check if a given graph is tree or not

Write a function that returns true if a given undirected graph is tree and false otherwise. For example, the following graph is a tree.

But the following graph is not a tree.

An undirected graph is tree if it has following properties.

1) There is no cycle.
2) The graph is connected.
For an undirected graph we can either use BFS or DFS to detect above two properties.
How to detect cycle in an undirected graph?
We can either use BFS or DFS. For every visited vertex v, if there is an adjacent u such that u is already visited and u is not parent of v, then there
is a cycle in graph. If we dont find such an adjacent for any vertex, we say that there is no cycle (See Detect cycle in an undirected graph for more
details).
How to check for connectivity?
Since the graph is undirected, we can start BFS or DFS from any vertex and check if all vertices are reachable or not. If all vertices are reachable,
then graph is connected, otherwise not.

C++
// A C++ Program to check whether a graph is tree or not
#include<iostream>
#include <list>
#include <limits.h>
using namespace std;
// Class for an undirected graph
class Graph
{
int V;
// No. of vertices
list<int> *adj; // Pointer to an array for adjacency lists
bool isCyclicUtil(int v, bool visited[], int parent);
public:
Graph(int V); // Constructor
void addEdge(int v, int w); // to add an edge to graph
bool isTree(); // returns true if graph is tree
};
Graph::Graph(int V)
{
this->V = V;
adj = new list<int>[V];
}
void Graph::addEdge(int v, int w)
{
adj[v].push_back(w); // Add w to vs list.
adj[w].push_back(v); // Add v to ws list.
}
// A recursive function that uses visited[] and parent to
// detect cycle in subgraph reachable from vertex v.
bool Graph::isCyclicUtil(int v, bool visited[], int parent)

{
// Mark the current node as visited
visited[v] = true;
// Recur for all the vertices adjacent to this vertex
list<int>::iterator i;
for (i = adj[v].begin(); i != adj[v].end(); ++i)
{
// If an adjacent is not visited, then recur for
// that adjacent
if (!visited[*i])
{
if (isCyclicUtil(*i, visited, v))
return true;
}
// If an adjacent is visited and not parent of current
// vertex, then there is a cycle.
else if (*i != parent)
return true;
}
return false;
}
// Returns true if the graph is a tree, else false.
bool Graph::isTree()
{
// Mark all the vertices as not visited and not part of
// recursion stack
bool *visited = new bool[V];
for (int i = 0; i < V; i++)
visited[i] = false;
//
//
//
//
if

The call to isCyclicUtil serves multiple purposes.

It returns true if graph reachable from vertex 0
is cyclcic. It also marks all vertices reachable
from 0.
(isCyclicUtil(0, visited, -1))
return false;

// If we find a vertex which is not reachable from 0

// (not marked by isCyclicUtil(), then we return false
for (int u = 0; u < V; u++)
if (!visited[u])
return false;
return true;
}
// Driver program to test above functions
int main()
{
Graph g1(5);
g1.addEdge(1, 0);
g1.addEdge(0, 2);
g1.addEdge(0, 3);
g1.addEdge(3, 4);
g1.isTree()? cout << "Graph is Tree\n":
cout << "Graph is not Tree\n";
Graph g2(5);
g2.addEdge(1, 0);
g2.addEdge(0, 2);
g2.addEdge(2, 1);
g2.addEdge(0, 3);
g2.addEdge(3, 4);
g2.isTree()? cout << "Graph is Tree\n":
cout << "Graph is not Tree\n";
return 0;
}

Java
// A Java Program to check whether a graph is tree or not
import java.io.*;
import java.util.*;
// This class represents a directed graph using adjacency
// list representation

class Graph
{
private int V; // No. of vertices
private LinkedList<Integer> adj[]; //Adjacency List
// Constructor
Graph(int v)
{
V = v;
adj = new LinkedList[v];
for (int i=0; i<v; ++i)
adj[i] = new LinkedList();
}
// Function to add an edge into the graph
void addEdge(int v,int w)
{
adj[v].add(w);
adj[w].add(v);
}
// A recursive function that uses visited[] and parent
// to detect cycle in subgraph reachable from vertex v.
Boolean isCyclicUtil(int v, Boolean visited[], int parent)
{
// Mark the current node as visited
visited[v] = true;
Integer i;
// Recur for all the vertices adjacent to this vertex
Iterator<Integer> it = adj[v].iterator();
while (it.hasNext())
{
i = it.next();
// If an adjacent is not visited, then recur for
// that adjacent
if (!visited[i])
{
if (isCyclicUtil(i, visited, v))
return true;
}
// If an adjacent is visited and not parent of
// current vertex, then there is a cycle.
else if (i != parent)
return true;
}
return false;
}
// Returns true if the graph is a tree, else false.
Boolean isTree()
{
// Mark all the vertices as not visited and not part
// of recursion stack
Boolean visited[] = new Boolean[V];
for (int i = 0; i < V; i++)
visited[i] = false;
//
//
//
//
if

The call to isCyclicUtil serves multiple purposes

It returns true if graph reachable from vertex 0
is cyclcic. It also marks all vertices reachable
from 0.
(isCyclicUtil(0, visited, -1))
return false;

// If we find a vertex which is not reachable from 0

// (not marked by isCyclicUtil(), then we return false
for (int u = 0; u < V; u++)
if (!visited[u])
return false;
return true;
}
// Driver method
public static void main(String args[])
{
// Create a graph given in the above diagram
Graph g1 = new Graph(5);

g1.addEdge(1, 0);
g1.addEdge(0, 2);
g1.addEdge(0, 3);
g1.addEdge(3, 4);
if (g1.isTree())
System.out.println("Graph is Tree");
else
System.out.println("Graph is not Tree");
Graph g2 = new Graph(5);
g2.addEdge(1, 0);
g2.addEdge(0, 2);
g2.addEdge(2, 1);
g2.addEdge(0, 3);
g2.addEdge(3, 4);
if (g2.isTree())
System.out.println("Graph is Tree");
else
System.out.println("Graph is not Tree");
}
}
// This code is contributed by Aakash Hasija

Graph is Tree
Graph is not Tree

Greedy Algorithms | Set 5 (Prims Minimum Spanning Tree (MST))

We have discussedKruskals algorithm for Minimum Spanning Tree. Like Kruskals algorithm, Prims algorithm is also aGreedy algorithm. It starts
with an empty spanning tree. The idea is to maintain two sets of vertices. The first set contains the vertices already included in the MST, the other
set contains the vertices not yet included. At every step, it considers all the edges that connect the two sets, and picks the minimum weight edge
from these edges. After picking the edge, it moves the other endpoint of the edge to the set containing MST.
A group of edges that connects two set of vertices in a graph is calledcut in graph theory.So, at every step of Prims algorithm, we find a cut (of
two sets, one contains the vertices already included in MST and other contains rest of the verices), pick the minimum weight edge from
the cut and include this vertex to MST Set (the set that contains already included vertices).
How does Prims Algorithm Work? The idea behind Prims algorithm is simple, a spanning tree means all vertices must be connected. So the two
disjoint subsets (discussed above) of vertices must be connected to make aSpanningTree. And they must be connected with the minimum weight
edge to make it aMinimumSpanning Tree.
Algorithm
1) Create a set mstSet that keeps track of vertices already included in MST.
2) Assign a key value to all vertices in the input graph. Initialize all key values as INFINITE. Assign key value as 0 for the first vertex so that it is
picked first.
3) While mstSet doesnt include all vertices
.a) Pick a vertex u which is not there in mstSet and has minimum key value.
.b) Include u to mstSet.
.c) Update key value of all adjacent vertices of u. To update the key values, iterate through all adjacent vertices. For every adjacent vertex v, if
weight of edge u-v is less than the previous key value of v, update the key value as weight of u-v
The idea of using key values is to pick the minimum weight edge from cut. The key values are used only for vertices which are not yet included in
MST, the key value for these vertices indicate the minimum weight edges connecting them to the set of vertices included in MST.
Let us understand with the following example:

The set mstSet is initially empty and keys assigned to vertices are {0, INF, INF, INF, INF, INF, INF, INF} where INF indicates infinite. Now
pick the vertex with minimum key value. The vertex 0 is picked, include it in mstSet. So mstSet becomes {0}. After including to mstSet, update
key values of adjacent vertices. Adjacent vertices of 0 are 1 and 7. The key values of 1 and 7 are updated as 4 and 8. Following subgraph shows
vertices and their key values, only the vertices with finite key values are shown. The vertices included in MST are shown in green color.

Pick the vertex with minimum key value and not already included in MST (not in mstSET). The vertex 1 is picked and added to mstSet. So mstSet
now becomes {0, 1}. Update the key values of adjacent vertices of 1. The key value of vertex 2 becomes 8.

Pick the vertex with minimum key value and not already included in MST (not in mstSET). We can either pick vertex 7 or vertex 2, let vertex 7 is
picked. So mstSet now becomes {0, 1, 7}. Update the key values of adjacent vertices of 7. The key value of vertex 6 and 8 becomes finite (7
and 1 respectively).

Pick the vertex with minimum key value and not already included in MST (not in mstSET). Vertex 6 is picked. So mstSet now becomes {0, 1, 7,
6}. Update the key values of adjacent vertices of 6. The key value of vertex 5 and 8 are updated.

We repeat the above steps until mstSet includes all vertices of given graph. Finally, we get the following graph.

How to implement the above algorithm?

We use a boolean array mstSet[] to represent the set of vertices included in MST. If a value mstSet[v] is true, then vertex v is included in MST,
otherwise not. Array key[] is used to store key values of all vertices. Another array parent[] to store indexes of parent nodes in MST. The parent
array is the output array which is used to show the constructed MST.

C/C++
// A C / C++ program for Prim's Minimum Spanning Tree (MST) algorithm.
// The program is for adjacency matrix representation of the graph
#include <stdio.h>
#include <limits.h>
// Number of vertices in the graph
#define V 5
// A utility function to find the vertex with minimum key value, from
// the set of vertices not yet included in MST
int minKey(int key[], bool mstSet[])
{
// Initialize min value
int min = INT_MAX, min_index;
for (int v = 0; v < V; v++)
if (mstSet[v] == false && key[v] < min)
min = key[v], min_index = v;
return min_index;
}
// A utility function to print the constructed MST stored in parent[]
int printMST(int parent[], int n, int graph[V][V])
{
printf("Edge Weight\n");
for (int i = 1; i < V; i++)
printf("%d - %d
%d \n", parent[i], i, graph[i][parent[i]]);
}
// Function to construct and print MST for a graph represented using adjacency
// matrix representation
void primMST(int graph[V][V])
{
int parent[V]; // Array to store constructed MST
int key[V]; // Key values used to pick minimum weight edge in cut
bool mstSet[V]; // To represent set of vertices not yet included in MST

// Initialize all keys as INFINITE

for (int i = 0; i < V; i++)
key[i] = INT_MAX, mstSet[i] = false;
// Always include first 1st vertex in MST.
key[0] = 0;
// Make key 0 so that this vertex is picked as first vertex
parent[0] = -1; // First node is always root of MST
// The MST will have V vertices
for (int count = 0; count < V-1; count++)
{
// Pick thd minimum key vertex from the set of vertices
// not yet included in MST
int u = minKey(key, mstSet);
// Add the picked vertex to the MST Set
mstSet[u] = true;
// Update key value and parent index of the adjacent vertices of
// the picked vertex. Consider only those vertices which are not yet
// included in MST
for (int v = 0; v < V; v++)
// graph[u][v] is non zero only for adjacent vertices of m
// mstSet[v] is false for vertices not yet included in MST
// Update the key only if graph[u][v] is smaller than key[v]
if (graph[u][v] && mstSet[v] == false && graph[u][v] < key[v])
parent[v] = u, key[v] = graph[u][v];
}
// print the constructed MST
printMST(parent, V, graph);
}
// driver program to test above function
int main()
{
/* Let us create the following graph
2
3
(0)--(1)--(2)
| / \ |
6| 8/ \5 |7
| /
\ |
(3)-------(4)
9
*/
int graph[V][V] = {{0, 2, 0, 6, 0},
{2, 0, 3, 8, 5},
{0, 3, 0, 0, 7},
{6, 8, 0, 0, 9},
{0, 5, 7, 9, 0},
};
// Print the solution
primMST(graph);
return 0;
}

Java
// A Java program for Prim's Minimum Spanning Tree (MST) algorithm.
// The program is for adjacency matrix representation of the graph
import java.util.*;
import java.lang.*;
import java.io.*;
class MST
{
// Number of vertices in the graph
private static final int V=5;
// A utility function to find the vertex with minimum key
// value, from the set of vertices not yet included in MST
int minKey(int key[], Boolean mstSet[])
{
// Initialize min value
int min = Integer.MAX_VALUE, min_index=-1;

for (int v = 0; v < V; v++)

if (mstSet[v] == false && key[v] < min)
{
min = key[v];
min_index = v;
}
return min_index;
}
// A utility function to print the constructed MST stored in
// parent[]
void printMST(int parent[], int n, int graph[][])
{
System.out.println("Edge Weight");
for (int i = 1; i < V; i++)
System.out.println(parent[i]+" - "+ i+"
"+
graph[i][parent[i]]);
}
// Function to construct and print MST for a graph represented
// using adjacency matrix representation
void primMST(int graph[][])
{
// Array to store constructed MST
int parent[] = new int[V];
// Key values used to pick minimum weight edge in cut
int key[] = new int [V];
// To represent set of vertices not yet included in MST
Boolean mstSet[] = new Boolean[V];
// Initialize all keys as INFINITE
for (int i = 0; i < V; i++)
{
key[i] = Integer.MAX_VALUE;
mstSet[i] = false;
}
// Always include first 1st vertex in MST.
key[0] = 0;
// Make key 0 so that this vertex is
// picked as first vertex
parent[0] = -1; // First node is always root of MST
// The MST will have V vertices
for (int count = 0; count < V-1; count++)
{
// Pick thd minimum key vertex from the set of vertices
// not yet included in MST
int u = minKey(key, mstSet);
// Add the picked vertex to the MST Set
mstSet[u] = true;
// Update key value and parent index of the adjacent
// vertices of the picked vertex. Consider only those
// vertices which are not yet included in MST
for (int v = 0; v < V; v++)
//
//
//
if

graph[u][v] is non zero only for adjacent vertices of m

mstSet[v] is false for vertices not yet included in MST
Update the key only if graph[u][v] is smaller than key[v]
(graph[u][v]!=0 && mstSet[v] == false &&
graph[u][v] < key[v])

{
parent[v] = u;
key[v] = graph[u][v];
}
}
// print the constructed MST
printMST(parent, V, graph);
}
public static void main (String[] args)
{
/* Let us create the following graph
2
3
(0)--(1)--(2)

|
/ \ |
6| 8/ \5 |7
| /
\ |
(3)-------(4)
9
*/
MST t = new MST();
int graph[][] = new int[][] {{0, 2, 0, 6, 0},
{2, 0, 3, 8, 5},
{0, 3, 0, 0, 7},
{6, 8, 0, 0, 9},
{0, 5, 7, 9, 0},
};
// Print the solution
t.primMST(graph);
}
}
// This code is contributed by Aakash Hasija

Edge Weight
0 - 1
2
1 - 2
3
0 - 3
6
1 - 4
5

Time Complexity of the above program is O(V^2). If the input graph is represented using adjacency list, then the time complexity of Prims
algorithm can be reduced to O(E log V) with the help of binary heap. Please see Prims MST for Adjacency List Representation for more details.

Applications of Minimum Spanning Tree Problem

Minimum Spanning Tree (MST) problem: Given connected graph G with positive edge weights, find a min weight set of edges that connects all of
the vertices.
MST is fundamental problem with diverse applications.

Network design.
telephone, electrical, hydraulic, TV cable, computer, road
The standard application is to a problem like phone network design. You have a business with several offices; you want to lease phone lines to
connect them up with each other; and the phone company charges different amounts of money to connect different pairs of cities. You want a set
of lines that connects all your offices with a minimum total cost. It should be a spanning tree, since if a network isnt a tree you can always remove
some edges and save money.
Approximation algorithms for NP-hard problems.
traveling salesperson problem, Steiner tree
A less obvious application is that the minimum spanning tree can be used to approximately solve the traveling salesman problem. A convenient
formal way of defining this problem is to find the shortest path that visits each point at least once.
Note that if you have a path visiting all points exactly once, its a special kind of tree. For instance in the example above, twelve of sixteen spanning
trees are actually paths. If you have a path visiting some vertices more than once, you can always drop some edges to get a tree. So in general the
MST weight is less than the TSP weight, because its a minimization over a strictly larger set.
On the other hand, if you draw a path tracing around the minimum spanning tree, you trace each edge twice and visit all points, so the TSP weight
is less than twice the MST weight. Therefore this tour is within a factor of two of optimal.
Indirect applications.
max bottleneck paths
LDPC codes for error correction
image registration with Renyi entropy
learning salient features for real-time face verification
reducing data storage in sequencing amino acids in a protein
model locality of particle interactions in turbulent fluid flows
autoconfig protocol for Ethernet bridging to avoid cycles in a network
Cluster analysis
k clustering problem can be viewed as finding an MST and deleting the k-1 most
expensive edges.
Sources:
http://www.cs.princeton.edu/courses/archive/spr07/cos226/lectures/mst.pdf
http://www.ics.uci.edu/~eppstein/161/960206.html

Greedy Algorithms | Set 6 (Prims MST for Adjacency List Representation)

We recommend to read following two posts as a prerequisite of this post.
1. Greedy Algorithms | Set 5 (Prims Minimum Spanning Tree (MST))
2. Graph and its representations
We have discussed Prims algorithm and its implementation for adjacency matrix representation of graphs. The time complexity for the matrix
representation is O(V^2). In this post, O(ELogV) algorithm for adjacency list representation is discussed.
As discussed in the previous post, in Prims algorithm, two sets are maintained, one set contains list of vertices already included in MST, other set
contains vertices not yet included. With adjacency list representation, all vertices of a graph can be traversed in O(V+E) time using BFS. The idea
is to traverse all vertices of graph using BFS and use a Min Heap to store the vertices not yet included in MST. Min Heap is used as a priority
queue to get the minimum weight edge from the cut. Min Heap is used as time complexity of operations like extracting minimum element and
decreasing key value is O(LogV) in Min Heap.
Following are the detailed steps.
1) Create a Min Heap of size V where V is the number of vertices in the given graph. Every node of min heap contains vertex number and key
value of the vertex.
2) Initialize Min Heap with first vertex as root (the key value assigned to first vertex is 0). The key value assigned to all other vertices is INF
(infinite).
3) While Min Heap is not empty, do following
..a) Extract the min value node from Min Heap. Let the extracted vertex be u.
..b) For every adjacent vertex v of u, check if v is in Min Heap (not yet included in MST). If v is in Min Heap and its key value is more than weight
of u-v, then update the key value of v as weight of u-v.
Let us understand the above algorithm with the following example:

Initially, key value of first vertex is 0 and INF (infinite) for all other vertices. So vertex 0 is extracted from Min Heap and key values of vertices
adjacent to 0 (1 and 7) are updated. Min Heap contains all vertices except vertex 0.
The vertices in green color are the vertices included in MST.

Since key value of vertex 1 is minimum among all nodes in Min Heap, it is extracted from Min Heap and key values of vertices adjacent to 1 are
updated (Key is updated if the a vertex is not in Min Heap and previous key value is greater than the weight of edge from 1 to the adjacent). Min
Heap contains all vertices except vertex 0 and 1.

Since key value of vertex 7 is minimum among all nodes in Min Heap, it is extracted from Min Heap and key values of vertices adjacent to 7 are
updated (Key is updated if the a vertex is not in Min Heap and previous key value is greater than the weight of edge from 7 to the adjacent). Min
Heap contains all vertices except vertex 0, 1 and 7.

Since key value of vertex 6 is minimum among all nodes in Min Heap, it is extracted from Min Heap and key values of vertices adjacent to 6 are
updated (Key is updated if the a vertex is not in Min Heap and previous key value is greater than the weight of edge from 6 to the adjacent). Min
Heap contains all vertices except vertex 0, 1, 7 and 6.

The above steps are repeated for rest of the nodes in Min Heap till Min Heap becomes empty

// C / C++ program for Prim's MST for adjacency list representation of graph
#include <stdio.h>
#include <stdlib.h>
#include <limits.h>
// A structure to represent a node in adjacency list
struct AdjListNode
{
int dest;
int weight;
struct AdjListNode* next;
};
// A structure to represent an adjacency liat
struct AdjList
{
struct AdjListNode *head; // pointer to head node of list
};
// A structure to represent a graph. A graph is an array of adjacency lists.
// Size of array will be V (number of vertices in graph)
struct Graph
{
int V;
struct AdjList* array;
};
// A utility function to create a new adjacency list node
struct AdjListNode* newAdjListNode(int dest, int weight)
{
struct AdjListNode* newNode =
(struct AdjListNode*) malloc(sizeof(struct AdjListNode));
newNode->dest = dest;
newNode->weight = weight;
newNode->next = NULL;
return newNode;
}
// A utility function that creates a graph of V vertices
struct Graph* createGraph(int V)
{
struct Graph* graph = (struct Graph*) malloc(sizeof(struct Graph));
graph->V = V;
// Create an array of adjacency lists. Size of array will be V
graph->array = (struct AdjList*) malloc(V * sizeof(struct AdjList));
// Initialize each adjacency list as empty by making head as NULL
for (int i = 0; i < V; ++i)
graph->array[i].head = NULL;
return graph;
}
// Adds an edge to an undirected graph
void addEdge(struct Graph* graph, int src, int dest, int weight)
{
// Add an edge from src to dest. A new node is added to the adjacency
// list of src. The node is added at the begining
struct AdjListNode* newNode = newAdjListNode(dest, weight);
newNode->next = graph->array[src].head;

graph->array[src].head = newNode;
// Since graph is undirected, add an edge from dest to src also
newNode = newAdjListNode(src, weight);
newNode->next = graph->array[dest].head;
graph->array[dest].head = newNode;
}
// Structure to represent a min heap node
struct MinHeapNode
{
int v;
int key;
};
// Structure to represent a min heap
struct MinHeap
{
int size;
// Number of heap nodes present currently
int capacity; // Capacity of min heap
int *pos;
// This is needed for decreaseKey()
struct MinHeapNode **array;
};
// A utility function to create a new Min Heap Node
struct MinHeapNode* newMinHeapNode(int v, int key)
{
struct MinHeapNode* minHeapNode =
(struct MinHeapNode*) malloc(sizeof(struct MinHeapNode));
minHeapNode->v = v;
minHeapNode->key = key;
return minHeapNode;
}
// A utilit function to create a Min Heap
struct MinHeap* createMinHeap(int capacity)
{
struct MinHeap* minHeap =
(struct MinHeap*) malloc(sizeof(struct MinHeap));
minHeap->pos = (int *)malloc(capacity * sizeof(int));
minHeap->size = 0;
minHeap->capacity = capacity;
minHeap->array =
(struct MinHeapNode**) malloc(capacity * sizeof(struct MinHeapNode*));
return minHeap;
}
// A utility function to swap two nodes of min heap. Needed for min heapify
void swapMinHeapNode(struct MinHeapNode** a, struct MinHeapNode** b)
{
struct MinHeapNode* t = *a;
*a = *b;
*b = t;
}
// A standard function to heapify at given idx
// This function also updates position of nodes when they are swapped.
// Position is needed for decreaseKey()
void minHeapify(struct MinHeap* minHeap, int idx)
{
int smallest, left, right;
smallest = idx;
left = 2 * idx + 1;
right = 2 * idx + 2;
if (left < minHeap->size &&
minHeap->array[left]->key < minHeap->array[smallest]->key )
smallest = left;
if (right < minHeap->size &&
minHeap->array[right]->key < minHeap->array[smallest]->key )
smallest = right;
if (smallest != idx)
{
// The nodes to be swapped in min heap
MinHeapNode *smallestNode = minHeap->array[smallest];
MinHeapNode *idxNode = minHeap->array[idx];
// Swap positions
minHeap->pos[smallestNode->v] = idx;

minHeap->pos[idxNode->v] = smallest;
// Swap nodes
swapMinHeapNode(&minHeap->array[smallest], &minHeap->array[idx]);
minHeapify(minHeap, smallest);
}
}
// A utility function to check if the given minHeap is ampty or not
int isEmpty(struct MinHeap* minHeap)
{
return minHeap->size == 0;
}
// Standard function to extract minimum node from heap
struct MinHeapNode* extractMin(struct MinHeap* minHeap)
{
if (isEmpty(minHeap))
return NULL;
// Store the root node
struct MinHeapNode* root = minHeap->array[0];
// Replace root node with last node
struct MinHeapNode* lastNode = minHeap->array[minHeap->size - 1];
minHeap->array[0] = lastNode;
// Update position of last node
minHeap->pos[root->v] = minHeap->size-1;
minHeap->pos[lastNode->v] = 0;
// Reduce heap size and heapify root
--minHeap->size;
minHeapify(minHeap, 0);
return root;
}
// Function to decreasy key value of a given vertex v. This function
// uses pos[] of min heap to get the current index of node in min heap
void decreaseKey(struct MinHeap* minHeap, int v, int key)
{
// Get the index of v in heap array
int i = minHeap->pos[v];
// Get the node and update its key value
minHeap->array[i]->key = key;
// Travel up while the complete tree is not hepified.
// This is a O(Logn) loop
while (i && minHeap->array[i]->key < minHeap->array[(i - 1) / 2]->key)
{
// Swap this node with its parent
minHeap->pos[minHeap->array[i]->v] = (i-1)/2;
minHeap->pos[minHeap->array[(i-1)/2]->v] = i;
swapMinHeapNode(&minHeap->array[i], &minHeap->array[(i - 1) / 2]);
// move to parent index
i = (i - 1) / 2;
}
}
// A utility function to check if a given vertex
// 'v' is in min heap or not
bool isInMinHeap(struct MinHeap *minHeap, int v)
{
if (minHeap->pos[v] < minHeap->size)
return true;
return false;
}
// A utility function used to print the constructed MST
void printArr(int arr[], int n)
{
for (int i = 1; i < n; ++i)
printf("%d - %d\n", arr[i], i);
}
// The main function that constructs Minimum Spanning Tree (MST)
// using Prim's algorithm

void PrimMST(struct Graph* graph)

{
int V = graph->V;// Get the number of vertices in graph
int parent[V]; // Array to store constructed MST
int key[V];
// Key values used to pick minimum weight edge in cut
// minHeap represents set E
struct MinHeap* minHeap = createMinHeap(V);
// Initialize min heap with all vertices. Key value of
// all vertices (except 0th vertex) is initially infinite
for (int v = 1; v < V; ++v)
{
parent[v] = -1;
key[v] = INT_MAX;
minHeap->array[v] = newMinHeapNode(v, key[v]);
minHeap->pos[v] = v;
}
// Make key value of 0th vertex as 0 so that it
// is extracted first
key[0] = 0;
minHeap->array[0] = newMinHeapNode(0, key[0]);
minHeap->pos[0] = 0;
// Initially size of min heap is equal to V
minHeap->size = V;
// In the followin loop, min heap contains all nodes
// not yet added to MST.
while (!isEmpty(minHeap))
{
// Extract the vertex with minimum key value
struct MinHeapNode* minHeapNode = extractMin(minHeap);
int u = minHeapNode->v; // Store the extracted vertex number
// Traverse through all adjacent vertices of u (the extracted
// vertex) and update their key values
struct AdjListNode* pCrawl = graph->array[u].head;
while (pCrawl != NULL)
{
int v = pCrawl->dest;
//
//
//
if
{

If v is not yet included in MST and weight of u-v is

less than key value of v, then update key value and
parent of v
(isInMinHeap(minHeap, v) && pCrawl->weight < key[v])
key[v] = pCrawl->weight;
parent[v] = u;
decreaseKey(minHeap, v, key[v]);

}
pCrawl = pCrawl->next;
}
}
// print edges of MST
printArr(parent, V);
}
// Driver program to test above functions
int main()
{
// Let us create the graph given in above fugure
int V = 9;
struct Graph* graph = createGraph(V);
addEdge(graph, 0, 1, 4);
addEdge(graph, 0, 7, 8);
addEdge(graph, 1, 2, 8);
addEdge(graph, 1, 7, 11);
addEdge(graph, 2, 3, 7);
addEdge(graph, 2, 8, 2);
addEdge(graph, 2, 5, 4);
addEdge(graph, 3, 4, 9);
addEdge(graph, 3, 5, 14);
addEdge(graph, 4, 5, 10);
addEdge(graph, 5, 6, 2);
addEdge(graph, 6, 7, 1);
addEdge(graph, 6, 8, 6);
addEdge(graph, 7, 8, 7);

PrimMST(graph);
return 0;
}

Output:
0
5
2
3
6
7
0
2

1
2
3
4
5
6
7
8

Time Complexity: The time complexity of the above code/algorithm looks O(V^2) as there are two nested while loops. If we take a closer look,
we can observe that the statements in inner loop are executed O(V+E) times (similar to BFS). The inner loop has decreaseKey() operation which
takes O(LogV) time. So overall time complexity is O(E+V)*O(LogV) which is O((E+V)*LogV) = O(ELogV) (For a connected graph, V =
O(E))
References:
Introduction to Algorithms by Clifford Stein, Thomas H. Cormen, Charles E. Leiserson, Ronald L.
http://en.wikipedia.org/wiki/Prims_algorithm

Greedy Algorithms | Set 2 (Kruskals Minimum Spanning Tree Algorithm)

What is Minimum Spanning Tree?
Given a connected and undirected graph, a spanning tree of that graph is a subgraph that is a tree and connects all the vertices together. A single
graph can have many different spanning trees. A minimum spanning tree (MST) or minimum weight spanning tree for a weighted, connected and
undirected graph is a spanning tree with weight less than or equal to the weight of every other spanning tree. The weight of a spanning tree is the
sum of weights given to each edge of the spanning tree.
How many edges does a minimum spanning tree has?
A minimum spanning tree has (V 1) edges where V is the number of vertices in the given graph.
What are the applications of Minimum Spanning Tree?
See this for applications of MST.
Below are the steps for finding MST using Kruskals algorithm
1. Sort all the edges in non-decreasing order of their weight.
2. Pick the smallest edge. Check if it forms a cycle with the spanning tree
formed so far. If cycle is not formed, include this edge. Else, discard it.
3. Repeat step#2 until there are (V-1) edges in the spanning tree.

The step#2 uses Union-Find algorithm to detect cycle. So we recommend to read following post as a prerequisite.
Union-Find Algorithm | Set 1 (Detect Cycle in a Graph)
Union-Find Algorithm | Set 2 (Union By Rank and Path Compression)
The algorithm is a Greedy Algorithm. The Greedy Choice is to pick the smallest weight edge that does not cause a cycle in the MST constructed
so far. Let us understand it with an example: Consider the below input graph.

The graph contains 9 vertices and 14 edges. So, the minimum spanning tree formed will be having (9 1) = 8 edges.
After sorting:
Weight Src
Dest
1
7
6
2
8
2
2
6
5
4
0
1
4
2
5
6
8
6
7
2
3
7
7
8
8
0
7
8
1
2
9
3
4
10
5
4
11
1
7
14
3
5

Now pick all edges one by one from sorted list of edges
1. Pick edge 7-6: No cycle is formed, include it.

2. Pick edge 8-2: No cycle is formed, include it.

3. Pick edge 6-5: No cycle is formed, include it.

4. Pick edge 0-1: No cycle is formed, include it.

5. Pick edge 2-5: No cycle is formed, include it.

6. Pick edge 8-6: Since including this edge results in cycle, discard it.
7. Pick edge 2-3: No cycle is formed, include it.

8. Pick edge 7-8: Since including this edge results in cycle, discard it.
9. Pick edge 0-7: No cycle is formed, include it.

10. Pick edge 1-2: Since including this edge results in cycle, discard it.
11. Pick edge 3-4: No cycle is formed, include it.

Since the number of edges included equals (V 1), the algorithm stops here.

C/C++
// C++ program for Kruskal's algorithm to find Minimum Spanning Tree
// of a given connected, undirected and weighted graph
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
// a structure to represent a weighted edge in graph
struct Edge
{
int src, dest, weight;

};
// a structure to represent a connected, undirected and weighted graph
struct Graph
{
// V-> Number of vertices, E-> Number of edges
int V, E;
// graph is represented as an array of edges. Since the graph is
// undirected, the edge from src to dest is also edge from dest
// to src. Both are counted as 1 edge here.
struct Edge* edge;
};
// Creates a graph with V vertices and E edges
struct Graph* createGraph(int V, int E)
{
struct Graph* graph = (struct Graph*) malloc( sizeof(struct Graph) );
graph->V = V;
graph->E = E;
graph->edge = (struct Edge*) malloc( graph->E * sizeof( struct Edge ) );
return graph;
}
// A structure to represent a subset for union-find
struct subset
{
int parent;
int rank;
};
// A utility function to find set of an element i
// (uses path compression technique)
int find(struct subset subsets[], int i)
{
// find root and make root as parent of i (path compression)
if (subsets[i].parent != i)
subsets[i].parent = find(subsets, subsets[i].parent);
return subsets[i].parent;
}
// A function that does union of two sets of x and y
// (uses union by rank)
void Union(struct subset subsets[], int x, int y)
{
int xroot = find(subsets, x);
int yroot = find(subsets, y);
// Attach smaller rank tree under root of high rank tree
// (Union by Rank)
if (subsets[xroot].rank < subsets[yroot].rank)
subsets[xroot].parent = yroot;
else if (subsets[xroot].rank > subsets[yroot].rank)
subsets[yroot].parent = xroot;
// If ranks are same, then make one as root and increment
// its rank by one
else
{
subsets[yroot].parent = xroot;
subsets[xroot].rank++;
}
}
// Compare two edges according to their weights.
// Used in qsort() for sorting an array of edges
int myComp(const void* a, const void* b)
{
struct Edge* a1 = (struct Edge*)a;
struct Edge* b1 = (struct Edge*)b;
return a1->weight > b1->weight;
}
// The main function to construct MST using Kruskal's algorithm
void KruskalMST(struct Graph* graph)
{
int V = graph->V;
struct Edge result[V]; // Tnis will store the resultant MST

int e = 0; // An index variable, used for result[]

int i = 0; // An index variable, used for sorted edges
// Step 1: Sort all the edges in non-decreasing order of their weight
// If we are not allowed to change the given graph, we can create a copy of
// array of edges
qsort(graph->edge, graph->E, sizeof(graph->edge[0]), myComp);
// Allocate memory for creating V ssubsets
struct subset *subsets =
(struct subset*) malloc( V * sizeof(struct subset) );
// Create V subsets with single elements
for (int v = 0; v < V; ++v)
{
subsets[v].parent = v;
subsets[v].rank = 0;
}
// Number of edges to be taken is equal to V-1
while (e < V - 1)
{
// Step 2: Pick the smallest edge. And increment the index
// for next iteration
struct Edge next_edge = graph->edge[i++];
int x = find(subsets, next_edge.src);
int y = find(subsets, next_edge.dest);
// If including this edge does't cause cycle, include it
// in result and increment the index of result for next edge
if (x != y)
{
result[e++] = next_edge;
Union(subsets, x, y);
}
// Else discard the next_edge
}
// print the contents of result[] to display the built MST
printf("Following are the edges in the constructed MST\n");
for (i = 0; i < e; ++i)
printf("%d -- %d == %d\n", result[i].src, result[i].dest,
result[i].weight);
return;
}
// Driver program to test above functions
int main()
{
/* Let us create following weighted graph
10
0--------1
| \
|
6| 5\ |15
|
\ |
2--------3
4
*/
int V = 4; // Number of vertices in graph
int E = 5; // Number of edges in graph
struct Graph* graph = createGraph(V, E);
// add edge 0-1
graph->edge[0].src = 0;
graph->edge[0].dest = 1;
graph->edge[0].weight = 10;
// add edge 0-2
graph->edge[1].src = 0;
graph->edge[1].dest = 2;
graph->edge[1].weight = 6;
// add edge 0-3
graph->edge[2].src = 0;
graph->edge[2].dest = 3;
graph->edge[2].weight = 5;
// add edge 1-3
graph->edge[3].src = 1;
graph->edge[3].dest = 3;

graph->edge[3].weight = 15;
// add edge 2-3
graph->edge[4].src = 2;
graph->edge[4].dest = 3;
graph->edge[4].weight = 4;
KruskalMST(graph);
return 0;
}

Java
// Java program for Kruskal's algorithm to find Minimum Spanning Tree
// of a given connected, undirected and weighted graph
import java.util.*;
import java.lang.*;
import java.io.*;
class Graph
{
// A class to represent a graph edge
class Edge implements Comparable<Edge>
{
int src, dest, weight;
// Comparator function used for sorting edges based on
// their weight
public int compareTo(Edge compareEdge)
{
return this.weight-compareEdge.weight;
}
};
// A class to represent a subset for union-find
class subset
{
int parent, rank;
};
int V, E;
// V-> no. of vertices & E->no.of edges
Edge edge[]; // collection of all edges
// Creates a graph with V vertices and E edges
Graph(int v, int e)
{
V = v;
E = e;
edge = new Edge[E];
for (int i=0; i<e; ++i)
edge[i] = new Edge();
}
// A utility function to find set of an element i
// (uses path compression technique)
int find(subset subsets[], int i)
{
// find root and make root as parent of i (path compression)
if (subsets[i].parent != i)
subsets[i].parent = find(subsets, subsets[i].parent);
return subsets[i].parent;
}
// A function that does union of two sets of x and y
// (uses union by rank)
void Union(subset subsets[], int x, int y)
{
int xroot = find(subsets, x);
int yroot = find(subsets, y);
// Attach smaller rank tree under root of high rank tree
// (Union by Rank)
if (subsets[xroot].rank < subsets[yroot].rank)
subsets[xroot].parent = yroot;
else if (subsets[xroot].rank > subsets[yroot].rank)
subsets[yroot].parent = xroot;
// If ranks are same, then make one as root and increment

// its rank by one

else
{
subsets[yroot].parent = xroot;
subsets[xroot].rank++;
}
}
// The main function to construct MST using Kruskal's algorithm
void KruskalMST()
{
Edge result[] = new Edge[V]; // Tnis will store the resultant MST
int e = 0; // An index variable, used for result[]
int i = 0; // An index variable, used for sorted edges
for (i=0; i<V; ++i)
result[i] = new Edge();
// Step 1: Sort all the edges in non-decreasing order of their
// weight. If we are not allowed to change the given graph, we
// can create a copy of array of edges
Arrays.sort(edge);
// Allocate memory for creating V ssubsets
subset subsets[] = new subset[V];
for(i=0; i<V; ++i)
subsets[i]=new subset();
// Create V subsets with single elements
for (int v = 0; v < V; ++v)
{
subsets[v].parent = v;
subsets[v].rank = 0;
}
i = 0; // Index used to pick next edge
// Number of edges to be taken is equal to V-1
while (e < V - 1)
{
// Step 2: Pick the smallest edge. And increment the index
// for next iteration
Edge next_edge = new Edge();
next_edge = edge[i++];
int x = find(subsets, next_edge.src);
int y = find(subsets, next_edge.dest);
// If including this edge does't cause cycle, include it
// in result and increment the index of result for next edge
if (x != y)
{
result[e++] = next_edge;
Union(subsets, x, y);
}
// Else discard the next_edge
}
// print the contents of result[] to display the built MST
System.out.println("Following are the edges in the constructed MST");
for (i = 0; i < e; ++i)
System.out.println(result[i].src+" -- "+result[i].dest+" == "+
result[i].weight);
}
// Driver Program
public static void main (String[] args)
{
/* Let us create following weighted graph
10
0--------1
| \
|
6| 5\ |15
|
\ |
2--------3
4
*/
int V = 4; // Number of vertices in graph
int E = 5; // Number of edges in graph
Graph graph = new Graph(V, E);
// add edge 0-1

graph.edge[0].src = 0;
graph.edge[0].dest = 1;
graph.edge[0].weight = 10;
// add edge 0-2
graph.edge[1].src = 0;
graph.edge[1].dest = 2;
graph.edge[1].weight = 6;
// add edge 0-3
graph.edge[2].src = 0;
graph.edge[2].dest = 3;
graph.edge[2].weight = 5;
// add edge 1-3
graph.edge[3].src = 1;
graph.edge[3].dest = 3;
graph.edge[3].weight = 15;
// add edge 2-3
graph.edge[4].src = 2;
graph.edge[4].dest = 3;
graph.edge[4].weight = 4;
graph.KruskalMST();
}
}
//This code is contributed by Aakash Hasija
Following
2 -- 3 ==
0 -- 3 ==
0 -- 1 ==

are the edges in the constructed MST

4
5
10

Time Complexity: O(ElogE) or O(ElogV). Sorting of edges takes O(ELogE) time. After sorting, we iterate through all edges and apply findunion algorithm. The find and union operations can take atmost O(LogV) time. So overall complexity is O(ELogE + ELogV) time. The value of E
can be atmost V^2, so O(LogV) are O(LogE) same. Therefore, overall time complexity is O(ElogE) or O(ElogV)
References:
http://www.ics.uci.edu/~eppstein/161/960206.html
http://en.wikipedia.org/wiki/Minimum_spanning_tree

Greedy Algorithms | Set 9 (Boruvkas algorithm)

We have discussed following topics on Minimum Spanning Tree.
Applications of Minimum Spanning Tree Problem
Kruskals Minimum Spanning Tree Algorithm
Prims Minimum Spanning Tree Algorithm
In this post, Boruvkas algorithm is discussed. Like Prims and Kruskals, Boruvkas algorithm is also a Greedy algorithm. Below is complete
algorithm.
1)
2)
3)
4)

Input is a connected, weighted and directed graph.

Initialize all vertices as individual components (or sets).
Initialize MST as empty.
While there are more than one components, do following
for each component.
a) Find the closest weight edge that connects this
component to any other component.
b) Add this closest edge to MST if not already added.
5) Return MST.

Below is the idea behind above algorithm (The idea is same as Prims MST algorithm).
A spanning tree means all vertices must be connected. So the two disjoint subsets (discussed above) of vertices must be connected to
make a Spanning Tree. And they must be connected with the minimum weight edge to make it a Minimum Spanning Tree.
Let us understand the algorithm with below example.

Initially MST is empty. Every vertex is singe component as highlighted in blue color in below diagram.

For every component, find the cheapest edge that connects it to some other component.
Component
{0}
{1}
{2}
{3}
{4}
{5}
{6}
{7}
{8}

Cheapest Edge that connects

it to some other component
0-1
0-1
2-8
2-3
3-4
5-6
6-7
6-7
2-8

The cheapest edges are highlighted with green color. Now MST becomes {0-1, 2-8, 2-3, 3-4, 5-6, 6-7}.

After above step, components are {{0,1}, {2,3,4,8}, {5,6,7}}. The components are encircled with blue color.

We again repeat the step, i.e., for every component, find the cheapest edge that connects it to some other component.
Component
{0,1}
{2,3,4,8}
{5,6,7}

Cheapest Edge that connects

it to some other component
1-2 (or 0-7)
2-5
2-5

The cheapest edges are highlighted with green color. Now MST becomes {0-1, 2-8, 2-3, 3-4, 5-6, 6-7, 1-2, 2-5}

At this stage, there is only one component {0, 1, 2, 3, 4, 5, 6, 7, 8} which has all edges. Since there is only one component left, we stop and
return MST.
Implementation:
Below is C++ implementation of above algorithm. The input graph is represented as a collection of edges and union-find data structure is used to
keep track of components.
// Boruvka's algorithm to find Minimum Spanning
// Tree of a given connected, undirected and
// weighted graph
#include <stdio.h>
// a structure to represent a weighted edge in graph
struct Edge
{
int src, dest, weight;
};
// a structure to represent a connected, undirected
// and weighted graph as a collection of edges.
struct Graph
{
// V-> Number of vertices, E-> Number of edges
int V, E;
// graph is represented as an array of edges.
// Since the graph is undirected, the edge
// from src to dest is also edge from dest
// to src. Both are counted as 1 edge here.
Edge* edge;
};
// A structure to represent a subset for union-find
struct subset
{
int parent;
int rank;
};
// Function prototypes for union-find (These functions are defined
// after boruvkaMST() )
int find(struct subset subsets[], int i);
void Union(struct subset subsets[], int x, int y);
// The main function for MST using Boruvka's algorithm
void boruvkaMST(struct Graph* graph)
{
// Get data of given graph
int V = graph->V, E = graph->E;
Edge *edge = graph->edge;

// Allocate memory for creating V subsets.

struct subset *subsets = new subset[V];
// An array to store index of the cheapest edge of
// subset. The stored index for indexing array 'edge[]'
int *cheapest = new int[V];
// Create V subsets with single elements
for (int v = 0; v < V; ++v)
{
subsets[v].parent = v;
subsets[v].rank = 0;
cheapest[v] = -1;
}
// Initially there are V different trees.
// Finally there will be one tree that will be MST
int numTrees = V;
int MSTweight = 0;
// Keep combining components (or sets) until all
// compnentes are not combined into single MST.
while (numTrees > 1)
{
// Traverse through all edges and update
// cheapest of every component
for (int i=0; i<E; i++)
{
// Find components (or sets) of two corners
// of current edge
int set1 = find(subsets, edge[i].src);
int set2 = find(subsets, edge[i].dest);
// If two corners of current edge belong to
// same set, ignore current edge
if (set1 == set2)
continue;
// Else check if current edge is closer to previous
// cheapest edges of set1 and set2
else
{
if (cheapest[set1] == -1 ||
edge[cheapest[set1]].weight > edge[i].weight)
cheapest[set1] = i;
if (cheapest[set1] == -1 ||
edge[cheapest[set2]].weight > edge[i].weight)
cheapest[set2] = i;
}
}
// Consider the above picked cheapest edges and add them
// to MST
for (int i=0; i<V; i++)
{
// Check if cheapest for current set exists
if (cheapest[i] != -1)
{
int set1 = find(subsets, edge[cheapest[i]].src);
int set2 = find(subsets, edge[cheapest[i]].dest);
if (set1 == set2)
continue;
MSTweight += edge[cheapest[i]].weight;
printf("Edge %d-%d included in MST\n",
edge[cheapest[i]].src, edge[cheapest[i]].dest,
edge[cheapest[i]].weight);
// Do a union of set1 and set2 and decrease number
// of trees
Union(subsets, set1, set2);
numTrees--;
}
}
}
printf("Weight of MST is %d\n", MSTweight);
return;
}

// Creates a graph with V vertices and E edges

struct Graph* createGraph(int V, int E)
{
Graph* graph = new Graph;
graph->V = V;
graph->E = E;
graph->edge = new Edge[E];
return graph;
}
// A utility function to find set of an element i
// (uses path compression technique)
int find(struct subset subsets[], int i)
{
// find root and make root as parent of i
// (path compression)
if (subsets[i].parent != i)
subsets[i].parent =
find(subsets, subsets[i].parent);
return subsets[i].parent;
}
// A function that does union of two sets of x and y
// (uses union by rank)
void Union(struct subset subsets[], int x, int y)
{
int xroot = find(subsets, x);
int yroot = find(subsets, y);
// Attach smaller rank tree under root of high
// rank tree (Union by Rank)
if (subsets[xroot].rank < subsets[yroot].rank)
subsets[xroot].parent = yroot;
else if (subsets[xroot].rank > subsets[yroot].rank)
subsets[yroot].parent = xroot;
// If ranks are same, then make one as root and
// increment its rank by one
else
{
subsets[yroot].parent = xroot;
subsets[xroot].rank++;
}
}
// Driver program to test above functions
int main()
{
/* Let us create following weighted graph
10
0--------1
| \
|
6| 5\ |15
|
\ |
2--------3
4
*/
int V = 4; // Number of vertices in graph
int E = 5; // Number of edges in graph
struct Graph* graph = createGraph(V, E);
// add edge 0-1
graph->edge[0].src = 0;
graph->edge[0].dest = 1;
graph->edge[0].weight = 10;
// add edge 0-2
graph->edge[1].src = 0;
graph->edge[1].dest = 2;
graph->edge[1].weight = 6;
// add edge 0-3
graph->edge[2].src = 0;
graph->edge[2].dest = 3;
graph->edge[2].weight = 5;
// add edge 1-3
graph->edge[3].src = 1;
graph->edge[3].dest = 3;

graph->edge[3].weight = 15;
// add edge 2-3
graph->edge[4].src = 2;
graph->edge[4].dest = 3;
graph->edge[4].weight = 4;
boruvkaMST(graph);
return 0;
}

Output:
Edge 0-3 included in MST
Edge 0-1 included in MST
Edge 2-3 included in MST
Weight of MST is 19

Interesting Facts about Boruvkas algorithm:

1) Time Complexity of Boruvkas algorithm is O(E log V) which is same as Kruskals and Prims algorithms.
2) Boruvkas algorithm is used as a step in a faster randomized algorithm that works in linear time O(E).
3) Boruvkas algorithm is the oldest minimum spanning tree algorithm was discovered by Boruuvka in 1926, long before computers even existed.
The algorithm was published as a method of constructing an efficient electricity network.
Exercise:
The above code assumes that input graph is connected and it fails if a disconnected graph is given. Extend the above algorithm so that it works for
a disconnected graph also and produces a forest.
References:
http://en.wikipedia.org/wiki/Bor%C5%AFvka%27s_algorithm

Greedy Algorithms | Set 7 (Dijkstras shortest path algorithm)

Given a graph and a source vertex in graph, find shortest paths from source to all vertices in the given graph.
Dijkstras algorithm is very similar to Prims algorithm for minimum spanning tree. Like Prims MST, we generate a SPT (shortest path tree) with
given source as root. We maintain two sets, one set contains vertices included in shortest path tree, other set includes vertices not yet included in
shortest path tree. At every step of the algorithm, we find a vertex which is in the other set (set of not yet included) and has minimum distance from
source.
Below are the detailed steps used in Dijkstras algorithm to find the shortest path from a single source vertex to all other vertices in the given graph.
Algorithm
1) Create a set sptSet (shortest path tree set) that keeps track of vertices included in shortest path tree, i.e., whose minimum distance from source
is calculated and finalized. Initially, this set is empty.
2) Assign a distance value to all vertices in the input graph. Initialize all distance values as INFINITE. Assign distance value as 0 for the source
vertex so that it is picked first.
3) While sptSet doesnt include all vertices
.a) Pick a vertex u which is not there in sptSetand has minimum distance value.
.b) Include u to sptSet.
.c) Update distance value of all adjacent vertices of u. To update the distance values, iterate through all adjacent vertices. For every adjacent
vertex v, if sum of distance value of u (from source) and weight of edge u-v, is less than the distance value of v, then update the distance value of v.
Let us understand with the following example:

The set sptSetis initially empty and distances assigned to vertices are {0, INF, INF, INF, INF, INF, INF, INF} where INF indicates infinite.
Now pick the vertex with minimum distance value. The vertex 0 is picked, include it in sptSet. So sptSet becomes {0}. After including 0 to
sptSet, update distance values of its adjacent vertices. Adjacent vertices of 0 are 1 and 7. The distance values of 1 and 7 are updated as 4 and 8.
Following subgraph shows vertices and their distance values, only the vertices with finite distance values are shown. The vertices included in SPT
are shown in green color.

Pick the vertex with minimum distance value and not already included in SPT (not in sptSET). The vertex 1 is picked and added to sptSet. So
sptSet now becomes {0, 1}. Update the distance values of adjacent vertices of 1. The distance value of vertex 2 becomes 12.

Pick the vertex with minimum distance value and not already included in SPT (not in sptSET). Vertex 7 is picked. So sptSet now becomes {0, 1,
7}. Update the distance values of adjacent vertices of 7. The distance value of vertex 6 and 8 becomes finite (15 and 9 respectively).

Pick the vertex with minimum distance value and not already included in SPT (not in sptSET). Vertex 6 is picked. So sptSet now becomes {0, 1,
7, 6}. Update the distance values of adjacent vertices of 6. The distance value of vertex 5 and 8 are updated.

We repeat the above steps until sptSet doesnt include all vertices of given graph. Finally, we get the following Shortest Path Tree (SPT).

How to implement the above algorithm?

We use a boolean array sptSet[] to represent the set of vertices included in SPT. If a value sptSet[v] is true, then vertex v is included in SPT,
otherwise not. Array dist[] is used to store shortest distance values of all vertices.

C/C++
// A C / C++ program for Dijkstra's single source shortest path algorithm.
// The program is for adjacency matrix representation of the graph
#include <stdio.h>
#include <limits.h>
// Number of vertices in the graph
#define V 9
// A utility function to find the vertex with minimum distance value, from
// the set of vertices not yet included in shortest path tree
int minDistance(int dist[], bool sptSet[])
{
// Initialize min value
int min = INT_MAX, min_index;
for (int v = 0; v < V; v++)
if (sptSet[v] == false && dist[v] <= min)
min = dist[v], min_index = v;
return min_index;
}
// A utility function to print the constructed distance array
int printSolution(int dist[], int n)
{
printf("Vertex Distance from Source\n");
for (int i = 0; i < V; i++)
printf("%d \t\t %d\n", i, dist[i]);
}
// Funtion that implements Dijkstra's single source shortest path algorithm
// for a graph represented using adjacency matrix representation
void dijkstra(int graph[V][V], int src)
{
int dist[V];
// The output array. dist[i] will hold the shortest
// distance from src to i
bool sptSet[V]; // sptSet[i] will true if vertex i is included in shortest
// path tree or shortest distance from src to i is finalized
// Initialize all distances as INFINITE and stpSet[] as false
for (int i = 0; i < V; i++)
dist[i] = INT_MAX, sptSet[i] = false;
// Distance of source vertex from itself is always 0
dist[src] = 0;
// Find shortest path for all vertices
for (int count = 0; count < V-1; count++)
{
// Pick the minimum distance vertex from the set of vertices not

// yet processed. u is always equal to src in first iteration.

int u = minDistance(dist, sptSet);
// Mark the picked vertex as processed
sptSet[u] = true;
// Update dist value of the adjacent vertices of the picked vertex.
for (int v = 0; v < V; v++)
//
//
//
if

Update dist[v] only if is not in sptSet, there is an edge from

u to v, and total weight of path from src to v through u is
smaller than current value of dist[v]
(!sptSet[v] && graph[u][v] && dist[u] != INT_MAX
&& dist[u]+graph[u][v] < dist[v])
dist[v] = dist[u] + graph[u][v];

}
// print the constructed distance array
printSolution(dist, V);
}
// driver program to test above function
int main()
{
/* Let us create the example graph discussed above */
int graph[V][V] = {{0, 4, 0, 0, 0, 0, 0, 8, 0},
{4, 0, 8, 0, 0, 0, 0, 11, 0},
{0, 8, 0, 7, 0, 4, 0, 0, 2},
{0, 0, 7, 0, 9, 14, 0, 0, 0},
{0, 0, 0, 9, 0, 10, 0, 0, 0},
{0, 0, 4, 0, 10, 0, 2, 0, 0},
{0, 0, 0, 14, 0, 2, 0, 1, 6},
{8, 11, 0, 0, 0, 0, 1, 0, 7},
{0, 0, 2, 0, 0, 0, 6, 7, 0}
};
dijkstra(graph, 0);
return 0;
}

Java
// A Java program for Dijkstra's single source shortest path algorithm.
// The program is for adjacency matrix representation of the graph
import java.util.*;
import java.lang.*;
import java.io.*;
class ShortestPath
{
// A utility function to find the vertex with minimum distance value,
// from the set of vertices not yet included in shortest path tree
static final int V=9;
int minDistance(int dist[], Boolean sptSet[])
{
// Initialize min value
int min = Integer.MAX_VALUE, min_index=-1;
for (int v = 0; v < V; v++)
if (sptSet[v] == false && dist[v] <= min)
{
min = dist[v];
min_index = v;
}
return min_index;
}
// A utility function to print the constructed distance array
void printSolution(int dist[], int n)
{
System.out.println("Vertex Distance from Source");
for (int i = 0; i < V; i++)
System.out.println(i+" \t\t "+dist[i]);
}
// Funtion that implements Dijkstra's single source shortest path
// algorithm for a graph represented using adjacency matrix
// representation

void dijkstra(int graph[][], int src)

{
int dist[] = new int[V]; // The output array. dist[i] will hold
// the shortest distance from src to i
// sptSet[i] will true if vertex i is included in shortest
// path tree or shortest distance from src to i is finalized
Boolean sptSet[] = new Boolean[V];
// Initialize all distances as INFINITE and stpSet[] as false
for (int i = 0; i < V; i++)
{
dist[i] = Integer.MAX_VALUE;
sptSet[i] = false;
}
// Distance of source vertex from itself is always 0
dist[src] = 0;
// Find shortest path for all vertices
for (int count = 0; count < V-1; count++)
{
// Pick the minimum distance vertex from the set of vertices
// not yet processed. u is always equal to src in first
// iteration.
int u = minDistance(dist, sptSet);
// Mark the picked vertex as processed
sptSet[u] = true;
// Update dist value of the adjacent vertices of the
// picked vertex.
for (int v = 0; v < V; v++)
//
//
//
if

Update dist[v] only if is not in sptSet, there is an

edge from u to v, and total weight of path from src to
v through u is smaller than current value of dist[v]
(!sptSet[v] && graph[u][v]!=0 &&
dist[u] != Integer.MAX_VALUE &&
dist[u]+graph[u][v] < dist[v])
dist[v] = dist[u] + graph[u][v];

}
// print the constructed distance array
printSolution(dist, V);
}
// Driver method
public static void main (String[] args)
{
/* Let us create the example graph discussed above */
int graph[][] = new int[][]{{0, 4, 0, 0, 0, 0, 0, 8, 0},
{4, 0, 8, 0, 0, 0, 0, 11, 0},
{0, 8, 0, 7, 0, 4, 0, 0, 2},
{0, 0, 7, 0, 9, 14, 0, 0, 0},
{0, 0, 0, 9, 0, 10, 0, 0, 0},
{0, 0, 4, 0, 10, 0, 2, 0, 0},
{0, 0, 0, 14, 0, 2, 0, 1, 6},
{8, 11, 0, 0, 0, 0, 1, 0, 7},
{0, 0, 2, 0, 0, 0, 6, 7, 0}
};
ShortestPath t = new ShortestPath();
t.dijkstra(graph, 0);
}
}
//This code is contributed by Aakash Hasija

Vertex
0
1
2
3
4
5
6
7
8

Notes:

Distance from Source

0
4
12
19
21
11
9
8
14

1) The code calculates shortest distance, but doesnt calculate the path information. We can create a parent array, update the parent array when
distance is updated (like prims implementation) and use it show the shortest path from source to different vertices.
2) The code is for undirected graph, same dijekstra function can be used for directed graphs also.
3) The code finds shortest distances from source to all vertices. If we are interested only in shortest distance from source to a single target, we can
break the for loop when the picked minimum distance vertex is equal to target (Step 3.a of algorithm).
4) Time Complexity of the implementation is O(V^2). If the input graph is represented using adjacency list, it can be reduced to O(E log V) with
the help of binary heap. Please see
Dijkstras Algorithm for Adjacency List Representation for more details.
5) Dijkstras algorithm doesnt work for graphs with negative weight edges. For graphs with negative weight edges, BellmanFord algorithm can be
used, we will soon be discussing it as a separate post.
Dijkstras Algorithm for Adjacency List Representation

Greedy Algorithms | Set 8 (Dijkstras Algorithm for Adjacency List Representation)

We recommend to read following two posts as a prerequisite of this post.
1. Greedy Algorithms | Set 7 (Dijkstras shortest path algorithm)
2.Graph and its representations
We have discussed Dijkstras algorithm and its implementation for adjacency matrix representation of graphs. The time complexity for the matrix
representation is O(V^2). In this post, O(ELogV) algorithm for adjacency list representation is discussed.
As discussed in the previous post, in Dijkstras algorithm, two sets are maintained, one set contains list of vertices already included in SPT (Shortest
Path Tree), other set contains vertices not yet included. With adjacency list representation, all vertices of a graph can be traversed in O(V+E) time
usingBFS. The idea is to traverse all vertices of graph usingBFSand use a Min Heap to store the vertices not yet included in SPT (or the vertices
for which shortest distance is not finalized yet). Min Heap is used as a priority queue to get the minimum distance vertex from set of not
yetincludedvertices. Time complexity of operations like extract-min and decrease-key value is O(LogV) for Min Heap.
Following are the detailed steps.
1)Create a Min Heap of size V where V is the number of vertices in the given graph. Every node of min heap contains vertex number and distance
value of the vertex.
2)Initialize Min Heap with source vertex as root (the distance value assigned to source vertex is 0). The distance value assigned to all other vertices
is INF (infinite).
3)While Min Heap is not empty, do following
..a)Extract the vertex with minimum distance value node from Min Heap. Let the extracted vertex be u.
..b)For every adjacent vertex v of u, check if v is in Min Heap. If v is in Min Heap and distance value is more than weight of u-v plus distance
value of u, then update the distance value of v.
Let us understand with the following example. Let the given source vertex be 0

Initially, distance value of source vertex is 0 and INF (infinite) for all other vertices. So source vertex is extracted from Min Heap and distance
values of vertices adjacent to 0 (1 and 7) are updated. Min Heap contains all vertices except vertex 0.
The vertices in green color are the vertices for which minimum distances are finalized and are not in Min Heap

Since distance value of vertex 1 is minimum among all nodes in Min Heap, it is extracted from Min Heap and distance values of vertices adjacent
to 1 are updated (distance is updated if the a vertex is not in Min Heap and distance through 1 is shorter than the previous distance). Min Heap
contains all vertices except vertex 0 and 1.

Pick the vertex with minimum distance value from min heap. Vertex 7 is picked. So min heap now contains all vertices except 0, 1 and 7. Update
the distance values of adjacent vertices of 7. The distance value of vertex 6 and 8 becomes finite (15 and 9 respectively).

Pick the vertex with minimum distance from min heap. Vertex 6 is picked. So min heap now contains all vertices except 0, 1, 7 and 6. Update the
distance values of adjacent vertices of 6. The distance value of vertex 5 and 8 are updated.

Above steps are repeated till min heap doesnt become empty. Finally, we get the following shortest path tree.

// C / C++ program for Dijkstra's shortest path algorithm for adjacency

// list representation of graph
#include <stdio.h>
#include <stdlib.h>
#include <limits.h>
// A structure to represent a node in adjacency list
struct AdjListNode
{
int dest;
int weight;
struct AdjListNode* next;
};
// A structure to represent an adjacency liat
struct AdjList
{
struct AdjListNode *head; // pointer to head node of list
};
// A structure to represent a graph. A graph is an array of adjacency lists.
// Size of array will be V (number of vertices in graph)
struct Graph
{
int V;
struct AdjList* array;
};
// A utility function to create a new adjacency list node
struct AdjListNode* newAdjListNode(int dest, int weight)
{
struct AdjListNode* newNode =
(struct AdjListNode*) malloc(sizeof(struct AdjListNode));
newNode->dest = dest;
newNode->weight = weight;
newNode->next = NULL;
return newNode;
}
// A utility function that creates a graph of V vertices
struct Graph* createGraph(int V)
{
struct Graph* graph = (struct Graph*) malloc(sizeof(struct Graph));
graph->V = V;
// Create an array of adjacency lists. Size of array will be V
graph->array = (struct AdjList*) malloc(V * sizeof(struct AdjList));
// Initialize each adjacency list as empty by making head as NULL
for (int i = 0; i < V; ++i)
graph->array[i].head = NULL;
return graph;
}

// Adds an edge to an undirected graph

void addEdge(struct Graph* graph, int src, int dest, int weight)
{
// Add an edge from src to dest. A new node is added to the adjacency
// list of src. The node is added at the begining
struct AdjListNode* newNode = newAdjListNode(dest, weight);
newNode->next = graph->array[src].head;
graph->array[src].head = newNode;
// Since graph is undirected, add an edge from dest to src also
newNode = newAdjListNode(src, weight);
newNode->next = graph->array[dest].head;
graph->array[dest].head = newNode;
}
// Structure to represent a min heap node
struct MinHeapNode
{
int v;
int dist;
};
// Structure to represent a min heap
struct MinHeap
{
int size;
// Number of heap nodes present currently
int capacity; // Capacity of min heap
int *pos;
// This is needed for decreaseKey()
struct MinHeapNode **array;
};
// A utility function to create a new Min Heap Node
struct MinHeapNode* newMinHeapNode(int v, int dist)
{
struct MinHeapNode* minHeapNode =
(struct MinHeapNode*) malloc(sizeof(struct MinHeapNode));
minHeapNode->v = v;
minHeapNode->dist = dist;
return minHeapNode;
}
// A utility function to create a Min Heap
struct MinHeap* createMinHeap(int capacity)
{
struct MinHeap* minHeap =
(struct MinHeap*) malloc(sizeof(struct MinHeap));
minHeap->pos = (int *)malloc(capacity * sizeof(int));
minHeap->size = 0;
minHeap->capacity = capacity;
minHeap->array =
(struct MinHeapNode**) malloc(capacity * sizeof(struct MinHeapNode*));
return minHeap;
}
// A utility function to swap two nodes of min heap. Needed for min heapify
void swapMinHeapNode(struct MinHeapNode** a, struct MinHeapNode** b)
{
struct MinHeapNode* t = *a;
*a = *b;
*b = t;
}
// A standard function to heapify at given idx
// This function also updates position of nodes when they are swapped.
// Position is needed for decreaseKey()
void minHeapify(struct MinHeap* minHeap, int idx)
{
int smallest, left, right;
smallest = idx;
left = 2 * idx + 1;
right = 2 * idx + 2;
if (left < minHeap->size &&
minHeap->array[left]->dist < minHeap->array[smallest]->dist )
smallest = left;
if (right < minHeap->size &&
minHeap->array[right]->dist < minHeap->array[smallest]->dist )
smallest = right;
if (smallest != idx)

{
// The nodes to be swapped in min heap
MinHeapNode *smallestNode = minHeap->array[smallest];
MinHeapNode *idxNode = minHeap->array[idx];
// Swap positions
minHeap->pos[smallestNode->v] = idx;
minHeap->pos[idxNode->v] = smallest;
// Swap nodes
swapMinHeapNode(&minHeap->array[smallest], &minHeap->array[idx]);
minHeapify(minHeap, smallest);
}
}
// A utility function to check if the given minHeap is ampty or not
int isEmpty(struct MinHeap* minHeap)
{
return minHeap->size == 0;
}
// Standard function to extract minimum node from heap
struct MinHeapNode* extractMin(struct MinHeap* minHeap)
{
if (isEmpty(minHeap))
return NULL;
// Store the root node
struct MinHeapNode* root = minHeap->array[0];
// Replace root node with last node
struct MinHeapNode* lastNode = minHeap->array[minHeap->size - 1];
minHeap->array[0] = lastNode;
// Update position of last node
minHeap->pos[root->v] = minHeap->size-1;
minHeap->pos[lastNode->v] = 0;
// Reduce heap size and heapify root
--minHeap->size;
minHeapify(minHeap, 0);
return root;
}
// Function to decreasy dist value of a given vertex v. This function
// uses pos[] of min heap to get the current index of node in min heap
void decreaseKey(struct MinHeap* minHeap, int v, int dist)
{
// Get the index of v in heap array
int i = minHeap->pos[v];
// Get the node and update its dist value
minHeap->array[i]->dist = dist;
// Travel up while the complete tree is not hepified.
// This is a O(Logn) loop
while (i && minHeap->array[i]->dist < minHeap->array[(i - 1) / 2]->dist)
{
// Swap this node with its parent
minHeap->pos[minHeap->array[i]->v] = (i-1)/2;
minHeap->pos[minHeap->array[(i-1)/2]->v] = i;
swapMinHeapNode(&minHeap->array[i], &minHeap->array[(i - 1) / 2]);
// move to parent index
i = (i - 1) / 2;
}
}
// A utility function to check if a given vertex
// 'v' is in min heap or not
bool isInMinHeap(struct MinHeap *minHeap, int v)
{
if (minHeap->pos[v] < minHeap->size)
return true;
return false;
}
// A utility function used to print the solution
void printArr(int dist[], int n)

{
printf("Vertex Distance from Source\n");
for (int i = 0; i < n; ++i)
printf("%d \t\t %d\n", i, dist[i]);
}
// The main function that calulates distances of shortest paths from src to all
// vertices. It is a O(ELogV) function
void dijkstra(struct Graph* graph, int src)
{
int V = graph->V;// Get the number of vertices in graph
int dist[V];
// dist values used to pick minimum weight edge in cut
// minHeap represents set E
struct MinHeap* minHeap = createMinHeap(V);
// Initialize min heap with all vertices. dist value of all vertices
for (int v = 0; v < V; ++v)
{
dist[v] = INT_MAX;
minHeap->array[v] = newMinHeapNode(v, dist[v]);
minHeap->pos[v] = v;
}
// Make dist value of src vertex as 0 so that it is extracted first
minHeap->array[src] = newMinHeapNode(src, dist[src]);
minHeap->pos[src] = src;
dist[src] = 0;
decreaseKey(minHeap, src, dist[src]);
// Initially size of min heap is equal to V
minHeap->size = V;
// In the followin loop, min heap contains all nodes
// whose shortest distance is not yet finalized.
while (!isEmpty(minHeap))
{
// Extract the vertex with minimum distance value
struct MinHeapNode* minHeapNode = extractMin(minHeap);
int u = minHeapNode->v; // Store the extracted vertex number
// Traverse through all adjacent vertices of u (the extracted
// vertex) and update their distance values
struct AdjListNode* pCrawl = graph->array[u].head;
while (pCrawl != NULL)
{
int v = pCrawl->dest;
// If shortest distance to v is not finalized yet, and distance to v
// through u is less than its previously calculated distance
if (isInMinHeap(minHeap, v) && dist[u] != INT_MAX &&
pCrawl->weight + dist[u] < dist[v])
{
dist[v] = dist[u] + pCrawl->weight;
// update distance value in min heap also
decreaseKey(minHeap, v, dist[v]);
}
pCrawl = pCrawl->next;
}
}
// print the calculated shortest distances
printArr(dist, V);
}
// Driver program to test above functions
int main()
{
// create the graph given in above fugure
int V = 9;
struct Graph* graph = createGraph(V);
addEdge(graph, 0, 1, 4);
addEdge(graph, 0, 7, 8);
addEdge(graph, 1, 2, 8);
addEdge(graph, 1, 7, 11);
addEdge(graph, 2, 3, 7);
addEdge(graph, 2, 8, 2);
addEdge(graph, 2, 5, 4);
addEdge(graph, 3, 4, 9);

addEdge(graph,
addEdge(graph,
addEdge(graph,
addEdge(graph,
addEdge(graph,
addEdge(graph,

3,
4,
5,
6,
6,
7,

5,
5,
6,
7,
8,
8,

14);
10);
2);
1);
6);
7);

dijkstra(graph, 0);
return 0;
}

Output:
Vertex
0
1
2
3
4
5
6
7
8

Distance from Source

0
4
12
19
21
11
9
8
14

Time Complexity:The time complexity of the above code/algorithm looks O(V^2) as there are two nested while loops. If we take a closer look,
we can observe that the statements in inner loop are executed O(V+E) times (similar to BFS). The inner loop has decreaseKey() operation which
takes O(LogV) time. So overall time complexity is O(E+V)*O(LogV) which is O((E+V)*LogV) = O(ELogV)
Note that the above code uses Binary Heap for Priority Queue implementation. Time complexity can be reduced to O(E + VLogV) using
Fibonacci Heap. The reason is, Fibonacci Heap takes O(1) time for decrease-key operation while Binary Heap takes O(Logn) time.
Notes:
1)The code calculates shortest distance, but doesnt calculate the path information. We can create a parent array, update the parent array when
distance is updated (likeprims implementation) and use it show the shortest path from source to different vertices.
2)The code is for undirected graph, same dijekstra function can be used for directed graphs also.
3)The code finds shortest distances from source to all vertices. If we are interested only in shortest distance from source to a single target, we can
break the for loop when the picked minimum distance vertex is equal to target (Step 3.a of algorithm).
4)Dijkstras algorithm doesnt work for graphs with negative weight edges. For graphs with negative weight edges,BellmanFord algorithmcan be
used, we will soon be discussing it as a separate post.
References:
Introduction to Algorithms by Clifford Stein, Thomas H. Cormen, Charles E. Leiserson, Ronald L.
Algorithms by Sanjoy Dasgupta, Christos Papadimitriou, Umesh Vazirani

Dynamic Programming | Set 23 (BellmanFord Algorithm)

Given a graph and a source vertex src in graph, find shortest paths from src to all vertices in the given graph. The graph may contain negative
weight edges.
We have discussed Dijkstras algorithm for this problem. Dijksras algorithm is a Greedy algorithm and time complexity is O(VLogV) (with the use
of Fibonacci heap). Dijkstra doesnt work for Graphs with negative weight edges, Bellman-Ford works for such graphs. Bellman-Ford is
also simpler than Dijkstra and suites well for distributed systems. But time complexity of Bellman-Ford is O(VE), which is more than
Dijkstra.
Algorithm
Following are the detailed steps.
Input: Graph and a source vertex src
Output: Shortest distance to all vertices from src. If there is a negative weight cycle, then shortest distances are not calculated, negative weight
cycle is reported.
1) This step initializes distances from source to all vertices as infinite and distance to source itself as 0. Create an array dist[] of size |V| with all
values as infinite except dist[src] where src is source vertex.
2) This step calculates shortest distances. Do following |V|-1 times where |V| is the number of vertices in given graph.
..a) Do following for each edge u-v
If dist[v] > dist[u] + weight of edge uv, then update dist[v]
.dist[v] = dist[u] + weight of edge uv
3) This step reports if there is a negative weight cycle in graph. Do following for each edge u-v
If dist[v] > dist[u] + weight of edge uv, then Graph contains negative weight cycle
The idea of step 3 is, step 2 guarantees shortest distances if graph doesnt contain negative weight cycle. If we iterate through all edges one more
time and get a shorter path for any vertex, then there is a negative weight cycle
How does this work? Like other Dynamic Programming Problems, the algorithm calculate shortest paths in bottom-up manner. It first calculates
the shortest distances for the shortest paths which have at-most one edge in the path. Then, it calculates shortest paths with at-nost 2 edges, and so
on. After the ith iteration of outer loop, the shortest paths with at most i edges are calculated. There can be maximum |V| 1 edges in any simple
path, that is why the outer loop runs |v| 1 times. The idea is, assuming that there is no negative weight cycle, if we have calculated shortest paths
with at most i edges, then an iteration over all edges guarantees to give shortest path with at-most (i+1) edges (Proof is simple, you can refer this or
MIT Video Lecture)
Example
Let us understand the algorithm with following example graph. The images are taken from this source.
Let the given source vertex be 0. Initialize all distances as infinite, except the distance to source itself. Total number of vertices in the graph is 5, so
all edges must be processed 4 times.

Let all edges are processed in following order: (B,E), (D,B), (B,D), (A,B), (A,C), (D,C), (B,C), (E,D). We get following distances when all edges
are processed first time. The first row in shows initial distances. The second row shows distances when edges (B,E), (D,B), (B,D) and (A,B) are
processed. The third row shows distances when (A,C) is processed. The fourth row shows when (D,C), (B,C) and (E,D) are processed.

The first iteration guarantees to give all shortest paths which are at most 1 edge long. We get following distances when all edges are processed
second time (The last row shows final values).

The second iteration guarantees to give all shortest paths which are at most 2 edges long. The algorithm processes all edges 2 more times. The
distances are minimized after the second iteration, so third and fourth iterations dont update the distances.
Implementation:

C++
// A C / C++ program for Bellman-Ford's single source
// shortest path algorithm.
#include
#include
#include
#include

<stdio.h>
<stdlib.h>
<string.h>
<limits.h>

// a structure to represent a weighted edge in graph

struct Edge
{
int src, dest, weight;
};
// a structure to represent a connected, directed and
// weighted graph
struct Graph
{
// V-> Number of vertices, E-> Number of edges
int V, E;
// graph is represented as an array of edges.
struct Edge* edge;
};
// Creates a graph with V vertices and E edges
struct Graph* createGraph(int V, int E)
{
struct Graph* graph =
(struct Graph*) malloc( sizeof(struct Graph) );
graph->V = V;
graph->E = E;
graph->edge =
(struct Edge*) malloc( graph->E * sizeof( struct Edge ) );
return graph;
}
// A utility function used to print the solution
void printArr(int dist[], int n)
{
printf("Vertex Distance from Source\n");
for (int i = 0; i < n; ++i)
printf("%d \t\t %d\n", i, dist[i]);
}
// The main function that finds shortest distances from src to
// all other vertices using Bellman-Ford algorithm. The function
// also detects negative weight cycle
void BellmanFord(struct Graph* graph, int src)
{
int V = graph->V;
int E = graph->E;
int dist[V];
// Step 1: Initialize distances from src to all other vertices
// as INFINITE
for (int i = 0; i < V; i++)
dist[i] = INT_MAX;
dist[src] = 0;

// Step 2: Relax all edges |V| - 1 times. A simple shortest

// path from src to any other vertex can have at-most |V| - 1
// edges
for (int i = 1; i <= V-1; i++)
{
for (int j = 0; j < E; j++)
{
int u = graph->edge[j].src;
int v = graph->edge[j].dest;
int weight = graph->edge[j].weight;
if (dist[u] != INT_MAX && dist[u] + weight < dist[v])
dist[v] = dist[u] + weight;
}
}
// Step 3: check for negative-weight cycles. The above step
// guarantees shortest distances if graph doesn't contain
// negative weight cycle. If we get a shorter path, then there
// is a cycle.
for (int i = 0; i < E; i++)
{
int u = graph->edge[i].src;
int v = graph->edge[i].dest;
int weight = graph->edge[i].weight;
if (dist[u] != INT_MAX && dist[u] + weight < dist[v])
printf("Graph contains negative weight cycle");
}
printArr(dist, V);
return;
}
// Driver program to test above functions
int main()
{
/* Let us create the graph given in above example */
int V = 5; // Number of vertices in graph
int E = 8; // Number of edges in graph
struct Graph* graph = createGraph(V, E);
// add edge 0-1 (or A-B in above figure)
graph->edge[0].src = 0;
graph->edge[0].dest = 1;
graph->edge[0].weight = -1;
// add edge 0-2 (or A-C in above figure)
graph->edge[1].src = 0;
graph->edge[1].dest = 2;
graph->edge[1].weight = 4;
// add edge 1-2 (or B-C in above figure)
graph->edge[2].src = 1;
graph->edge[2].dest = 2;
graph->edge[2].weight = 3;
// add edge 1-3 (or B-D in above figure)
graph->edge[3].src = 1;
graph->edge[3].dest = 3;
graph->edge[3].weight = 2;
// add edge 1-4 (or A-E in above figure)
graph->edge[4].src = 1;
graph->edge[4].dest = 4;
graph->edge[4].weight = 2;
// add edge 3-2 (or D-C in above figure)
graph->edge[5].src = 3;
graph->edge[5].dest = 2;
graph->edge[5].weight = 5;
// add edge 3-1 (or D-B in above figure)
graph->edge[6].src = 3;
graph->edge[6].dest = 1;
graph->edge[6].weight = 1;
// add edge 4-3 (or E-D in above figure)
graph->edge[7].src = 4;
graph->edge[7].dest = 3;
graph->edge[7].weight = -3;

BellmanFord(graph, 0);
return 0;
}

Java
// A Java program for Bellman-Ford's single source shortest path
// algorithm.
import java.util.*;
import java.lang.*;
import java.io.*;
// A class to represent a connected, directed and weighted graph
class Graph
{
// A class to represent a weighted edge in graph
class Edge {
int src, dest, weight;
Edge() {
src = dest = weight = 0;
}
};
int V, E;
Edge edge[];
// Creates a graph with V vertices and E edges
Graph(int v, int e)
{
V = v;
E = e;
edge = new Edge[e];
for (int i=0; i<e; ++i)
edge[i] = new Edge();
}
// The main function that finds shortest distances from src
// to all other vertices using Bellman-Ford algorithm. The
// function also detects negative weight cycle
void BellmanFord(Graph graph,int src)
{
int V = graph.V, E = graph.E;
int dist[] = new int[V];
// Step 1: Initialize distances from src to all other
// vertices as INFINITE
for (int i=0; i<V; ++i)
dist[i] = Integer.MAX_VALUE;
dist[src] = 0;
// Step 2: Relax all edges |V| - 1 times. A simple
// shortest path from src to any other vertex can
// have at-most |V| - 1 edges
for (int i=1; i<V; ++i)
{
for (int j=0; j<E; ++j)
{
int u = graph.edge[j].src;
int v = graph.edge[j].dest;
int weight = graph.edge[j].weight;
if (dist[u]!=Integer.MAX_VALUE &&
dist[u]+weight<dist[v])
dist[v]=dist[u]+weight;
}
}
// Step 3: check for negative-weight cycles. The above
// step guarantees shortest distances if graph doesn't
// contain negative weight cycle. If we get a shorter
// path, then there is a cycle.
for (int j=0; j<E; ++j)
{
int u = graph.edge[j].src;
int v = graph.edge[j].dest;
int weight = graph.edge[j].weight;
if (dist[u]!=Integer.MAX_VALUE &&
dist[u]+weight<dist[v])
System.out.println("Graph contains negative weight cycle");

}
printArr(dist, V);
}
// A utility function used to print the solution
void printArr(int dist[], int V)
{
System.out.println("Vertex Distance from Source");
for (int i=0; i<V; ++i)
System.out.println(i+"\t\t"+dist[i]);
}
// Driver method to test above function
public static void main(String[] args)
{
int V = 5; // Number of vertices in graph
int E = 8; // Number of edges in graph
Graph graph = new Graph(V, E);
// add edge 0-1 (or A-B in above figure)
graph.edge[0].src = 0;
graph.edge[0].dest = 1;
graph.edge[0].weight = -1;
// add edge 0-2 (or A-C in above figure)
graph.edge[1].src = 0;
graph.edge[1].dest = 2;
graph.edge[1].weight = 4;
// add edge 1-2 (or B-C in above figure)
graph.edge[2].src = 1;
graph.edge[2].dest = 2;
graph.edge[2].weight = 3;
// add edge 1-3 (or B-D in above figure)
graph.edge[3].src = 1;
graph.edge[3].dest = 3;
graph.edge[3].weight = 2;
// add edge 1-4 (or A-E in above figure)
graph.edge[4].src = 1;
graph.edge[4].dest = 4;
graph.edge[4].weight = 2;
// add edge 3-2 (or D-C in above figure)
graph.edge[5].src = 3;
graph.edge[5].dest = 2;
graph.edge[5].weight = 5;
// add edge 3-1 (or D-B in above figure)
graph.edge[6].src = 3;
graph.edge[6].dest = 1;
graph.edge[6].weight = 1;
// add edge 4-3 (or E-D in above figure)
graph.edge[7].src = 4;
graph.edge[7].dest = 3;
graph.edge[7].weight = -3;
graph.BellmanFord(graph, 0);
}
}
// Contributed by Aakash Hasija

Vertex
0
1
2
3
4

Distance from Source

0
-1
2
-2
1

Notes
1) Negative weights are found in various applications of graphs. For example, instead of paying cost for a path, we may get some advantage if we
follow the path.
2) Bellman-Ford works better (better than Dijksras) for distributed systems. Unlike Dijksras where we need to find minimum value of all vertices,
in Bellman-Ford, edges are considered one by one.

Exercise
1) The standard Bellman-Ford algorithm reports shortest path only if there is no negative weight cycles. Modify it so that it reports minimum
distances even if there is a negative weight cycle.
2) Can we use Dijksras algorithm for shortest paths for graphs with negative weights one idea can be, calculate the minimum weight value, add a
positive value (equal to absolute value of minimum weight value) to all weights and run the Dijksras algorithm for the modified graph. Will this
algorithm work?
References:
http://www.youtube.com/watch?v=Ttezuzs39nk
http://en.wikipedia.org/wiki/Bellman%E2%80%93Ford_algorithm
http://www.cs.arizona.edu/classes/cs445/spring07/ShortestPath2.prn.pdf

Dynamic Programming | Set 16 (Floyd Warshall Algorithm)

The Floyd Warshall Algorithm is for solving the All Pairs Shortest Path problem. The problem is to find shortest distances between every pair of
vertices in a given edge weighted directed Graph.
Example:
Input:
graph[][] = { {0, 5, INF, 10},
{INF, 0, 3, INF},
{INF, INF, 0, 1},
{INF, INF, INF, 0} }
which represents the following graph
10
(0)------->(3)
|
/|\
5 |
|
|
| 1
\|/
|
(1)------->(2)
3
Note that the value of graph[i][j] is 0 if i is equal to j
And graph[i][j] is INF (infinite) if there is no edge from vertex i to j.
Output:
Shortest distance matrix
0
5
8
INF
0
3
INF
INF
0
INF
INF
INF

9
4
1
0

Floyd Warshall Algorithm

We initialize the solution matrix same as the input graph matrix as a first step. Then we update the solution matrix by considering all vertices as an
intermediate vertex. The idea is to one by one pick all vertices and update all shortest paths which include the picked vertex as an intermediate
vertex in the shortest path. When we pick vertex number k as an intermediate vertex, we already have considered vertices {0, 1, 2, .. k-1} as
intermediate vertices. For every pair (i, j) of source and destination vertices respectively, there are two possible cases.
1) k is not an intermediate vertex in shortest path from i to j. We keep the value of dist[i][j] as it is.
2) k is an intermediate vertex in shortest path from i to j. We update the value of dist[i][j] as dist[i][k] + dist[k][j].
The following figure is taken from the Cormen book. It shows the above optimal substructure property in the all-pairs shortest path problem.

Following is implementations of the Floyd Warshall algorithm.

C/C++
// C Program for Floyd Warshall Algorithm
#include<stdio.h>
// Number of vertices in the graph
#define V 4
/* Define Infinite as a large enough value. This value will be used
for vertices not connected to each other */
#define INF 99999
// A function to print the solution matrix
void printSolution(int dist[][V]);
// Solves the all-pairs shortest path problem using Floyd Warshall algorithm
void floydWarshell (int graph[][V])
{
/* dist[][] will be the output matrix that will finally have the shortest
distances between every pair of vertices */
int dist[V][V], i, j, k;
/* Initialize the solution matrix same as input graph matrix. Or
we can say the initial values of shortest distances are based
on shortest paths considering no intermediate vertex. */

for (i = 0; i < V; i++)

for (j = 0; j < V; j++)
dist[i][j] = graph[i][j];
/* Add all vertices one by one to the set of intermediate vertices.
---> Before start of a iteration, we have shortest distances between all
pairs of vertices such that the shortest distances consider only the
vertices in set {0, 1, 2, .. k-1} as intermediate vertices.
----> After the end of a iteration, vertex no. k is added to the set of
intermediate vertices and the set becomes {0, 1, 2, .. k} */
for (k = 0; k < V; k++)
{
// Pick all vertices as source one by one
for (i = 0; i < V; i++)
{
// Pick all vertices as destination for the
// above picked source
for (j = 0; j < V; j++)
{
// If vertex k is on the shortest path from
// i to j, then update the value of dist[i][j]
if (dist[i][k] + dist[k][j] < dist[i][j])
dist[i][j] = dist[i][k] + dist[k][j];
}
}
}
// Print the shortest distance matrix
printSolution(dist);
}
/* A utility function to print solution */
void printSolution(int dist[][V])
{
printf ("Following matrix shows the shortest distances"
" between every pair of vertices \n");
for (int i = 0; i < V; i++)
{
for (int j = 0; j < V; j++)
{
if (dist[i][j] == INF)
printf("%7s", "INF");
else
printf ("%7d", dist[i][j]);
}
printf("\n");
}
}
// driver program to test above function
int main()
{
/* Let us create the following weighted graph
10
(0)------->(3)
|
/|\
5 |
|
|
| 1
\|/
|
(1)------->(2)
3
*/
int graph[V][V] = { {0, 5, INF, 10},
{INF, 0, 3, INF},
{INF, INF, 0, 1},
{INF, INF, INF, 0}
};
// Print the solution
floydWarshell(graph);
return 0;
}

Java
// A Java program for Floyd Warshall All Pairs Shortest
// Path algorithm.
import java.util.*;
import java.lang.*;
import java.io.*;

class AllPairShortestPath
{
final static int INF = 99999, V = 4;
void floydWarshall(int graph[][])
{
int dist[][] = new int[V][V];
int i, j, k;
/* Initialize the solution matrix same as input graph matrix.
Or we can say the initial values of shortest distances
are based on shortest paths considering no intermediate
vertex. */
for (i = 0; i < V; i++)
for (j = 0; j < V; j++)
dist[i][j] = graph[i][j];
/* Add all vertices one by one to the set of intermediate
vertices.
---> Before start of a iteration, we have shortest
distances between all pairs of vertices such that
the shortest distances consider only the vertices in
set {0, 1, 2, .. k-1} as intermediate vertices.
----> After the end of a iteration, vertex no. k is added
to the set of intermediate vertices and the set
becomes {0, 1, 2, .. k} */
for (k = 0; k < V; k++)
{
// Pick all vertices as source one by one
for (i = 0; i < V; i++)
{
// Pick all vertices as destination for the
// above picked source
for (j = 0; j < V; j++)
{
// If vertex k is on the shortest path from
// i to j, then update the value of dist[i][j]
if (dist[i][k] + dist[k][j] < dist[i][j])
dist[i][j] = dist[i][k] + dist[k][j];
}
}
}
// Print the shortest distance matrix
printSolution(dist);
}
void printSolution(int dist[][])
{
System.out.println("Following matrix shows the shortest "+
"distances between every pair of vertices");
for (int i=0; i<V; ++i)
{
for (int j=0; j<V; ++j)
{
if (dist[i][j]==INF)
System.out.print("INF ");
else
System.out.print(dist[i][j]+" ");
}
System.out.println();
}
}
// Driver program to test above function
public static void main (String[] args)
{
/* Let us create the following weighted graph
10
(0)------->(3)
|
/|\
5 |
|
|
| 1
\|/
|
(1)------->(2)
3
*/
int graph[][] = { {0, 5, INF, 10},
{INF, 0, 3, INF},
{INF, INF, 0, 1},
{INF, INF, INF, 0}

};
AllPairShortestPath a = new AllPairShortestPath();
// Print the solution
a.floydWarshall(graph);
}
}
// Contributed by Aakash Hasija

Output:
Following matrix shows the shortest distances between every pair of vertices
0
5
8
9
INF
0
3
4
INF
INF
0
1
INF
INF
INF
0

Time Complexity: O(V^3)

The above program only prints the shortest distances. We can modify the solution to print the shortest paths also by storing the predecessor
information in a separate 2D matrix.
Also, the value of INF can be taken as INT_MAX from limits.h to make sure that we handle maximum possible value. When we take INF as
INT_MAX, we need to change the if condition in the above program to avoid arithmatic overflow.
#include<limits.h>
#define INF INT_MAX
..........................
if (dist[i][k] != INF && dist[k][j] != INF && dist[i][k] + dist[k][j] < dist[i][j])
dist[i][j] = dist[i][k] + dist[k][j];
...........................

Johnsons algorithm for All-pairs shortest paths

The problem is to find shortest paths between every pair of vertices in a given weighted directed Graph and weights may be negative. We have
discussed Floyd Warshall Algorithm for this problem. Time complexity of Floyd Warshall Algorithm is ?(V3). Using Johnsons algorithm, we can
find all pair shortest paths in O(V2log V + VE) time. Johnsons algorithm uses both Dijkstra and Bellman-Ford as subroutines.
If we apply Dijkstras Single Source shortest path algorithm for every vertex, considering every vertex as source, we can find all pair shortest paths
in O(V*VLogV) time. So using Dijkstras single source shortest path seems to be a better option than Floyd Warshell, but the problem with
Dijkstras algorithm is, it doesnt work for negative weight edge.
The idea of Johnsons algorithm is to re-weight all edges and make them all positive, then apply Dijkstras algorithm for every vertex.
How to transform a given graph to a graph with all non-negative weight edges?
One may think of a simple approach of finding the minimum weight edge and adding this weight to all edges. Unfortunately, this doesnt work as
there may be different number of edges in different paths (See this for an example). If there are multiple paths from a vertex u to v, then all paths
must be increased by same amount, so that the shortest path remains the shortest in the transformed graph.
The idea of Johnsons algorithm is to assign a weight to every vertex. Let the weight assigned to vertex u be h[u]. We reweight edges using vertex
weights. For example, for an edge (u, v) of weight w(u, v), the new weight becomes w(u, v) + h[u] h[v]. The great thing about this reweighting is,
all set of paths between any two vertices are increased by same amount and all negative weights become non-negative. Consider any path between
two vertices s and t, weight of every path is increased by h[s] h[t], all h[] values of vertices on path from s to t cancel each other.
How do we calculate h[] values? Bellman-Ford algorithm is used for this purpose. Following is the complete algorithm. A new vertex is added to
the graph and connected to all existing vertices. The shortest distance values from new vertex to all existing vertices are h[] values.
Algorithm:
1) Let the given graph be G. Add a new vertex s to the graph, add edges from new vertex to all vertices of G. Let the modified graph be G.
2) Run Bellman-Ford algorithm on G with s as source. Let the distances calculated by Bellman-Ford be h[0], h[1], .. h[V-1]. If we find a negative
weight cycle, then return. Note that the negative weight cycle cannot be created by new vertex s as there is no edge to s. All edges are from s.
3) Reweight the edges of original graph. For each edge (u, v), assign the new weight as original weight + h[u] h[v].
4) Remove the added vertex s and run Dijkstras algorithm for every vertex.
How does the transformation ensure nonnegative weight edges?
The following property is always true about h[] values as they are shortest distances.
h[v] <= h[u] + w(u, v)

The property simply means, shortest distance from s to v must be smaller than or equal to shortest distance from s to u plus weight of edge (u, v).
The new weights are w(u, v) + h[u] - h[v]. The value of the new weights must be greater than or equal to zero because of the inequality "h[v] <=
h[u] + w(u, v)". Example:
Let us consider the following graph.

We add a source s and add edges from s to all vertices of the original graph. In the following diagram s is 4.

We calculate the shortest distances from 4 to all other vertices using Bellman-Ford algorithm. The shortest distances from 4 to 0, 1, 2 and 3 are 0,
-5, -1 and 0 respectively, i.e., h[] = {0, -5, -1, 0}. Once we get these distances, we remove the source vertex 4 and reweight the edges using
following formula. w(u, v) = w(u, v) + h[u] - h[v].

Since all weights are positive now, we can run Dijkstra's shortest path algorithm for every vertex as source.
Time Complexity: The main steps in algorithm are Bellman Ford Algorithm called once and Dijkstra called V times. Time complexity of Bellman
Ford is O(VE) and time complexity of Dijkstra is O(VLogV). So overall time complexity is O(V2log V + VE).
The time complexity of Johnson's algorithm becomes same as Floyd Warshell when the graphs is complete (For a complete graph E = O(V2). But
for sparse graphs, the algorithm performs much better than Floyd Warshell.
References:
Introduction to Algorithms 3rd Edition by Clifford Stein, Thomas H. Cormen, Charles E. Leiserson, Ronald L. Rivest
http://www.youtube.com/watch?v=b6LOHvCzmkI
http://www.youtube.com/watch?v=TV2Z6nbo1ic
http://en.wikipedia.org/wiki/Johnson%27s_algorithm
http://www.youtube.com/watch?v=Sygq1e0xWnM

Shortest Path in Directed Acyclic Graph

Given a Weighted Directed Acyclic Graph and a source vertex in the graph, find the shortest paths from given source to all other vertices.
For a general weighted graph, we can calculate single source shortest distances in O(VE) time using BellmanFord Algorithm. For a graph with no
negative weights, we can do better and calculate single source shortest distances in O(E + VLogV) time using Dijkstras algorithm. Can we do even
better for Directed Acyclic Graph (DAG)? We can calculate single source shortest distances in O(V+E) time for DAGs. The idea is to use
Topological Sorting.
We initialize distances to all vertices as infinite and distance to source as 0, then we find a topological sorting of the graph. Topological Sorting of a
graph represents a linear ordering of the graph (See below, figure (b) is a linear representation of figure (a) ). Once we have topological order (or
linear representation), we one by one process all vertices in topological order. For every vertex being processed, we update distances of its
adjacent using distance of current vertex.
Following figure is taken from this source. It shows step by step process of finding shortest paths.

Following is complete algorithm for finding shortest distances.

1) Initialize dist[] = {INF, INF, .} and dist[s] = 0 where s is the source vertex.
2) Create a toplogical order of all vertices.
3) Do following for every vertex u in topological order.
..Do following for every adjacent vertex v of u
if (dist[v] > dist[u] + weight(u, v))
dist[v] = dist[u] + weight(u, v)

C++
// C++ program to find single source shortest paths for Directed Acyclic Graphs
#include<iostream>
#include <list>
#include <stack>
#include <limits.h>
#define INF INT_MAX
using namespace std;
// Graph is represented using adjacency list. Every node of adjacency list
// contains vertex number of the vertex to which edge connects. It also
// contains weight of the edge
class AdjListNode
{
int v;

int weight;
public:
AdjListNode(int _v, int _w) { v = _v; weight = _w;}
int getV()
{ return v; }
int getWeight() { return weight; }
};
// Class to represent a graph using adjacency list representation
class Graph
{
int V;
// No. of vertices'
// Pointer to an array containing adjacency lists
list<AdjListNode> *adj;
// A function used by shortestPath
void topologicalSortUtil(int v, bool visited[], stack<int> &Stack);
public:
Graph(int V); // Constructor
// function to add an edge to graph
void addEdge(int u, int v, int weight);
// Finds shortest paths from given source vertex
void shortestPath(int s);
};
Graph::Graph(int V)
{
this->V = V;
adj = new list<AdjListNode>[V];
}
void Graph::addEdge(int u, int v, int weight)
{
AdjListNode node(v, weight);
adj[u].push_back(node); // Add v to u's list
}
// A recursive function used by shortestPath. See below link for details
// http://www.geeksforgeeks.org/topological-sorting/
void Graph::topologicalSortUtil(int v, bool visited[], stack<int> &Stack)
{
// Mark the current node as visited
visited[v] = true;
// Recur for all the vertices adjacent to this vertex
list<AdjListNode>::iterator i;
for (i = adj[v].begin(); i != adj[v].end(); ++i)
{
AdjListNode node = *i;
if (!visited[node.getV()])
topologicalSortUtil(node.getV(), visited, Stack);
}
// Push current vertex to stack which stores topological sort
Stack.push(v);
}
// The function to find shortest paths from given vertex. It uses recursive
// topologicalSortUtil() to get topological sorting of given graph.
void Graph::shortestPath(int s)
{
stack<int> Stack;
int dist[V];
// Mark all the vertices as not visited
bool *visited = new bool[V];
for (int i = 0; i < V; i++)
visited[i] = false;
// Call the recursive helper function to store Topological Sort
// starting from all vertices one by one
for (int i = 0; i < V; i++)
if (visited[i] == false)
topologicalSortUtil(i, visited, Stack);
// Initialize distances to all vertices as infinite and distance
// to source as 0
for (int i = 0; i < V; i++)
dist[i] = INF;

dist[s] = 0;
// Process vertices in topological order
while (Stack.empty() == false)
{
// Get the next vertex from topological order
int u = Stack.top();
Stack.pop();
// Update distances of all adjacent vertices
list<AdjListNode>::iterator i;
if (dist[u] != INF)
{
for (i = adj[u].begin(); i != adj[u].end(); ++i)
if (dist[i->getV()] > dist[u] + i->getWeight())
dist[i->getV()] = dist[u] + i->getWeight();
}
}
// Print the calculated shortest distances
for (int i = 0; i < V; i++)
(dist[i] == INF)? cout << "INF ": cout << dist[i] << " ";
}
// Driver program to test above functions
int main()
{
// Create a graph given in the above diagram. Here vertex numbers are
// 0, 1, 2, 3, 4, 5 with following mappings:
// 0=r, 1=s, 2=t, 3=x, 4=y, 5=z
Graph g(6);
g.addEdge(0, 1, 5);
g.addEdge(0, 2, 3);
g.addEdge(1, 3, 6);
g.addEdge(1, 2, 2);
g.addEdge(2, 4, 4);
g.addEdge(2, 5, 2);
g.addEdge(2, 3, 7);
g.addEdge(3, 4, -1);
g.addEdge(4, 5, -2);
int s = 1;
cout << "Following are shortest distances from source " << s <<" \n";
g.shortestPath(s);
return 0;
}

Java
// Java program to find single source shortest paths in Directed Acyclic Graphs
import java.io.*;
import java.util.*;
class ShortestPath
{
static final int INF=Integer.MAX_VALUE;
class AdjListNode
{
private int v;
private int weight;
AdjListNode(int _v, int _w) { v = _v; weight = _w; }
int getV() { return v; }
int getWeight() { return weight; }
}
// Class to represent graph as an adjcency list of
// nodes of type AdjListNode
class Graph
{
private int V;
private LinkedList<AdjListNode>adj[];
Graph(int v)
{
V=v;
adj = new LinkedList[V];
for (int i=0; i<v; ++i)
adj[i] = new LinkedList<AdjListNode>();
}
void addEdge(int u, int v, int weight)

{
AdjListNode node = new AdjListNode(v,weight);
adj[u].add(node);// Add v to u's list
}
// A recursive function used by shortestPath.
// See below link for details
void topologicalSortUtil(int v, Boolean visited[], Stack stack)
{
// Mark the current node as visited.
visited[v] = true;
Integer i;
// Recur for all the vertices adjacent to this vertex
Iterator<AdjListNode> it = adj[v].iterator();
while (it.hasNext())
{
AdjListNode node =it.next();
if (!visited[node.getV()])
topologicalSortUtil(node.getV(), visited, stack);
}
// Push current vertex to stack which stores result
stack.push(new Integer(v));
}
// The function to find shortest paths from given vertex. It
// uses recursive topologicalSortUtil() to get topological
// sorting of given graph.
void shortestPath(int s)
{
Stack stack = new Stack();
int dist[] = new int[V];
// Mark all the vertices as not visited
Boolean visited[] = new Boolean[V];
for (int i = 0; i < V; i++)
visited[i] = false;
// Call the recursive helper function to store Topological
// Sort starting from all vertices one by one
for (int i = 0; i < V; i++)
if (visited[i] == false)
topologicalSortUtil(i, visited, stack);
// Initialize distances to all vertices as infinite and
// distance to source as 0
for (int i = 0; i < V; i++)
dist[i] = INF;
dist[s] = 0;
// Process vertices in topological order
while (stack.empty() == false)
{
// Get the next vertex from topological order
int u = (int)stack.pop();
// Update distances of all adjacent vertices
Iterator<AdjListNode> it;
if (dist[u] != INF)
{
it = adj[u].iterator();
while (it.hasNext())
{
AdjListNode i= it.next();
if (dist[i.getV()] > dist[u] + i.getWeight())
dist[i.getV()] = dist[u] + i.getWeight();
}
}
}
// Print the calculated shortest distances
for (int i = 0; i < V; i++)
{
if (dist[i] == INF)
System.out.print( "INF ");
else
System.out.print( dist[i] + " ");
}
}
}

// Method to create a new graph instance through an object

// of ShortestPath class.
Graph newGraph(int number)
{
return new Graph(number);
}
public static void main(String args[])
{
// Create a graph given in the above diagram. Here vertex
// numbers are 0, 1, 2, 3, 4, 5 with following mappings:
// 0=r, 1=s, 2=t, 3=x, 4=y, 5=z
ShortestPath t = new ShortestPath();
Graph g = t.newGraph(6);
g.addEdge(0, 1, 5);
g.addEdge(0, 2, 3);
g.addEdge(1, 3, 6);
g.addEdge(1, 2, 2);
g.addEdge(2, 4, 4);
g.addEdge(2, 5, 2);
g.addEdge(2, 3, 7);
g.addEdge(3, 4, -1);
g.addEdge(4, 5, -2);
int s = 1;
System.out.println("Following are shortest distances "+
"from source " + s );
g.shortestPath(s);
}
}
//This code is contributed by Aakash Hasija

Following are shortest distances from source 1

INF 0 2 6 5 3

Time Complexity: Time complexity of topological sorting is O(V+E). After finding topological order, the algorithm process all vertices and for
every vertex, it runs a loop for all adjacent vertices. Total adjacent vertices in a graph is O(E). So the inner loop runs O(V+E) times. Therefore,
overall time complexity of this algorithm is O(V+E).
References:
http://www.utdallas.edu/~sizheng/CS4349.d/l-notes.d/L17.pdf

Some interesting shortest path questions | Set 1

Question 1: Given a directed weighted graph. You are also given the shortest path from a source vertex s to a destination vertex t. If
weight of every edge is increased by 10 units, does the shortest path remain same in the modified graph?
The shortest path may change. The reason is, there may be different number of edges in different paths from s to t. For example, let shortest path
be of weight 15 and has 5 edges. Let there be another path with 2 edges and total weight 25. The weight of the shortest path is increased by 5*10
and becomes 15 + 50. Weight of the other path is increased by 2*10 and becomes 25 + 20. So the shortest path changes to the other path with
weight as 45.
Question 2:This is similar to above question. Does the shortest path change when weights of all edges are multiplied by 10?
If we multiply all edge weights by 10, the shortest path doesnt change. The reason is simple, weights of all paths from s to t get multiplied by same
amount. The number of edges on a path doesnt matter. It is like changing unit of weights.
Question 3:Given a directed graph where every edge has weight as either 1 or 2, find the shortest path from a given source vertex s to
a given destination vertex t. Expected time complexity is O(V+E).
If we apply Dijkstras shortest path algorithm, we can get a shortest path in O(E + VLogV) time. How to do it in O(V+E) time? The idea is to use
BFS . One important observation about BFS is, the path used in BFS always has least number of edges between any two vertices. So if all edges
are of same weight, we can use BFS to find the shortest path. For this problem, we can modify the graph and split all edges of weight 2 into two
edges of weight 1 each. In the modified graph, we can use BFS to find the shortest path. How is this approach O(V+E)? In worst case, all edges
are of weight 2 and we need to do O(E) operations to split all edges, so the time complexity becomes O(E) + O(V+E) which is O(V+E).
Question 4:Given a directed acyclic weighted graph, how to find the shortest path from a source s to a destination t in O(V+E) time?
See: Shortest Path in Directed Acyclic Graph
More Questions See following links for more questions.
http://algs4.cs.princeton.edu/44sp/
http://geeksquiz.com/graph-shortest-paths/

Shortest path with exactly k edges in a directed and weighted graph

Given a directed and two vertices u and v in it, find shortest path from u to v with exactly k edges on the path.
The graph is given as adjacency matrix representation where value of graph[i][j] indicates the weight of an edge from vertex i to vertex j and a
value INF(infinite) indicates no edge from i to j.
For example consider the following graph. Let source u be vertex 0, destination v be 3 and k be 2. There are two walks of length 2, the walks are
{0, 2, 3} and {0, 1, 3}. The shortest among the two is {0, 2, 3} and weight of path is 3+6 = 9.

The idea is to browse through all paths of length k from u to v using the approach discussed in the previous post and return weight of the shortest
path. A simple solution is to start from u, go to all adjacent vertices and recur for adjacent vertices with k as k-1, source as adjacent vertex and
destination as v. Following are C++ and Java implementations of this simple solution.

C++
// C++ program to find shortest path with exactly k edges
#include <iostream>
#include <climits>
using namespace std;
// Define number of vertices in the graph and inifinite value
#define V 4
#define INF INT_MAX
// A naive recursive function to count walks from u to v with k edges
int shortestPath(int graph[][V], int u, int v, int k)
{
// Base cases
if (k == 0 && u == v)
return 0;
if (k == 1 && graph[u][v] != INF) return graph[u][v];
if (k <= 0)
return INF;
// Initialize result
int res = INF;
// Go to all adjacents of u and recur
for (int i = 0; i < V; i++)
{
if (graph[u][i] != INF && u != i && v != i)
{
int rec_res = shortestPath(graph, i, v, k-1);
if (rec_res != INF)
res = min(res, graph[u][i] + rec_res);
}
}
return res;
}
// driver program to test above function
int main()
{
/* Let us create the graph shown in above diagram*/
int graph[V][V] = { {0, 10, 3, 2},
{INF, 0, INF, 7},
{INF, INF, 0, 6},
{INF, INF, INF, 0}
};
int u = 0, v = 3, k = 2;
cout << "Weight of the shortest path is " <<
shortestPath(graph, u, v, k);
return 0;
}

Java
// Dynamic Programming based Java program to find shortest path
// with exactly k edges
import java.util.*;
import java.lang.*;
import java.io.*;
class ShortestPath
{
// Define number of vertices in the graph and inifinite value
static final int V = 4;
static final int INF = Integer.MAX_VALUE;
// A naive recursive function to count walks from u to v
// with k edges
int shortestPath(int graph[][], int u, int v, int k)
{
// Base cases
if (k == 0 && u == v)
return 0;
if (k == 1 && graph[u][v] != INF) return graph[u][v];
if (k <= 0)
return INF;
// Initialize result
int res = INF;
// Go to all adjacents of u and recur
for (int i = 0; i < V; i++)
{
if (graph[u][i] != INF && u != i && v != i)
{
int rec_res = shortestPath(graph, i, v, k-1);
if (rec_res != INF)
res = Math.min(res, graph[u][i] + rec_res);
}
}
return res;
}
public static void main (String[] args)
{
/* Let us create the graph shown in above diagram*/
int graph[][] = new int[][]{ {0, 10, 3, 2},
{INF, 0, INF, 7},
{INF, INF, 0, 6},
{INF, INF, INF, 0}
};
ShortestPath t = new ShortestPath();
int u = 0, v = 3, k = 2;
System.out.println("Weight of the shortest path is "+
t.shortestPath(graph, u, v, k));
}
}

Weight of the shortest path is 9

The worst case time complexity of the above function is O(Vk) where V is the number of vertices in the given graph. We can simply analyze the
time complexity by drawing recursion tree. The worst occurs for a complete graph. In worst case, every internal node of recursion tree would have
exactly V children.
We can optimize the above solution using Dynamic Programming. The idea is to build a 3D table where first dimension is source, second
dimension is destination, third dimension is number of edges from source to destination, and the value is count of walks. Like other Dynamic
Programming problems, we fill the 3D table in bottom up manner.

C++
// Dynamic Programming based C++ program to find shortest path with
// exactly k edges
#include <iostream>
#include <climits>
using namespace std;
// Define number of vertices in the graph and inifinite value
#define V 4
#define INF INT_MAX
// A Dynamic programming based function to find the shortest path from

// u to v with exactly k edges.

int shortestPath(int graph[][V], int u, int v, int k)
{
// Table to be filled up using DP. The value sp[i][j][e] will store
// weight of the shortest path from i to j with exactly k edges
int sp[V][V][k+1];
// Loop for number of edges from 0 to k
for (int e = 0; e <= k; e++)
{
for (int i = 0; i < V; i++) // for source
{
for (int j = 0; j < V; j++) // for destination
{
// initialize value
sp[i][j][e] = INF;
// from base cases
if (e == 0 && i == j)
sp[i][j][e] = 0;
if (e == 1 && graph[i][j] != INF)
sp[i][j][e] = graph[i][j];
//go to adjacent only when number of edges is more than 1
if (e > 1)
{
for (int a = 0; a < V; a++)
{
// There should be an edge from i to a and a
// should not be same as either i or j
if (graph[i][a] != INF && i != a &&
j!= a && sp[a][j][e-1] != INF)
sp[i][j][e] = min(sp[i][j][e], graph[i][a] +
sp[a][j][e-1]);
}
}
}
}
}
return sp[u][v][k];
}
// driver program to test above function
int main()
{
/* Let us create the graph shown in above diagram*/
int graph[V][V] = { {0, 10, 3, 2},
{INF, 0, INF, 7},
{INF, INF, 0, 6},
{INF, INF, INF, 0}
};
int u = 0, v = 3, k = 2;
cout << shortestPath(graph, u, v, k);
return 0;
}

Java
// Dynamic Programming based Java program to find shortest path with
// exactly k edges
import java.util.*;
import java.lang.*;
import java.io.*;
class ShortestPath
{
// Define number of vertices in the graph and inifinite value
static final int V = 4;
static final int INF = Integer.MAX_VALUE;
// A Dynamic programming based function to find the shortest path
// from u to v with exactly k edges.
int shortestPath(int graph[][], int u, int v, int k)
{
// Table to be filled up using DP. The value sp[i][j][e] will
// store weight of the shortest path from i to j with exactly
// k edges
int sp[][][] = new int[V][V][k+1];
// Loop for number of edges from 0 to k

for (int e = 0; e <= k; e++)

{
for (int i = 0; i < V; i++) // for source
{
for (int j = 0; j < V; j++) // for destination
{
// initialize value
sp[i][j][e] = INF;
// from base cases
if (e == 0 && i == j)
sp[i][j][e] = 0;
if (e == 1 && graph[i][j] != INF)
sp[i][j][e] = graph[i][j];
// go to adjacent only when number of edges is
// more than 1
if (e > 1)
{
for (int a = 0; a < V; a++)
{
// There should be an edge from i to a and
// a should not be same as either i or j
if (graph[i][a] != INF && i != a &&
j!= a && sp[a][j][e-1] != INF)
sp[i][j][e] = Math.min(sp[i][j][e],
graph[i][a] + sp[a][j][e-1]);
}
}
}
}
}
return sp[u][v][k];
}
public static void main (String[] args)
{
/* Let us create the graph shown in above diagram*/
int graph[][] = new int[][]{ {0, 10, 3, 2},
{INF, 0, INF, 7},
{INF, INF, 0, 6},
{INF, INF, INF, 0}
};
ShortestPath t = new ShortestPath();
int u = 0, v = 3, k = 2;
System.out.println("Weight of the shortest path is "+
t.shortestPath(graph, u, v, k));
}
}
//This code is contributed by Aakash Hasija

Weight of the shortest path is 9

Time complexity of the above DP based solution is O(V3K) which is much better than the naive solution.

Find if there is a path between two vertices in a directed graph

Given a Directed Graph and two vertices in it, check whether there is a path from the first given vertex to second. For example, in the following
graph, there is a path from vertex 1 to 3. As another example, there is no path from 3 to 0.

We can either use Breadth First Search (BFS) or Depth First Search (DFS) to find path between two vertices. Take the first vertex as source in
BFS (or DFS), follow the standard BFS (or DFS). If we see the second vertex in our traversal, then return true. Else return false.
Following are C++ and Java codes that use BFS for finding reachability of second vertex from first vertex.

C++
// C++ program to check if there is exist a path between two vertices
// of a graph.
#include<iostream>
#include <list>
using namespace std;
// This class represents a directed graph using adjacency list
// representation
class Graph
{
int V;
// No. of vertices
list<int> *adj;
// Pointer to an array containing adjacency lists
public:
Graph(int V); // Constructor
void addEdge(int v, int w); // function to add an edge to graph
bool isReachable(int s, int d);
};
Graph::Graph(int V)
{
this->V = V;
adj = new list<int>[V];
}
void Graph::addEdge(int v, int w)
{
adj[v].push_back(w); // Add w to vs list.
}
// A BFS based function to check whether d is reachable from s.
bool Graph::isReachable(int s, int d)
{
// Base case
if (s == d)
return true;
// Mark all the vertices as not visited
bool *visited = new bool[V];
for (int i = 0; i < V; i++)
visited[i] = false;
// Create a queue for BFS
list<int> queue;
// Mark the current node as visited and enqueue it
visited[s] = true;
queue.push_back(s);
// it will be used to get all adjacent vertices of a vertex
list<int>::iterator i;
while (!queue.empty())
{
// Dequeue a vertex from queue and print it

s = queue.front();
queue.pop_front();
// Get all adjacent vertices of the dequeued vertex s
// If a adjacent has not been visited, then mark it visited
// and enqueue it
for (i = adj[s].begin(); i != adj[s].end(); ++i)
{
// If this adjacent node is the destination node, then
// return true
if (*i == d)
return true;
// Else, continue to do BFS
if (!visited[*i])
{
visited[*i] = true;
queue.push_back(*i);
}
}
}
// If BFS is complete without visiting d
return false;
}
// Driver program to test methods of graph class
int main()
{
// Create a graph given in the above diagram
Graph g(4);
g.addEdge(0, 1);
g.addEdge(0, 2);
g.addEdge(1, 2);
g.addEdge(2, 0);
g.addEdge(2, 3);
g.addEdge(3, 3);
int u = 1, v = 3;
if(g.isReachable(u, v))
cout<< "\n There is a path from " << u << " to " << v;
else
cout<< "\n There is no path from " << u << " to " << v;
u = 3, v = 1;
if(g.isReachable(u, v))
cout<< "\n There is a path from " << u << " to " << v;
else
cout<< "\n There is no path from " << u << " to " << v;
return 0;
}

Java
// Java program to check if there is exist a path between two vertices
// of a graph.
import java.io.*;
import java.util.*;
import java.util.LinkedList;
// This class represents a directed graph using adjacency list
// representation
class Graph
{
private int V; // No. of vertices
private LinkedList<Integer> adj[]; //Adjacency List
//Constructor
Graph(int v)
{
V = v;
adj = new LinkedList[v];
for (int i=0; i<v; ++i)
adj[i] = new LinkedList();
}
//Function to add an edge into the graph
void addEdge(int v,int w) { adj[v].add(w);

//prints BFS traversal from a given source s

Boolean isReachable(int s, int d)
{
LinkedList<Integer>temp;
// Mark all the vertices as not visited(By default set
// as false)
boolean visited[] = new boolean[V];
// Create a queue for BFS
LinkedList<Integer> queue = new LinkedList<Integer>();
// Mark the current node as visited and enqueue it
visited[s]=true;
queue.add(s);
// 'i' will be used to get all adjacent vertices of a vertex
Iterator<Integer> i;
while (queue.size()!=0)
{
// Dequeue a vertex from queue and print it
s = queue.poll();
int n;
i = adj[s].listIterator();
// Get all adjacent vertices of the dequeued vertex s
// If a adjacent has not been visited, then mark it
// visited and enqueue it
while (i.hasNext())
{
n = i.next();
// If this adjacent node is the destination node,
// then return true
if (n==d)
return true;
// Else, continue to do BFS
if (!visited[n])
{
visited[n] = true;
queue.add(n);
}
}
}
// If BFS is complete without visited d
return false;
}
// Driver method
public static void main(String args[])
{
// Create a graph given in the above diagram
Graph g = new Graph(4);
g.addEdge(0, 1);
g.addEdge(0, 2);
g.addEdge(1, 2);
g.addEdge(2, 0);
g.addEdge(2, 3);
g.addEdge(3, 3);
int u = 1;
int v = 3;
if (g.isReachable(u, v))
System.out.println("There is a path from " + u +" to " + v);
else
System.out.println("There is no path from " + u +" to " + v);;
u = 3;
v = 1;
if (g.isReachable(u, v))
System.out.println("There is a path from " + u +" to " + v);
else
System.out.println("There is no path from " + u +" to " + v);;
}
}
// This code is contributed by Aakash Hasija

There is a path from 1 to 3

There is no path from 3 to 1

As an exercise, try an extended version of the problem where the complete path between two vertices is also needed.

Connectivity in a directed graph

Given a directed graph, find out whether the graph is strongly connected or not. A directed graph is strongly connected if there is a path between
any two pair of vertices. For example, following is a strongly connected graph.

It is easy for undirected graph, we can just do a BFS and DFS starting from any vertex. If BFS or DFS visits all vertices, then the given
undirected graph is connected. This approach wont work for a directed graph. For example, consider the following graph which is not strongly
connected. If we start DFS (or BFS) from vertex 0, we can reach all vertices, but if we start from any other vertex, we cannot reach all vertices.

How to do for directed graph?

A simple idea is to use a all pair shortest path algorithm like Floyd Warshall or find Transitive Closure of graph. Time complexity of this
method would be O(v3).
We can also do DFS V times starting from every vertex. If any DFS, doesnt visit all vertices, then graph is not strongly connected. This algorithm
takes O(V*(V+E)) time which can be same as transitive closure for a dense graph.
A better idea can be Strongly Connected Components (SCC) algorithm. We can find all SCCs in O(V+E) time. If number of SCCs is one,
then graph is strongly connected. The algorithm for SCC does extra work as it finds all SCCs.
Following is Kosarajus DFS based simple algorithm that does two DFS traversals of graph:
1) Initialize all vertices as not visited.
2) Do a DFS traversal of graph starting from any arbitrary vertex v. If DFS traversal doesnt visit all vertices, then return false.
3) Reverse all arcs (or find transpose or reverse of graph)
4) Mark all vertices as not-visited in reversed graph.
5) Do a DFS traversal of reversed graph starting from same vertex v (Same as step 2). If DFS traversal doesnt visit all vertices, then return false.
Otherwise return true.
The idea is, if every node can be reached from a vertex v, and every node can reach v, then the graph is strongly connected. In step 2, we check if
all vertices are reachable from v. In step 4, we check if all vertices can reach v (In reversed graph, if all vertices are reachable from v, then all
vertices can reach v in original graph).
Following is C++ implementation of above algorithm.

C++
// C++ program to check if a given directed graph is strongly
// connected or not
#include <iostream>
#include <list>
#include <stack>
using namespace std;
class Graph
{
int V;
// No. of vertices
list<int> *adj;
// An array of adjacency lists
// A recursive function to print DFS starting from v
void DFSUtil(int v, bool visited[]);
public:
// Constructor and Destructor
Graph(int V) { this->V = V; adj = new list<int>[V];}

~Graph() { delete [] adj; }

// Method to add an edge
void addEdge(int v, int w);
// The main function that returns true if the graph is strongly
// connected, otherwise false
bool isSC();
// Function that returns reverse (or transpose) of this graph
Graph getTranspose();
};
// A recursive function to print DFS starting from v
void Graph::DFSUtil(int v, bool visited[])
{
// Mark the current node as visited and print it
visited[v] = true;
// Recur for all the vertices adjacent to this vertex
list<int>::iterator i;
for (i = adj[v].begin(); i != adj[v].end(); ++i)
if (!visited[*i])
DFSUtil(*i, visited);
}
// Function that returns reverse (or transpose) of this graph
Graph Graph::getTranspose()
{
Graph g(V);
for (int v = 0; v < V; v++)
{
// Recur for all the vertices adjacent to this vertex
list<int>::iterator i;
for(i = adj[v].begin(); i != adj[v].end(); ++i)
{
g.adj[*i].push_back(v);
}
}
return g;
}
void Graph::addEdge(int v, int w)
{
adj[v].push_back(w); // Add w to vs list.
}
// The main function that returns true if graph is strongly connected
bool Graph::isSC()
{
// St1p 1: Mark all the vertices as not visited (For first DFS)
bool visited[V];
for (int i = 0; i < V; i++)
visited[i] = false;
// Step 2: Do DFS traversal starting from first vertex.
DFSUtil(0, visited);
// If DFS traversal doesnt visit all vertices, then return false.
for (int i = 0; i < V; i++)
if (visited[i] == false)
return false;
// Step 3: Create a reversed graph
Graph gr = getTranspose();
// Step 4: Mark all the vertices as not visited (For second DFS)
for(int i = 0; i < V; i++)
visited[i] = false;
// Step 5: Do DFS for reversed graph starting from first vertex.
// Staring Vertex must be same starting point of first DFS
gr.DFSUtil(0, visited);
// If all vertices are not visited in second DFS, then
// return false
for (int i = 0; i < V; i++)
if (visited[i] == false)
return false;
return true;

}
// Driver program to test above functions
int main()
{
// Create graphs given in the above diagrams
Graph g1(5);
g1.addEdge(0, 1);
g1.addEdge(1, 2);
g1.addEdge(2, 3);
g1.addEdge(3, 0);
g1.addEdge(2, 4);
g1.addEdge(4, 2);
g1.isSC()? cout << "Yes\n" : cout << "No\n";
Graph g2(4);
g2.addEdge(0, 1);
g2.addEdge(1, 2);
g2.addEdge(2, 3);
g2.isSC()? cout << "Yes\n" : cout << "No\n";
return 0;
}

Java
// Java program to check if a given directed graph is strongly
// connected or not
import java.io.*;
import java.util.*;
import java.util.LinkedList;
// This class represents a directed graph using adjacency
// list representation
class Graph
{
private int V; // No. of vertices
private LinkedList<Integer> adj[]; //Adjacency List
//Constructor
Graph(int v)
{
V = v;
adj = new LinkedList[v];
for (int i=0; i<v; ++i)
adj[i] = new LinkedList();
}
//Function to add an edge into the graph
void addEdge(int v,int w) { adj[v].add(w); }
// A recursive function to print DFS starting from v
void DFSUtil(int v,Boolean visited[])
{
// Mark the current node as visited and print it
visited[v] = true;
int n;
// Recur for all the vertices adjacent to this vertex
Iterator<Integer> i = adj[v].iterator();
while (i.hasNext())
{
n = i.next();
if (!visited[n])
DFSUtil(n,visited);
}
}
// Function that returns transpose of this graph
Graph getTranspose()
{
Graph g = new Graph(V);
for (int v = 0; v < V; v++)
{
// Recur for all the vertices adjacent to this vertex
Iterator<Integer> i = adj[v].listIterator();
while (i.hasNext())
g.adj[i.next()].add(v);
}

return g;
}
// The main function that returns true if graph is strongly
// connected
Boolean isSC()
{
// Step 1: Mark all the vertices as not visited
// (For first DFS)
Boolean visited[] = new Boolean[V];
for (int i = 0; i < V; i++)
visited[i] = false;
// Step 2: Do DFS traversal starting from first vertex.
DFSUtil(0, visited);
// If DFS traversal doesn't visit all vertices, then
// return false.
for (int i = 0; i < V; i++)
if (visited[i] == false)
return false;
// Step 3: Create a reversed graph
Graph gr = getTranspose();
// Step 4: Mark all the vertices as not visited (For
// second DFS)
for (int i = 0; i < V; i++)
visited[i] = false;
// Step 5: Do DFS for reversed graph starting from
// first vertex. Staring Vertex must be same starting
// point of first DFS
gr.DFSUtil(0, visited);
// If all vertices are not visited in second DFS, then
// return false
for (int i = 0; i < V; i++)
if (visited[i] == false)
return false;
return true;
}
public static void main(String args[])
{
// Create graphs given in the above diagrams
Graph g1 = new Graph(5);
g1.addEdge(0, 1);
g1.addEdge(1, 2);
g1.addEdge(2, 3);
g1.addEdge(3, 0);
g1.addEdge(2, 4);
g1.addEdge(4, 2);
if (g1.isSC())
System.out.println("Yes");
else
System.out.println("No");
Graph g2 = new Graph(4);
g2.addEdge(0, 1);
g2.addEdge(1, 2);
g2.addEdge(2, 3);
if (g2.isSC())
System.out.println("Yes");
else
System.out.println("No");
}
}
// This code is contributed by Aakash Hasija

Yes
No

Time Complexity: Time complexity of above implementation is sane as Depth First Search which is O(V+E) if the graph is represented using
adjacency matrix representation.
Exercise:

Can we use BFS instead of DFS in above algorithm?

References:
http://www.ieor.berkeley.edu/~hochbaum/files/ieor266-2012.pdf

Articulation Points (or Cut Vertices) in a Graph

A vertex in an undirected connected graph is an articulation point (or cut vertex) iff removing it (and edges through it) disconnects the graph.
Articulation points represent vulnerabilities in a connected network single points whose failure would split the network into 2 or more disconnected
components. They are useful for designing reliable networks.
For a disconnected undirected graph, an articulation point is a vertex removing which increases number of connected components.
Following are some example graphs with articulation points encircled with red color.

How to find all articulation points in a given graph?

A simple approach is to one by one remove all vertices and see if removal of a vertex causes disconnected graph. Following are steps of simple
approach for connected graph.
1) For every vertex v, do following
..a) Remove v from graph
..b) See if the graph remains connected (We can either use BFS or DFS)
..c) Add v back to the graph
Time complexity of above method is O(V*(V+E)) for a graph represented using adjacency list. Can we do better?
A O(V+E) algorithm to find all Articulation Points (APs)
The idea is to use DFS (Depth First Search). In DFS, we follow vertices in tree form called DFS tree. In DFS tree, a vertex u is parent of another
vertex v, if v is discovered by u (obviously v is an adjacent of u in graph). In DFS tree, a vertex u is articulation point if one of the following two
conditions is true.
1) u is root of DFS tree and it has at least two children.
2) u is not root of DFS tree and it has a child v such that no vertex in subtree rooted with v has a back edge to one of the ancestors (in DFS tree)
of u.
Following figure shows same points as above with one additional point that a leaf in DFS Tree can never be an articulation point. (Source Ref 2)

We do DFS traversal of given graph with additional code to find out Articulation Points (APs). In DFS traversal, we maintain a parent[] array
where parent[u] stores parent of vertex u. Among the above mentioned two cases, the first case is simple to detect. For every vertex, count

children. If currently visited vertex u is root (parent[u] is NIL) and has more than two children, print it.
How to handle second case? The second case is trickier. We maintain an array disc[] to store discovery time of vertices. For every node u, we
need to find out the earliest visited vertex (the vertex with minimum discovery time) that can be reached from subtree rooted with u. So we
maintain an additional array low[] which is defined as follows.
low[u] = min(disc[u], disc[w])
where w is an ancestor of u and there is a back edge from
some descendant of u to w.

Following are C++ and Java implementations of Tarjans algorithm for finding articulation points.

C++
// A C++ program to find articulation points in an undirected graph
#include<iostream>
#include <list>
#define NIL -1
using namespace std;
// A class that represents an undirected graph
class Graph
{
int V;
// No. of vertices
list<int> *adj;
// A dynamic array of adjacency lists
void APUtil(int v, bool visited[], int disc[], int low[],
int parent[], bool ap[]);
public:
Graph(int V); // Constructor
void addEdge(int v, int w); // function to add an edge to graph
void AP();
// prints articulation points
};
Graph::Graph(int V)
{
this->V = V;
adj = new list<int>[V];
}
void Graph::addEdge(int v, int w)
{
adj[v].push_back(w);
adj[w].push_back(v); // Note: the graph is undirected
}
// A recursive function that find articulation points using DFS traversal
// u --> The vertex to be visited next
// visited[] --> keeps tract of visited vertices
// disc[] --> Stores discovery times of visited vertices
// parent[] --> Stores parent vertices in DFS tree
// ap[] --> Store articulation points
void Graph::APUtil(int u, bool visited[], int disc[],
int low[], int parent[], bool ap[])
{
// A static variable is used for simplicity, we can avoid use of static
// variable by passing a pointer.
static int time = 0;
// Count of children in DFS Tree
int children = 0;
// Mark the current node as visited
visited[u] = true;
// Initialize discovery time and low value
disc[u] = low[u] = ++time;
// Go through all vertices aadjacent to this
list<int>::iterator i;
for (i = adj[u].begin(); i != adj[u].end(); ++i)
{
int v = *i; // v is current adjacent of u
// If v is not visited yet, then make it a child of u
// in DFS tree and recur for it
if (!visited[v])
{
children++;
parent[v] = u;
APUtil(v, visited, disc, low, parent, ap);

// Check if the subtree rooted with v has a connection to

// one of the ancestors of u
low[u] = min(low[u], low[v]);
// u is an articulation point in following cases
// (1) u is root of DFS tree and has two or more chilren.
if (parent[u] == NIL && children > 1)
ap[u] = true;
// (2) If u is not root and low value of one of its child is more
// than discovery value of u.
if (parent[u] != NIL && low[v] >= disc[u])
ap[u] = true;
}
// Update low value of u for parent function calls.
else if (v != parent[u])
low[u] = min(low[u], disc[v]);
}
}
// The function to do DFS traversal. It uses recursive function APUtil()
void Graph::AP()
{
// Mark all the vertices as not visited
bool *visited = new bool[V];
int *disc = new int[V];
int *low = new int[V];
int *parent = new int[V];
bool *ap = new bool[V]; // To store articulation points
// Initialize parent and visited, and ap(articulation point) arrays
for (int i = 0; i < V; i++)
{
parent[i] = NIL;
visited[i] = false;
ap[i] = false;
}
// Call the recursive helper function to find articulation points
// in DFS tree rooted with vertex 'i'
for (int i = 0; i < V; i++)
if (visited[i] == false)
APUtil(i, visited, disc, low, parent, ap);
// Now ap[] contains articulation points, print them
for (int i = 0; i < V; i++)
if (ap[i] == true)
cout << i << " ";
}
// Driver program to test above function
int main()
{
// Create graphs given in above diagrams
cout << "\nArticulation points in first graph \n";
Graph g1(5);
g1.addEdge(1, 0);
g1.addEdge(0, 2);
g1.addEdge(2, 1);
g1.addEdge(0, 3);
g1.addEdge(3, 4);
g1.AP();
cout << "\nArticulation points in second graph \n";
Graph g2(4);
g2.addEdge(0, 1);
g2.addEdge(1, 2);
g2.addEdge(2, 3);
g2.AP();
cout << "\nArticulation points in third graph \n";
Graph g3(7);
g3.addEdge(0, 1);
g3.addEdge(1, 2);
g3.addEdge(2, 0);
g3.addEdge(1, 3);
g3.addEdge(1, 4);
g3.addEdge(1, 6);

g3.addEdge(3, 5);
g3.addEdge(4, 5);
g3.AP();
return 0;
}

Java
// A Java program to find articulation points in an undirected graph
import java.io.*;
import java.util.*;
import java.util.LinkedList;
// This class represents an undirected graph using adjacency list
// representation
class Graph
{
private int V; // No. of vertices
// Array of lists for Adjacency List Representation
private LinkedList<Integer> adj[];
int time = 0;
static final int NIL = -1;
// Constructor
Graph(int v)
{
V = v;
adj = new LinkedList[v];
for (int i=0; i<v; ++i)
adj[i] = new LinkedList();
}
//Function to add an edge into the graph
void addEdge(int v, int w)
{
adj[v].add(w); // Add w to v's list.
adj[w].add(v); //Add v to w's list
}
// A recursive function that find articulation points using DFS
// u --> The vertex to be visited next
// visited[] --> keeps tract of visited vertices
// disc[] --> Stores discovery times of visited vertices
// parent[] --> Stores parent vertices in DFS tree
// ap[] --> Store articulation points
void APUtil(int u, boolean visited[], int disc[],
int low[], int parent[], boolean ap[])
{
// Count of children in DFS Tree
int children = 0;
// Mark the current node as visited
visited[u] = true;
// Initialize discovery time and low value
disc[u] = low[u] = ++time;
// Go through all vertices aadjacent to this
Iterator<Integer> i = adj[u].iterator();
while (i.hasNext())
{
int v = i.next(); // v is current adjacent of u
// If v is not visited yet, then make it a child of u
// in DFS tree and recur for it
if (!visited[v])
{
children++;
parent[v] = u;
APUtil(v, visited, disc, low, parent, ap);
// Check if the subtree rooted with v has a connection to
// one of the ancestors of u
low[u] = Math.min(low[u], low[v]);
// u is an articulation point in following cases

// (1) u is root of DFS tree and has two or more chilren.

if (parent[u] == NIL && children > 1)
ap[u] = true;
// (2) If u is not root and low value of one of its child
// is more than discovery value of u.
if (parent[u] != NIL && low[v] >= disc[u])
ap[u] = true;
}
// Update low value of u for parent function calls.
else if (v != parent[u])
low[u] = Math.min(low[u], disc[v]);
}
}
// The function to do DFS traversal. It uses recursive function APUtil()
void AP()
{
// Mark all the vertices as not visited
boolean visited[] = new boolean[V];
int disc[] = new int[V];
int low[] = new int[V];
int parent[] = new int[V];
boolean ap[] = new boolean[V]; // To store articulation points
// Initialize parent and visited, and ap(articulation point)
// arrays
for (int i = 0; i < V; i++)
{
parent[i] = NIL;
visited[i] = false;
ap[i] = false;
}
// Call the recursive helper function to find articulation
// points in DFS tree rooted with vertex 'i'
for (int i = 0; i < V; i++)
if (visited[i] == false)
APUtil(i, visited, disc, low, parent, ap);
// Now ap[] contains articulation points, print them
for (int i = 0; i < V; i++)
if (ap[i] == true)
System.out.print(i+" ");
}
// Driver method
public static void main(String args[])
{
// Create graphs given in above diagrams
System.out.println("Articulation points in first graph ");
Graph g1 = new Graph(5);
g1.addEdge(1, 0);
g1.addEdge(0, 2);
g1.addEdge(2, 1);
g1.addEdge(0, 3);
g1.addEdge(3, 4);
g1.AP();
System.out.println();
System.out.println("Articulation points in Second graph");
Graph g2 = new Graph(4);
g2.addEdge(0, 1);
g2.addEdge(1, 2);
g2.addEdge(2, 3);
g2.AP();
System.out.println();
System.out.println("Articulation points in Third graph ");
Graph g3 = new Graph(7);
g3.addEdge(0, 1);
g3.addEdge(1, 2);
g3.addEdge(2, 0);
g3.addEdge(1, 3);
g3.addEdge(1, 4);
g3.addEdge(1, 6);
g3.addEdge(3, 5);
g3.addEdge(4, 5);
g3.AP();
}

}
// This code is contributed by Aakash Hasija

Articulation points in first graph

0 3
Articulation points in second graph
1 2
Articulation points in third graph
1

Time Complexity: The above function is simple DFS with additional arrays. So time complexity is same as DFS which is O(V+E) for adjacency
list representation of graph.
References:
https://www.cs.washington.edu/education/courses/421/04su/slides/artic.pdf
http://www.slideshare.net/TraianRebedea/algorithm-design-and-complexity-course-8
http://faculty.simpson.edu/lydia.sinapova/www/cmsc250/LN250_Weiss/L25-Connectivity.htm

Biconnected graph
An undirected graph is called Biconnected if there are two vertex-disjoint paths between any two vertices. In a Biconnected Graph, there is a
simple cycle through any two vertices.
By convention, two nodes connected by an edge form a biconnected graph, but this does not verify the above properties. For a graph with more
than two vertices, the above properties must be there for it to be Biconnected.
Following are some examples.

See this for more examples.

How to find if a given graph is Biconnected or not?

A connected graph is Biconnected if it is connected and doesnt have any Articulation Point. We mainly need to check two things in a graph.
1) The graph is connected.
2) There is not articulation point in graph.
We start from any vertex and do DFS traversal. In DFS traversal, we check if there is any articulation point. If we dont find any articulation point,
then the graph is Biconnected. Finally, we need to check whether all vertices were reachable in DFS or not. If all vertices were not reachable, then
the graph is not even connected.
Following is C++ implementation of above approach.

C++
// A C++ program to find if a given undirected graph is
// biconnected
#include<iostream>
#include <list>
#define NIL -1
using namespace std;
// A class that represents an undirected graph
class Graph
{
int V;
// No. of vertices

list<int> *adj;
// A dynamic array of
bool isBCUtil(int v, bool visited[], int
int parent[]);
public:
Graph(int V); // Constructor
void addEdge(int v, int w); // to add an
bool isBC();
// returns true if graph
};

adjacency lists
disc[], int low[],

edge to graph
is Biconnected

Graph::Graph(int V)
{
this->V = V;
adj = new list<int>[V];
}
void Graph::addEdge(int v, int w)
{
adj[v].push_back(w);
adj[w].push_back(v); // Note: the graph is undirected
}
// A recursive function that returns true if there is an articulation
// point in given graph, otherwise returns false.
// This function is almost same as isAPUtil() here ( http://goo.gl/Me9Fw )
// u --> The vertex to be visited next
// visited[] --> keeps tract of visited vertices
// disc[] --> Stores discovery times of visited vertices
// parent[] --> Stores parent vertices in DFS tree
bool Graph::isBCUtil(int u, bool visited[], int disc[],int low[],int parent[])
{
// A static variable is used for simplicity, we can avoid use of static
// variable by passing a pointer.
static int time = 0;
// Count of children in DFS Tree
int children = 0;
// Mark the current node as visited
visited[u] = true;
// Initialize discovery time and low value
disc[u] = low[u] = ++time;
// Go through all vertices aadjacent to this
list<int>::iterator i;
for (i = adj[u].begin(); i != adj[u].end(); ++i)
{
int v = *i; // v is current adjacent of u
// If v is not visited yet, then make it a child of u
// in DFS tree and recur for it
if (!visited[v])
{
children++;
parent[v] = u;
// check if subgraph rooted with v has an articulation point
if (isBCUtil(v, visited, disc, low, parent))
return true;
// Check if the subtree rooted with v has a connection to
// one of the ancestors of u
low[u] = min(low[u], low[v]);
// u is an articulation point in following cases
// (1) u is root of DFS tree and has two or more chilren.
if (parent[u] == NIL && children > 1)
return true;
// (2) If u is not root and low value of one of its child is
// more than discovery value of u.
if (parent[u] != NIL && low[v] >= disc[u])
return true;
}
// Update low value of u for parent function calls.
else if (v != parent[u])
low[u] = min(low[u], disc[v]);
}
return false;

}
// The main function that returns true if graph is Biconnected,
// otherwise false. It uses recursive function isBCUtil()
bool Graph::isBC()
{
// Mark all the vertices as not visited
bool *visited = new bool[V];
int *disc = new int[V];
int *low = new int[V];
int *parent = new int[V];
// Initialize parent and visited, and ap(articulation point)
// arrays
for (int i = 0; i < V; i++)
{
parent[i] = NIL;
visited[i] = false;
}
// Call the recursive helper function to find if there is an articulation
// point in given graph. We do DFS traversal starring from vertex 0
if (isBCUtil(0, visited, disc, low, parent) == true)
return false;
// Now check whether the given graph is connected or not. An undirected
// graph is connected if all vertices are reachable from any starting
// point (we have taken 0 as starting point)
for (int i = 0; i < V; i++)
if (visited[i] == false)
return false;
return true;
}
// Driver program to test above function
int main()
{
// Create graphs given in above diagrams
Graph g1(2);
g1.addEdge(0, 1);
g1.isBC()? cout << "Yes\n" : cout << "No\n";
Graph g2(5);
g2.addEdge(1, 0);
g2.addEdge(0, 2);
g2.addEdge(2, 1);
g2.addEdge(0, 3);
g2.addEdge(3, 4);
g2.addEdge(2, 4);
g2.isBC()? cout << "Yes\n" : cout << "No\n";
Graph g3(3);
g3.addEdge(0, 1);
g3.addEdge(1, 2);
g3.isBC()? cout << "Yes\n" : cout << "No\n";
Graph g4(5);
g4.addEdge(1, 0);
g4.addEdge(0, 2);
g4.addEdge(2, 1);
g4.addEdge(0, 3);
g4.addEdge(3, 4);
g4.isBC()? cout << "Yes\n" : cout << "No\n";
Graph g5(3);
g5.addEdge(0, 1);
g5.addEdge(1, 2);
g5.addEdge(2, 0);
g5.isBC()? cout << "Yes\n" : cout << "No\n";
return 0;
}

Java
// A Java program to find if a given undirected graph is
// biconnected
import java.io.*;
import java.util.*;

import java.util.LinkedList;
// This class represents a directed graph using adjacency
// list representation
class Graph
{
private int V; // No. of vertices
// Array of lists for Adjacency List Representation
private LinkedList<Integer> adj[];
int time = 0;
static final int NIL = -1;
// Constructor
Graph(int v)
{
V = v;
adj = new LinkedList[v];
for (int i=0; i<v; ++i)
adj[i] = new LinkedList();
}
//Function to add an edge into the graph
void addEdge(int v, int w)
{
adj[v].add(w); //Note that the graph is undirected.
adj[w].add(v);
}
// A recursive function that returns true if there is an articulation
// point in given graph, otherwise returns false.
// This function is almost same as isAPUtil() @ http://goo.gl/Me9Fw
// u --> The vertex to be visited next
// visited[] --> keeps tract of visited vertices
// disc[] --> Stores discovery times of visited vertices
// parent[] --> Stores parent vertices in DFS tree
boolean isBCUtil(int u, boolean visited[], int disc[],int low[],
int parent[])
{
// Count of children in DFS Tree
int children = 0;
// Mark the current node as visited
visited[u] = true;
// Initialize discovery time and low value
disc[u] = low[u] = ++time;
// Go through all vertices aadjacent to this
Iterator<Integer> i = adj[u].iterator();
while (i.hasNext())
{
int v = i.next(); // v is current adjacent of u
// If v is not visited yet, then make it a child of u
// in DFS tree and recur for it
if (!visited[v])
{
children++;
parent[v] = u;
// check if subgraph rooted with v has an articulation point
if (isBCUtil(v, visited, disc, low, parent))
return true;
// Check if the subtree rooted with v has a connection to
// one of the ancestors of u
low[u] = Math.min(low[u], low[v]);
// u is an articulation point in following cases
// (1) u is root of DFS tree and has two or more chilren.
if (parent[u] == NIL && children > 1)
return true;
// (2) If u is not root and low value of one of its
// child is more than discovery value of u.
if (parent[u] != NIL && low[v] >= disc[u])
return true;

}
// Update low value of u for parent function calls.
else if (v != parent[u])
low[u] = Math.min(low[u], disc[v]);
}
return false;
}
// The main function that returns true if graph is Biconnected,
// otherwise false. It uses recursive function isBCUtil()
boolean isBC()
{
// Mark all the vertices as not visited
boolean visited[] = new boolean[V];
int disc[] = new int[V];
int low[] = new int[V];
int parent[] = new int[V];
// Initialize parent and visited, and ap(articulation point)
// arrays
for (int i = 0; i < V; i++)
{
parent[i] = NIL;
visited[i] = false;
}
//
//
//
if

Call the recursive helper function to find if there is an

articulation/ point in given graph. We do DFS traversal
starring from vertex 0
(isBCUtil(0, visited, disc, low, parent) == true)
return false;

// Now check whether the given graph is connected or not.

// An undirected graph is connected if all vertices are
// reachable from any starting point (we have taken 0 as
// starting point)
for (int i = 0; i < V; i++)
if (visited[i] == false)
return false;
return true;
}
// Driver method
public static void main(String args[])
{
// Create graphs given in above diagrams
Graph g1 =new Graph(2);
g1.addEdge(0, 1);
if (g1.isBC())
System.out.println("Yes");
else
System.out.println("No");
Graph g2 =new Graph(5);
g2.addEdge(1, 0);
g2.addEdge(0, 2);
g2.addEdge(2, 1);
g2.addEdge(0, 3);
g2.addEdge(3, 4);
g2.addEdge(2, 4);
if (g2.isBC())
System.out.println("Yes");
else
System.out.println("No");
Graph g3 = new Graph(3);
g3.addEdge(0, 1);
g3.addEdge(1, 2);
if (g3.isBC())
System.out.println("Yes");
else
System.out.println("No");
Graph g4 = new Graph(5);
g4.addEdge(1, 0);
g4.addEdge(0, 2);
g4.addEdge(2, 1);
g4.addEdge(0, 3);
g4.addEdge(3, 4);

if (g4.isBC())
System.out.println("Yes");
else
System.out.println("No");
Graph g5= new Graph(3);
g5.addEdge(0, 1);
g5.addEdge(1, 2);
g5.addEdge(2, 0);
if (g5.isBC())
System.out.println("Yes");
else
System.out.println("No");
}
}
// This code is contributed by Aakash Hasija

Yes
Yes
No
No
Yes

Time Complexity: The above function is a simple DFS with additional arrays. So time complexity is same as DFS which is O(V+E) for
adjacency list representation of graph.
References:
http://www.cs.purdue.edu/homes/ayg/CS251/slides/chap9d.pdf

Bridges in a graph
An edge in an undirected connected graph is a bridge iff removing it disconnects the graph. For a disconnected undirected graph, definition is
similar, a bridge is an edge removing which increases number of connected components.
Like Articulation Points,bridges represent vulnerabilities in a connected network and are useful for designing reliable networks. For example, in a
wired computer network, an articulation point indicates the critical computers and a bridge indicates the critical wires or connections.
Following are some example graphs with bridges highlighted with red color.

How to find all bridges in a given graph?

A simple approach is to one by one remove all edges and see if removal of a edge causes disconnected graph. Following are steps of simple
approach for connected graph.
1) For every edge (u, v), do following
..a) Remove (u, v) from graph
..b) See if the graph remains connected (We can either use BFS or DFS)
..c) Add (u, v) back to the graph.
Time complexity of above method is O(E*(V+E)) for a graph represented using adjacency list. Can we do better?
A O(V+E) algorithm to find all Bridges
The idea is similar to O(V+E) algorithm for Articulation Points. We do DFS traversal of the given graph. In DFS tree an edge (u, v) (u is parent of
v in DFS tree) is bridge if there does not exit any other alternative to reach u or an ancestor of u from subtree rooted with v. As discussed in the
previous post, the value low[v] indicates earliest visited vertex reachable from subtree rooted with v. The condition for an edge (u, v) to be a
bridge is, low[v] > disc[u].
Following are C++ and Java implementations of above approach.

C++
// A C++ program to find bridges in a given undirected graph
#include<iostream>
#include <list>
#define NIL -1
using namespace std;
// A class that represents an undirected graph
class Graph
{
int V;
// No. of vertices
list<int> *adj;
// A dynamic array of adjacency lists
void bridgeUtil(int v, bool visited[], int disc[], int low[],
int parent[]);
public:
Graph(int V); // Constructor
void addEdge(int v, int w); // to add an edge to graph
void bridge();
// prints all bridges
};
Graph::Graph(int V)
{
this->V = V;
adj = new list<int>[V];
}
void Graph::addEdge(int v, int w)
{
adj[v].push_back(w);
adj[w].push_back(v); // Note: the graph is undirected
}
// A recursive function that finds and prints bridges using
// DFS traversal
// u --> The vertex to be visited next

// visited[] --> keeps tract of visited vertices

// disc[] --> Stores discovery times of visited vertices
// parent[] --> Stores parent vertices in DFS tree
void Graph::bridgeUtil(int u, bool visited[], int disc[],
int low[], int parent[])
{
// A static variable is used for simplicity, we can
// avoid use of static variable by passing a pointer.
static int time = 0;
// Mark the current node as visited
visited[u] = true;
// Initialize discovery time and low value
disc[u] = low[u] = ++time;
// Go through all vertices aadjacent to this
list<int>::iterator i;
for (i = adj[u].begin(); i != adj[u].end(); ++i)
{
int v = *i; // v is current adjacent of u
// If v is not visited yet, then recur for it
if (!visited[v])
{
parent[v] = u;
bridgeUtil(v, visited, disc, low, parent);
// Check if the subtree rooted with v has a
// connection to one of the ancestors of u
low[u] = min(low[u], low[v]);
// If the lowest vertex reachable from subtree
// under v is below u in DFS tree, then u-v
// is a bridge
if (low[v] > disc[u])
cout << u <<" " << v << endl;
}
// Update low value of u for parent function calls.
else if (v != parent[u])
low[u] = min(low[u], disc[v]);
}
}
// DFS based function to find all bridges. It uses recursive
// function bridgeUtil()
void Graph::bridge()
{
// Mark all the vertices as not visited
bool *visited = new bool[V];
int *disc = new int[V];
int *low = new int[V];
int *parent = new int[V];
// Initialize parent and visited arrays
for (int i = 0; i < V; i++)
{
parent[i] = NIL;
visited[i] = false;
}
// Call the recursive helper function to find Bridges
// in DFS tree rooted with vertex 'i'
for (int i = 0; i < V; i++)
if (visited[i] == false)
bridgeUtil(i, visited, disc, low, parent);
}
// Driver program to test above function
int main()
{
// Create graphs given in above diagrams
cout << "\nBridges in first graph \n";
Graph g1(5);
g1.addEdge(1, 0);
g1.addEdge(0, 2);
g1.addEdge(2, 1);
g1.addEdge(0, 3);
g1.addEdge(3, 4);
g1.bridge();

cout << "\nBridges in second graph \n";

Graph g2(4);
g2.addEdge(0, 1);
g2.addEdge(1, 2);
g2.addEdge(2, 3);
g2.bridge();
cout << "\nBridges in third graph \n";
Graph g3(7);
g3.addEdge(0, 1);
g3.addEdge(1, 2);
g3.addEdge(2, 0);
g3.addEdge(1, 3);
g3.addEdge(1, 4);
g3.addEdge(1, 6);
g3.addEdge(3, 5);
g3.addEdge(4, 5);
g3.bridge();
return 0;
}

Java
// A Java program to find bridges in a given undirected graph
import java.io.*;
import java.util.*;
import java.util.LinkedList;
// This class represents a undirected graph using adjacency list
// representation
class Graph
{
private int V; // No. of vertices
// Array of lists for Adjacency List Representation
private LinkedList<Integer> adj[];
int time = 0;
static final int NIL = -1;
// Constructor
Graph(int v)
{
V = v;
adj = new LinkedList[v];
for (int i=0; i<v; ++i)
adj[i] = new LinkedList();
}
// Function to add an edge into the graph
void addEdge(int v, int w)
{
adj[v].add(w); // Add w to v's list.
adj[w].add(v); //Add v to w's list
}
// A recursive function that finds and prints bridges
// using DFS traversal
// u --> The vertex to be visited next
// visited[] --> keeps tract of visited vertices
// disc[] --> Stores discovery times of visited vertices
// parent[] --> Stores parent vertices in DFS tree
void bridgeUtil(int u, boolean visited[], int disc[],
int low[], int parent[])
{
// Count of children in DFS Tree
int children = 0;
// Mark the current node as visited
visited[u] = true;
// Initialize discovery time and low value
disc[u] = low[u] = ++time;
// Go through all vertices aadjacent to this
Iterator<Integer> i = adj[u].iterator();
while (i.hasNext())
{

int v = i.next(); // v is current adjacent of u

//
//
//
if
{

If v is not visited yet, then make it a child

of u in DFS tree and recur for it.
If v is not visited yet, then recur for it
(!visited[v])
parent[v] = u;
bridgeUtil(v, visited, disc, low, parent);
// Check if the subtree rooted with v has a
// connection to one of the ancestors of u
low[u] = Math.min(low[u], low[v]);
//
//
//
if

If the lowest vertex reachable from subtree

under v is below u in DFS tree, then u-v is
a bridge
(low[v] > disc[u])
System.out.println(u+" "+v);

}
// Update low value of u for parent function calls.
else if (v != parent[u])
low[u] = Math.min(low[u], disc[v]);
}
}
// DFS based function to find all bridges. It uses recursive
// function bridgeUtil()
void bridge()
{
// Mark all the vertices as not visited
boolean visited[] = new boolean[V];
int disc[] = new int[V];
int low[] = new int[V];
int parent[] = new int[V];
// Initialize parent and visited, and ap(articulation point)
// arrays
for (int i = 0; i < V; i++)
{
parent[i] = NIL;
visited[i] = false;
}
// Call the recursive helper function to find Bridges
// in DFS tree rooted with vertex 'i'
for (int i = 0; i < V; i++)
if (visited[i] == false)
bridgeUtil(i, visited, disc, low, parent);
}
public static void main(String args[])
{
// Create graphs given in above diagrams
System.out.println("Bridges in first graph ");
Graph g1 = new Graph(5);
g1.addEdge(1, 0);
g1.addEdge(0, 2);
g1.addEdge(2, 1);
g1.addEdge(0, 3);
g1.addEdge(3, 4);
g1.bridge();
System.out.println();
System.out.println("Bridges in Second graph");
Graph g2 = new Graph(4);
g2.addEdge(0, 1);
g2.addEdge(1, 2);
g2.addEdge(2, 3);
g2.bridge();
System.out.println();
System.out.println("Bridges in Third graph ");
Graph g3 = new Graph(7);
g3.addEdge(0, 1);
g3.addEdge(1, 2);
g3.addEdge(2, 0);
g3.addEdge(1, 3);

g3.addEdge(1,
g3.addEdge(1,
g3.addEdge(3,
g3.addEdge(4,
g3.bridge();

4);
6);
5);
5);

}
}
// This code is contributed by Aakash Hasija

Bridges in first graph

3 4
0 3
Bridges in second graph
2 3
1 2
0 1
Bridges in third graph
1 6

Time Complexity: The above function is simple DFS with additional arrays. So time complexity is same as DFS which is O(V+E) for adjacency
list representation of graph.
References:
https://www.cs.washington.edu/education/courses/421/04su/slides/artic.pdf
http://www.slideshare.net/TraianRebedea/algorithm-design-and-complexity-course-8
http://faculty.simpson.edu/lydia.sinapova/www/cmsc250/LN250_Weiss/L25-Connectivity.htm
http://www.youtube.com/watch?v=bmyyxNyZKzI

Eulerian path and circuit for undirected graph

Eulerian Path is a path in graph that visits every edge exactly once. Eulerian Circuit is an Eulerian Path which starts and ends on the same vertex.

How to find whether a given graph is Eulerian or not?

The problem is same as following question. Is it possible to draw a given graph without lifting pencil from the paper and without tracing any of the
edges more than once.
A graph is called Eulerian if it has an Eulerian Cycle and called Semi-Eulerian if it has an Eulerian Path. The problem seems similar to Hamiltonian
Path which is NP complete problem for a general graph. Fortunately, we can find whether a given graph has a Eulerian Path or not in polynomial
time. In fact, we can find it in O(V+E) time.
Following are some interesting properties of undirected graphs with an Eulerian path and cycle. We can use these properties to find whether a
graph is Eulerian or not.
Eulerian Cycle
An undirected graph has Eulerian cycle if following two conditions are true.
.a) All vertices with non-zero degree are connected. We dont care about vertices with zero degree because they dont belong to Eulerian Cycle or
Path (we only consider all edges).
.b) All vertices have even degree.
Eulerian Path
An undirected graph has Eulerian Path if following two conditions are true.
.a) Same as condition (a) for Eulerian Cycle
.b) If zero or two vertices have odd degree and all other vertices have even degree. Note that only one vertex with odd degree is not possible in an
undirected graph (sum of all degrees is always even in an undirected graph)
Note that a graph with no edges is considered Eulerian because there are no edges to traverse.
How does this work?
In Eulerian path, each time we visit a vertex v, we walk through two unvisited edges with one end point as v. Therefore, all middle vertices in
Eulerian Path must have even degree. For Eulerian Cycle, any vertex can be middle vertex, therefore all vertices must have even degree.

C++
// A C++ program to check if a given graph is Eulerian or not
#include<iostream>
#include <list>
using namespace std;
// A class that represents an undirected graph

class Graph
{
int V;
// No. of vertices
list<int> *adj;
// A dynamic array of adjacency lists
public:
// Constructor and destructor
Graph(int V) {this->V = V; adj = new list<int>[V]; }
~Graph() { delete [] adj; } // To avoid memory leak
// function to add an edge to graph
void addEdge(int v, int w);
// Method to check if this graph is Eulerian or not
int isEulerian();
// Method to check if all non-zero degree vertices are connected
bool isConnected();
// Function to do DFS starting from v. Used in isConnected();
void DFSUtil(int v, bool visited[]);
};
void Graph::addEdge(int v, int w)
{
adj[v].push_back(w);
adj[w].push_back(v); // Note: the graph is undirected
}
void Graph::DFSUtil(int v, bool visited[])
{
// Mark the current node as visited and print it
visited[v] = true;
// Recur for all the vertices adjacent to this vertex
list<int>::iterator i;
for (i = adj[v].begin(); i != adj[v].end(); ++i)
if (!visited[*i])
DFSUtil(*i, visited);
}
// Method to check if all non-zero degree vertices are connected.
// It mainly does DFS traversal starting from
bool Graph::isConnected()
{
// Mark all the vertices as not visited
bool visited[V];
int i;
for (i = 0; i < V; i++)
visited[i] = false;
// Find a vertex with non-zero degree
for (i = 0; i < V; i++)
if (adj[i].size() != 0)
break;
// If there are no edges in the graph, return true
if (i == V)
return true;
// Start DFS traversal from a vertex with non-zero degree
DFSUtil(i, visited);
// Check if all non-zero degree vertices are visited
for (i = 0; i < V; i++)
if (visited[i] == false && adj[i].size() > 0)
return false;
return true;
}
/* The function returns one of the following values
0 --> If grpah is not Eulerian
1 --> If graph has an Euler path (Semi-Eulerian)
2 --> If graph has an Euler Circuit (Eulerian) */
int Graph::isEulerian()
{
// Check if all non-zero degree vertices are connected
if (isConnected() == false)
return 0;
// Count vertices with odd degree

int odd = 0;
for (int i = 0; i < V; i++)
if (adj[i].size() & 1)
odd++;
// If count is more than 2, then graph is not Eulerian
if (odd > 2)
return 0;
// If odd count is 2, then semi-eulerian.
// If odd count is 0, then eulerian
// Note that odd count can never be 1 for undirected graph
return (odd)? 1 : 2;
}
// Function to run test cases
void test(Graph &g)
{
int res = g.isEulerian();
if (res == 0)
cout << "graph is not Eulerian\n";
else if (res == 1)
cout << "graph has a Euler path\n";
else
cout << "graph has a Euler cycle\n";
}
// Driver program to test above function
int main()
{
// Let us create and test graphs shown in above figures
Graph g1(5);
g1.addEdge(1, 0);
g1.addEdge(0, 2);
g1.addEdge(2, 1);
g1.addEdge(0, 3);
g1.addEdge(3, 4);
test(g1);
Graph g2(5);
g2.addEdge(1,
g2.addEdge(0,
g2.addEdge(2,
g2.addEdge(0,
g2.addEdge(3,
g2.addEdge(4,
test(g2);

0);
2);
1);
3);
4);
0);

Graph g3(5);
g3.addEdge(1,
g3.addEdge(0,
g3.addEdge(2,
g3.addEdge(0,
g3.addEdge(3,
g3.addEdge(1,
test(g3);

0);
2);
1);
3);
4);
3);

// Let us create a graph with 3 vertices

// connected in the form of cycle
Graph g4(3);
g4.addEdge(0, 1);
g4.addEdge(1, 2);
g4.addEdge(2, 0);
test(g4);
// Let us create a graph with all veritces
// with zero degree
Graph g5(3);
test(g5);
return 0;
}

Java
// A Java program to check if a given graph is Eulerian or not
import java.io.*;
import java.util.*;
import java.util.LinkedList;

// This class represents an undirected graph using adjacency list

// representation
class Graph
{
private int V; // No. of vertices
// Array of lists for Adjacency List Representation
private LinkedList<Integer> adj[];
// Constructor
Graph(int v)
{
V = v;
adj = new LinkedList[v];
for (int i=0; i<v; ++i)
adj[i] = new LinkedList();
}
//Function to add an edge into the graph
void addEdge(int v, int w)
{
adj[v].add(w);// Add w to v's list.
adj[w].add(v); //The graph is undirected
}
// A function used by DFS
void DFSUtil(int v,boolean visited[])
{
// Mark the current node as visited
visited[v] = true;
// Recur for all the vertices adjacent to this vertex
Iterator<Integer> i = adj[v].listIterator();
while (i.hasNext())
{
int n = i.next();
if (!visited[n])
DFSUtil(n, visited);
}
}
// Method to check if all non-zero degree vertices are
// connected. It mainly does DFS traversal starting from
boolean isConnected()
{
// Mark all the vertices as not visited
boolean visited[] = new boolean[V];
int i;
for (i = 0; i < V; i++)
visited[i] = false;
// Find a vertex with non-zero degree
for (i = 0; i < V; i++)
if (adj[i].size() != 0)
break;
// If there are no edges in the graph, return true
if (i == V)
return true;
// Start DFS traversal from a vertex with non-zero degree
DFSUtil(i, visited);
// Check if all non-zero degree vertices are visited
for (i = 0; i < V; i++)
if (visited[i] == false && adj[i].size() > 0)
return false;
return true;
}
/* The function returns one of the following values
0 --> If grpah is not Eulerian
1 --> If graph has an Euler path (Semi-Eulerian)
2 --> If graph has an Euler Circuit (Eulerian) */
int isEulerian()
{
// Check if all non-zero degree vertices are connected
if (isConnected() == false)
return 0;

// Count vertices with odd degree

int odd = 0;
for (int i = 0; i < V; i++)
if (adj[i].size()%2!=0)
odd++;
// If count is more than 2, then graph is not Eulerian
if (odd > 2)
return 0;
// If odd count is 2, then semi-eulerian.
// If odd count is 0, then eulerian
// Note that odd count can never be 1 for undirected graph
return (odd==2)? 1 : 2;
}
// Function to run test cases
void test()
{
int res = isEulerian();
if (res == 0)
System.out.println("graph is not Eulerian");
else if (res == 1)
System.out.println("graph has a Euler path");
else
System.out.println("graph has a Euler cycle");
}
// Driver method
public static void main(String args[])
{
// Let us create and test graphs shown in above figures
Graph g1 = new Graph(5);
g1.addEdge(1, 0);
g1.addEdge(0, 2);
g1.addEdge(2, 1);
g1.addEdge(0, 3);
g1.addEdge(3, 4);
g1.test();
Graph g2 = new Graph(5);
g2.addEdge(1, 0);
g2.addEdge(0, 2);
g2.addEdge(2, 1);
g2.addEdge(0, 3);
g2.addEdge(3, 4);
g2.addEdge(4, 0);
g2.test();
Graph g3 = new Graph(5);
g3.addEdge(1, 0);
g3.addEdge(0, 2);
g3.addEdge(2, 1);
g3.addEdge(0, 3);
g3.addEdge(3, 4);
g3.addEdge(1, 3);
g3.test();
// Let us create a graph with 3 vertices
// connected in the form of cycle
Graph g4 = new Graph(3);
g4.addEdge(0, 1);
g4.addEdge(1, 2);
g4.addEdge(2, 0);
g4.test();
// Let us create a graph with all veritces
// with zero degree
Graph g5 = new Graph(3);
g5.test();
}
}
// This code is contributed by Aakash Hasija

graph
graph
graph
graph
graph

has a Euler path

has a Euler cycle
is not Eulerian
has a Euler cycle
has a Euler cycle

Time Complexity: O(V+E)

We will soon be covering following topics on Eulerian Path and Circuit
1) Eulerian Path and Circuit for a Directed Graphs.
2) How to print a Eulerian Path or Circuit?
References:
http://en.wikipedia.org/wiki/Eulerian_path

Fleurys Algorithm for printing Eulerian Path or Circuit

Eulerian Pathis a path in graph that visits every edge exactly once. Eulerian Circuit is an Eulerian Path which starts and ends on the same vertex.
We strongly recommend to first read the following post on Euler Path and Circuit.
http://www.geeksforgeeks.org/eulerian-path-and-circuit/
In the above mentioned post, we discussed the problem of finding out whether a given graph is Eulerian or not. In this post, an algorithm to print
Eulerian trail or circuit is discussed.
Following is Fleurys Algorithm for printing Eulerian trail or cycle (Source Ref1).
1. Make sure the graph has either 0 or 2 odd vertices.
2. If there are 0 odd vertices, start anywhere. If there are 2 odd vertices, start at one of them.
3. Follow edges one at a time. If you have a choice between a bridge and a non-bridge, always choose the non-bridge.
4. Stop when you run out of edges.
The idea is, dont burn bridges so that we can come back to a vertex and traverse remaining edges. For example let us consider the following
graph.

There are two vertices with odd degree, 2 and 3, we can start path from any of them. Let us start tour from vertex 2.

There are three edges going out from vertex 2, which one to pick? We dont pick the edge 2-3? because that is a bridge (we wont be able to come
back to 3). We can pick any of the remaining two edge. Let us say we pick 2-0?. We remove this edge and move to vertex 0.

There is only one edge from vertex 0, so we pick it, remove it and move to vertex 1. Euler tour becomes 2-0 0-1?.

There is only one edge from vertex 1, so we pick it, remove it and move to vertex 2. Euler tour becomes 2-0 0-1 1-2?

Again there is only one edge from vertex 2, so we pick it, remove it and move to vertex 3. Euler tour becomes 2-0 0-1 1-2 2-3?

There are no more edges left, so we stop here. Final tour is 2-0 0-1 1-2 2-3?.
See this for and this fore more examples.
Following is C++ implementation of above algorithm. In the following code, it is assumed that the given graph has an Eulerian trail or Circuit. The
main focus is to print an Eulerian trail or circuit. We can use isEulerian() to first check whether there is an Eulerian Trail or Circuit in the given
graph.
We first find the starting point which must be an odd vertex (if there are odd vertices) and store it in variable u. If there are zero odd vertices, we
start from vertex 0. We call printEulerUtil() to print Euler tour starting with u. We traverse all adjacent vertices of u, if there is only one adjacent
vertex, we immediately consider it. If there are more than one adjacent vertices, we consider an adjacent v only if edge u-v is not a bridge. How to
find if a given is edge is bridge? We count number of vertices reachable from u. We remove edge u-v and again count number of reachable
vertices from u. If number of reachable vertices are reduced, then edge u-v is a bridge. To count reachable vertices, we can either use BFS or
DFS, we have used DFS in the above code. The function DFSCount(u) returns number of vertices reachable from u.
Once an edge is processed (included in Euler tour), we remove it from the graph. To remove the edge, we replace the vertex entry with -1 in
adjacency list. Note that simply deleting the node may not work as the code is recursive and a parent call may be in middle of adjacency list.
// A C++ program print Eulerian Trail in a given Eulerian or Semi-Eulerian Graph
#include <iostream>
#include <string.h>
#include <algorithm>
#include <list>
using namespace std;
// A class that represents an undirected graph
class Graph
{
int V;
// No. of vertices
list<int> *adj;
// A dynamic array of adjacency lists
public:
// Constructor and destructor
Graph(int V) { this->V = V; adj = new list<int>[V]; }
~Graph()
{ delete [] adj; }
// functions to add and remove edge
void addEdge(int u, int v) { adj[u].push_back(v); adj[v].push_back(u); }
void rmvEdge(int u, int v);
// Methods to print Eulerian tour
void printEulerTour();
void printEulerUtil(int s);
// This function returns count of vertices reachable from v. It does DFS
int DFSCount(int v, bool visited[]);
// Utility function to check if edge u-v is a valid next edge in
// Eulerian trail or circuit
bool isValidNextEdge(int u, int v);
};
/* The main function that print Eulerian Trail. It first finds an odd
degree vertex (if there is any) and then calls printEulerUtil()
to print the path */
void Graph::printEulerTour()
{
// Find a vertex with odd degree
int u = 0;
for (int i = 0; i < V; i++)
if (adj[i].size() & 1)
{ u = i; break; }
// Print tour starting from oddv
printEulerUtil(u);
cout << endl;
}
// Print Euler tour starting from vertex u
void Graph::printEulerUtil(int u)
{

// Recur for all the vertices adjacent to this vertex

list<int>::iterator i;
for (i = adj[u].begin(); i != adj[u].end(); ++i)
{
int v = *i;
// If edge u-v is not removed and it's a a valid next edge
if (v != -1 && isValidNextEdge(u, v))
{
cout << u << "-" << v << " ";
rmvEdge(u, v);
printEulerUtil(v);
}
}
}
// The function to check if edge u-v can be considered as next edge in
// Euler Tout
bool Graph::isValidNextEdge(int u, int v)
{
// The edge u-v is valid in one of the following two cases:
// 1) If v is the only adjacent vertex of u
int count = 0; // To store count of adjacent vertices
list<int>::iterator i;
for (i = adj[u].begin(); i != adj[u].end(); ++i)
if (*i != -1)
count++;
if (count == 1)
return true;
// 2) If there are multiple adjacents, then u-v is not a bridge
// Do following steps to check if u-v is a bridge
// 2.a) count of vertices reachable from u
bool visited[V];
memset(visited, false, V);
int count1 = DFSCount(u, visited);
// 2.b) Remove edge (u, v) and after removing the edge, count
// vertices reachable from u
rmvEdge(u, v);
memset(visited, false, V);
int count2 = DFSCount(u, visited);
// 2.c) Add the edge back to the graph
addEdge(u, v);
// 2.d) If count1 is greater, then edge (u, v) is a bridge
return (count1 > count2)? false: true;
}
// This function removes edge u-v from graph. It removes the edge by
// replacing adjcent vertex value with -1.
void Graph::rmvEdge(int u, int v)
{
// Find v in adjacency list of u and replace it with -1
list<int>::iterator iv = find(adj[u].begin(), adj[u].end(), v);
*iv = -1;
// Find u in adjacency list of v and replace it with -1
list<int>::iterator iu = find(adj[v].begin(), adj[v].end(), u);
*iu = -1;
}
// A DFS based function to count reachable vertices from v
int Graph::DFSCount(int v, bool visited[])
{
// Mark the current node as visited
visited[v] = true;
int count = 1;
// Recur for all vertices adjacent to this vertex
list<int>::iterator i;
for (i = adj[v].begin(); i != adj[v].end(); ++i)
if (*i != -1 && !visited[*i])
count += DFSCount(*i, visited);
return count;
}

// Driver program to test above function

int main()
{
// Let us first create and test graphs shown in above figure
Graph g1(4);
g1.addEdge(0, 1);
g1.addEdge(0, 2);
g1.addEdge(1, 2);
g1.addEdge(2, 3);
g1.printEulerTour();
Graph g2(3);
g2.addEdge(0, 1);
g2.addEdge(1, 2);
g2.addEdge(2, 0);
g2.printEulerTour();
Graph g3(5);
g3.addEdge(1, 0);
g3.addEdge(0, 2);
g3.addEdge(2, 1);
g3.addEdge(0, 3);
g3.addEdge(3, 4);
g3.addEdge(3, 2);
g3.addEdge(3, 1);
g3.addEdge(2, 4);
g3.printEulerTour();
return 0;
}

Output:
2-0 0-1 1-2 2-3
0-1 1-2 2-0
0-1 1-2 2-0 0-3 3-4 4-2 2-3 3-1

Note that the above code modifies given graph, we can create a copy of graph if we dont want the given graph to be modified.
Time Complexity: Time complexity of the above implementation is O ((V+E)2). The function printEulerUtil() is like DFS and it calls
isValidNextEdge() which also does DFS two times. Time complexity of DFS for adjacency list representation is O(V+E). Therefore overall time
complexity is O((V+E)*(V+E)) which can be written as O(E2) for a connected graph.
There are better algorithms to print Euler tour, we will soon be covering them as separate posts.
References:
http://www.math.ku.edu/~jmartin/courses/math105-F11/Lectures/chapter5-part2.pdf
http://en.wikipedia.org/wiki/Eulerian_path#Fleury.27s_algorithm

Strongly Connected Components

A directed graph is strongly connected if there is a path between all pairs of vertices. A strongly connected component (SCC) of a directed graph
is a maximal strongly connected subgraph. For example, there are 3 SCCs in the following graph.

We can find all strongly connected components in O(V+E) time using Kosarajus algorithm. Following is detailed Kosarajus algorithm.
1) Create an empty stack S and do DFS traversal of a graph. In DFS traversal, after calling recursive DFS for adjacent vertices of a vertex, push
the vertex to stack.
2) Reverse directions of all arcs to obtain the transpose graph.
3) One by one pop a vertex from S while S is not empty. Let the popped vertex be v. Take v as source and do DFS (call DFSUtil(v)). The DFS
starting from v prints strongly connected component of v.
How does this work?
The above algorithm is DFS based. It does DFS two times. DFS of a graph produces a single tree if all vertices are reachable from the DFS
starting point. Otherwise DFS produces a forest. So DFS of a graph with only one SCC always produces a tree. The important point to note is
DFS may produce a tree or a forest when there are more than one SCCs depending upon the chosen starting point. For example, in the above
diagram, if we start DFS from vertices 0 or 1 or 2, we get a tree as output. And if we start from 3 or 4, we get a forest. To find and print all
SCCs, we would want to start DFS from vertex 4 (which is a sink vertex), then move to 3 which is sink in the remaining set (set excluding 4) and
finally any of the remaining vertices (0, 1, 2). So how do we find this sequence of picking vertices as starting points of DFS? Unfortunately, there is
no direct way for getting this sequence. However, if we do a DFS of graph and store vertices according to their finish times, we make sure that the
finish time of a vertex that connects to other SCCs (other that its own SCC), will always be greater than finish time of vertices in the other SCC
(See this for proof). For example, in DFS of above example graph, finish time of 0 is always greater than 3 and 4 (irrespective of the sequence of
vertices considered for DFS). And finish time of 3 is always greater than 4. DFS doesnt guarantee about other vertices, for example finish times of
1 and 2 may be smaller or greater than 3 and 4 depending upon the sequence of vertices considered for DFS. So to use this property, we do DFS
traversal of complete graph and push every finished vertex to a stack. In stack, 3 always appears after 4, and 0 appear after both 3 and 4.
In the next step, we reverse the graph. Consider the graph of SCCs. In the reversed graph, the edges that connect two components are reversed.
So the SCC {0, 1, 2} becomes sink and the SCC {4} becomes source. As discussed above, in stack, we always have 0 before 3 and 4. So if we
do a DFS of the reversed graph using sequence of vertices in stack, we process vertices from sink to source. That is what we wanted to achieve
and that is all needed to print SCCs one by one.

Following is C++ implementation of Kosarajus algorithm.

C++
// C++ Implementation of Kosaraju's algorithm to print all SCCs
#include <iostream>
#include <list>
#include <stack>
using namespace std;
class Graph
{

int V;
// No. of vertices
list<int> *adj;
// An array of adjacency lists
// Fills Stack with vertices (in increasing order of finishing
// times). The top element of stack has the maximum finishing
// time
void fillOrder(int v, bool visited[], stack<int> &Stack);
// A recursive function to print DFS starting from v
void DFSUtil(int v, bool visited[]);
public:
Graph(int V);
void addEdge(int v, int w);
// The main function that finds and prints strongly connected
// components
void printSCCs();
// Function that returns reverse (or transpose) of this graph
Graph getTranspose();
};
Graph::Graph(int V)
{
this->V = V;
adj = new list<int>[V];
}
// A recursive function to print DFS starting from v
void Graph::DFSUtil(int v, bool visited[])
{
// Mark the current node as visited and print it
visited[v] = true;
cout << v << " ";
// Recur for all the vertices adjacent to this vertex
list<int>::iterator i;
for (i = adj[v].begin(); i != adj[v].end(); ++i)
if (!visited[*i])
DFSUtil(*i, visited);
}
Graph Graph::getTranspose()
{
Graph g(V);
for (int v = 0; v < V; v++)
{
// Recur for all the vertices adjacent to this vertex
list<int>::iterator i;
for(i = adj[v].begin(); i != adj[v].end(); ++i)
{
g.adj[*i].push_back(v);
}
}
return g;
}
void Graph::addEdge(int v, int w)
{
adj[v].push_back(w); // Add w to vs list.
}
void Graph::fillOrder(int v, bool visited[], stack<int> &Stack)
{
// Mark the current node as visited and print it
visited[v] = true;
// Recur for all the vertices adjacent to this vertex
list<int>::iterator i;
for(i = adj[v].begin(); i != adj[v].end(); ++i)
if(!visited[*i])
fillOrder(*i, visited, Stack);
// All vertices reachable from v are processed by now, push v
Stack.push(v);
}
// The main function that finds and prints all strongly connected
// components
void Graph::printSCCs()
{

stack<int> Stack;
// Mark all the vertices as not visited (For first DFS)
bool *visited = new bool[V];
for(int i = 0; i < V; i++)
visited[i] = false;
// Fill vertices in stack according to their finishing times
for(int i = 0; i < V; i++)
if(visited[i] == false)
fillOrder(i, visited, Stack);
// Create a reversed graph
Graph gr = getTranspose();
// Mark all the vertices as not visited (For second DFS)
for(int i = 0; i < V; i++)
visited[i] = false;
// Now process all vertices in order defined by Stack
while (Stack.empty() == false)
{
// Pop a vertex from stack
int v = Stack.top();
Stack.pop();
// Print Strongly connected component of the popped vertex
if (visited[v] == false)
{
gr.DFSUtil(v, visited);
cout << endl;
}
}
}
// Driver program to test above functions
int main()
{
// Create a graph given in the above diagram
Graph g(5);
g.addEdge(1, 0);
g.addEdge(0, 2);
g.addEdge(2, 1);
g.addEdge(0, 3);
g.addEdge(3, 4);
cout << "Following are strongly connected components in "
"given graph \n";
g.printSCCs();
return 0;
}

Java
// Java implementation of Kosaraju's algorithm to print all SCCs
import java.io.*;
import java.util.*;
import java.util.LinkedList;
// This class represents a directed graph using adjacency list
// representation
class Graph
{
private int V; // No. of vertices
private LinkedList<Integer> adj[]; //Adjacency List
//Constructor
Graph(int v)
{
V = v;
adj = new LinkedList[v];
for (int i=0; i<v; ++i)
adj[i] = new LinkedList();
}
//Function to add an edge into the graph
void addEdge(int v, int w) { adj[v].add(w); }
// A recursive function to print DFS starting from v

void DFSUtil(int v,boolean visited[])

{
// Mark the current node as visited and print it
visited[v] = true;
System.out.print(v + " ");
int n;
// Recur for all the vertices adjacent to this vertex
Iterator<Integer> i =adj[v].iterator();
while (i.hasNext())
{
n = i.next();
if (!visited[n])
DFSUtil(n,visited);
}
}
// Function that returns reverse (or transpose) of this graph
Graph getTranspose()
{
Graph g = new Graph(V);
for (int v = 0; v < V; v++)
{
// Recur for all the vertices adjacent to this vertex
Iterator<Integer> i =adj[v].listIterator();
while(i.hasNext())
g.adj[i.next()].add(v);
}
return g;
}
void fillOrder(int v, boolean visited[], Stack stack)
{
// Mark the current node as visited and print it
visited[v] = true;
// Recur for all the vertices adjacent to this vertex
Iterator<Integer> i = adj[v].iterator();
while (i.hasNext())
{
int n = i.next();
if(!visited[n])
fillOrder(n, visited, stack);
}
// All vertices reachable from v are processed by now,
// push v to Stack
stack.push(new Integer(v));
}
// The main function that finds and prints all strongly
// connected components
void printSCCs()
{
Stack stack = new Stack();
// Mark all the vertices as not visited (For first DFS)
boolean visited[] = new boolean[V];
for(int i = 0; i < V; i++)
visited[i] = false;
// Fill vertices in stack according to their finishing
// times
for (int i = 0; i < V; i++)
if (visited[i] == false)
fillOrder(i, visited, stack);
// Create a reversed graph
Graph gr = getTranspose();
// Mark all the vertices as not visited (For second DFS)
for (int i = 0; i < V; i++)
visited[i] = false;
// Now process all vertices in order defined by Stack
while (stack.empty() == false)
{
// Pop a vertex from stack
int v = (int)stack.pop();

// Print Strongly connected component of the popped vertex

if (visited[v] == false)
{
gr.DFSUtil(v, visited);
System.out.println();
}
}
}
// Driver method
public static void main(String args[])
{
// Create a graph given in the above diagram
Graph g = new Graph(5);
g.addEdge(1, 0);
g.addEdge(0, 2);
g.addEdge(2, 1);
g.addEdge(0, 3);
g.addEdge(3, 4);
System.out.println("Following are strongly connected components "+
"in given graph ");
g.printSCCs();
}
}
// This code is contributed by Aakash Hasija

Following are strongly connected components in given graph

0 1 2
3
4

Time Complexity: The above algorithm calls DFS, fins reverse of the graph and again calls DFS. DFS takes O(V+E) for a graph represented
using adjacency list. Reversing a graph also takes O(V+E) time. For reversing the graph, we simple traverse all adjacency lists.
The above algorithm is asymptotically best algorithm, but there are other algorithms like Tarjans algorithm and path-based which have same time
complexity but find SCCs using single DFS. The Tarjans algorithm is discussed in the following post.
Tarjans Algorithm to find Strongly Connected Components
Applications:
SCC algorithms can be used as a first step in many graph algorithms that work only on strongly connected graph.
In social networks, a group of people are generally strongly connected (For example, students of a class or any other common place). Many
people in these groups generally like some common pages or play common games. The SCC algorithms can be used to find such groups and
suggest the commonly liked pages or games to the people in the group who have not yet liked commonly liked a page or played a game.
References:
http://en.wikipedia.org/wiki/Kosaraju%27s_algorithm
https://www.youtube.com/watch?v=PZQ0Pdk15RA
You may also like to see Tarjans Algorithm to find Strongly Connected Components.

Transitive closure of a graph

Given a directed graph, find out if a vertex j is reachable from another vertex i for all vertex pairs (i, j) in the given graph. Here reachable mean that
there is a path from vertex i to j. The reach-ability matrix is called transitive closure of a graph.
The graph is given in the form of adjacency matrix say graph[V][V] where graph[i][j] is 1 if there is an edge from vertex i to vertex j or i is equal to
j, otherwise graph[i][j] is 0.
Floyd Warshall Algorithm can be used, we can calculate the distance matrix dist[V][V] using Floyd Warshall, if dist[i][j] is infinite, then j is not
reachable from i, otherwise j is reachable and value of dist[i][j] will be less than V.
Instead of directly using Floyd Warshall, we can optimize it in terms of space and time, for this particular problem. Following are the optimizations:
1) Instead of integer resultant matrix (dist[V][V] in floyd warshall), we can create a boolean reach-ability matrix reach[V][V] (we save space). The
value reach[i][j] will be 1 if j is reachable from i, otherwise 0.
2) Instead of using arithmetic operations, we can use logical operations. For arithmetic operation +, logical and && is used, and for min, logical or
|| is used. (We save time by a constant factor. Time complexity is same though)

C++
// Program for transitive closure using Floyd Warshall Algorithm
#include<stdio.h>
// Number of vertices in the graph
#define V 4
// A function to print the solution matrix
void printSolution(int reach[][V]);
// Prints transitive closure of graph[][] using Floyd Warshall algorithm
void transitiveClosure(int graph[][V])
{
/* reach[][] will be the output matrix that will finally have the
shortest distances between every pair of vertices */
int reach[V][V], i, j, k;
/* Initialize the solution matrix same as input graph matrix. Or
we can say the initial values of shortest distances are based
on shortest paths considering no intermediate vertex. */
for (i = 0; i < V; i++)
for (j = 0; j < V; j++)
reach[i][j] = graph[i][j];
/* Add all vertices one by one to the set of intermediate vertices.
---> Before start of a iteration, we have reachability values for
all pairs of vertices such that the reachability values
consider only the vertices in set {0, 1, 2, .. k-1} as
intermediate vertices.
----> After the end of a iteration, vertex no. k is added to the
set of intermediate vertices and the set becomes {0, 1, .. k} */
for (k = 0; k < V; k++)
{
// Pick all vertices as source one by one
for (i = 0; i < V; i++)
{
// Pick all vertices as destination for the
// above picked source
for (j = 0; j < V; j++)
{
// If vertex k is on a path from i to j,
// then make sure that the value of reach[i][j] is 1
reach[i][j] = reach[i][j] || (reach[i][k] && reach[k][j]);
}
}
}
// Print the shortest distance matrix
printSolution(reach);
}
/* A utility function to print solution */
void printSolution(int reach[][V])
{
printf ("Following matrix is transitive closure of the given graph\n");
for (int i = 0; i < V; i++)
{
for (int j = 0; j < V; j++)
printf ("%d ", reach[i][j]);

printf("\n");
}
}
// driver program to test above function
int main()
{
/* Let us create the following weighted graph
10
(0)------->(3)
|
/|\
5 |
|
|
| 1
\|/
|
(1)------->(2)
3
*/
int graph[V][V] = { {1, 1, 0, 1},
{0, 1, 1, 0},
{0, 0, 1, 1},
{0, 0, 0, 1}
};
// Print the solution
transitiveClosure(graph);
return 0;
}

Java
// Program for transitive closure using Floyd Warshall Algorithm
import java.util.*;
import java.lang.*;
import java.io.*;
class GraphClosure
{
final static int V = 4; //Number of vertices in a graph
// Prints transitive closure of graph[][] using Floyd
// Warshall algorithm
void transitiveClosure(int graph[][])
{
/* reach[][] will be the output matrix that will finally
have the shortest distances between every pair of
vertices */
int reach[][] = new int[V][V];
int i, j, k;
/* Initialize the solution matrix same as input graph
matrix. Or we can say the initial values of shortest
distances are based on shortest paths considering
no intermediate vertex. */
for (i = 0; i < V; i++)
for (j = 0; j < V; j++)
reach[i][j] = graph[i][j];
/* Add all vertices one by one to the set of intermediate
vertices.
---> Before start of a iteration, we have reachability
values for all pairs of vertices such that the
reachability values consider only the vertices in
set {0, 1, 2, .. k-1} as intermediate vertices.
----> After the end of a iteration, vertex no. k is
added to the set of intermediate vertices and the
set becomes {0, 1, 2, .. k} */
for (k = 0; k < V; k++)
{
// Pick all vertices as source one by one
for (i = 0; i < V; i++)
{
// Pick all vertices as destination for the
// above picked source
for (j = 0; j < V; j++)
{
// If vertex k is on a path from i to j,
// then make sure that the value of reach[i][j] is 1
reach[i][j] = (reach[i][j]!=0) ||
((reach[i][k]!=0) && (reach[k][j]!=0))?1:0;
}
}

}
// Print the shortest distance matrix
printSolution(reach);
}
/* A utility function to print solution */
void printSolution(int reach[][])
{
System.out.println("Following matrix is transitive closure"+
" of the given graph");
for (int i = 0; i < V; i++)
{
for (int j = 0; j < V; j++)
System.out.print(reach[i][j]+" ");
System.out.println();
}
}
// Driver program to test above function
public static void main (String[] args)
{
/* Let us create the following weighted graph
10
(0)------->(3)
|
/|\
5 |
|
|
| 1
\|/
|
(1)------->(2)
3
*/
/* Let us create the following weighted graph
10
(0)------->(3)
|
/|\
5 |
|
|
| 1
\|/
|
(1)------->(2)
3
*/
int graph[][] = new int[][]{ {1,
{0,
{0,
{0,
};

1,
1,
0,
0,

0,
1,
1,
0,

1},
0},
1},
1}

// Print the solution

GraphClosure g = new GraphClosure();
g.transitiveClosure(graph);
}
}
// This code is contributed by Aakash Hasija

Following matrix is transitive closure of the given graph

1 1 1 1
0 1 1 1
0 0 1 1
0 0 0 1

Time Complexity: O(V^3) where V is number of vertices in the given graph.

References:
Introduction to Algorithms by Clifford Stein, Thomas H. Cormen, Charles E. Leiserson, Ronald L.

Find the number of islands

Given a boolean 2D matrix, find the number of islands.
This is an variation of the standard problem: Counting number of connected components in a undirected graph.
Before we go to the problem, let us understand what is a connected component. A connected component of an undirected graph is a subgraph in
which every two vertices are connected to each other by a path(s), and which is connected to no other vertices outside the subgraph.
For example, the graph shown below has three connected components.

A graph where all vertices are connected with each other, has exactly one connected component, consisting of the whole graph. Such graph with
only one connected component is called as Strongly Connected Graph.
The problem can be easily solved by applying DFS() on each component. In each DFS() call, a component or a sub-graph is visited. We will call
DFS on the next un-visited component. The number of calls to DFS() gives the number of connected components. BFS can also be used.
What is an island?
A group of connected 1s forms an island. For example, the below matrix contains 5 islands
{1, 1, 0, 0, 0},
{0, 1, 0,
{1, 0, 0,
{0, 0, 0,
{1, 0, 1,

0,
1,
0,
0,

1},
1},
0},
1}

A cell in 2D matrix can be connected to 8 neighbors. So, unlike standard DFS(), where we recursively call for all adjacent vertices, here we can
recursive call for 8 neighbors only. We keep track of the visited 1s so that they are not visited again.

C/C++
// Program to count islands in boolean 2D matrix
#include <stdio.h>
#include <string.h>
#include <stdbool.h>
#define ROW 5
#define COL 5
// A function to check if a given cell (row, col) can be included in DFS
int isSafe(int M[][COL], int row, int col, bool visited[][COL])
{
// row number is in range, column number is in range and value is 1
// and not yet visited
return (row >= 0) && (row < ROW) &&
(col >= 0) && (col < COL) &&
(M[row][col] && !visited[row][col]);
}
// A utility function to do DFS for a 2D boolean matrix. It only considers
// the 8 neighbours as adjacent vertices
void DFS(int M[][COL], int row, int col, bool visited[][COL])
{
// These arrays are used to get row and column numbers of 8 neighbours
// of a given cell
static int rowNbr[] = {-1, -1, -1, 0, 0, 1, 1, 1};
static int colNbr[] = {-1, 0, 1, -1, 1, -1, 0, 1};
// Mark this cell as visited
visited[row][col] = true;
// Recur for all connected neighbours
for (int k = 0; k < 8; ++k)

if (isSafe(M, row + rowNbr[k], col + colNbr[k], visited) )

DFS(M, row + rowNbr[k], col + colNbr[k], visited);
}
// The main function that returns count of islands in a given boolean
// 2D matrix
int countIslands(int M[][COL])
{
// Make a bool array to mark visited cells.
// Initially all cells are unvisited
bool visited[ROW][COL];
memset(visited, 0, sizeof(visited));
// Initialize count as 0 and travese through the all cells of
// given matrix
int count = 0;
for (int i = 0; i < ROW; ++i)
for (int j = 0; j < COL; ++j)
if (M[i][j] && !visited[i][j]) // If a cell with value 1 is not
{
// visited yet, then new island found
DFS(M, i, j, visited);
// Visit all cells in this island.
++count;
// and increment island count
}
return count;
}
// Driver program
int main()
{
int M[][COL]=
{0, 1, 0,
{1, 0, 0,
{0, 0, 0,
{1, 0, 1,
};

to test above function

{
0,
1,
0,
0,

{1, 1, 0, 0, 0},
1},
1},
0},
1}

printf("Number of islands is: %d\n", countIslands(M));

return 0;
}

Java
// Java program to count islands in boolean 2D matrix
import java.util.*;
import java.lang.*;
import java.io.*;
class Islands
{
//No of rows and columns
static final int ROW = 5, COL = 5;
// A function to check if a given cell (row, col) can
// be included in DFS
boolean isSafe(int M[][], int row, int col,
boolean visited[][])
{
// row number is in range, column number is in range
// and value is 1 and not yet visited
return (row >= 0) && (row < ROW) &&
(col >= 0) && (col < COL) &&
(M[row][col]==1 && !visited[row][col]);
}
// A utility function to do DFS for a 2D boolean matrix.
// It only considers the 8 neighbors as adjacent vertices
void DFS(int M[][], int row, int col, boolean visited[][])
{
// These arrays are used to get row and column numbers
// of 8 neighbors of a given cell
int rowNbr[] = new int[] {-1, -1, -1, 0, 0, 1, 1, 1};
int colNbr[] = new int[] {-1, 0, 1, -1, 1, -1, 0, 1};
// Mark this cell as visited
visited[row][col] = true;
// Recur for all connected neighbours
for (int k = 0; k < 8; ++k)

if (isSafe(M, row + rowNbr[k], col + colNbr[k], visited) )

DFS(M, row + rowNbr[k], col + colNbr[k], visited);
}
// The main function that returns count of islands in a given
// boolean 2D matrix
int countIslands(int M[][])
{
// Make a bool array to mark visited cells.
// Initially all cells are unvisited
boolean visited[][] = new boolean[ROW][COL];
// Initialize count as 0 and travese through the all cells
// of given matrix
int count = 0;
for (int i = 0; i < ROW; ++i)
for (int j = 0; j < COL; ++j)
if (M[i][j]==1 && !visited[i][j]) // If a cell with
{
// value 1 is not
// visited yet, then new island found, Visit all
// cells in this island and increment island count
DFS(M, i, j, visited);
++count;
}
return count;
}
// Driver method
public static void main (String[] args) throws java.lang.Exception
{
int M[][]= new int[][] {{1, 1, 0, 0, 0},
{0, 1, 0, 0, 1},
{1, 0, 0, 1, 1},
{0, 0, 0, 0, 0},
{1, 0, 1, 0, 1}
};
Islands I = new Islands();
System.out.println("Number of islands is: "+ I.countIslands(M));
}
} //Contributed by Aakash Hasija

Output:
Number of islands is: 5

Time complexity: O(ROW x COL)

Reference:
http://en.wikipedia.org/wiki/Connected_component_%28graph_theory%29

Count all possible walks from a source to a destination with exactly k edges
Given a directed graph and two vertices u and v in it, count all possible walks from u to v with exactly k edges on the walk.
The graph is given as adjacency matrix representation where value of graph[i][j] as 1 indicates that there is an edge from vertex i to vertex j and a
value 0 indicates no edge from i to j.
For example consider the following graph. Let source u be vertex 0, destination v be 3 and k be 2. The output should be 2 as there are two walk
from 0 to 3 with exactly 2 edges. The walks are {0, 2, 3} and {0, 1, 3}

A simple solution is to start from u, go to all adjacent vertices and recur for adjacent vertices with k as k-1, source as adjacent vertex and
destination as v. Following is C++ implementation of this simple solution.

C++
// C++ program to count walks from u to v with exactly k edges
#include <iostream>
using namespace std;
// Number of vertices in the graph
#define V 4
// A naive recursive function to count walks from u to v with k edges
int countwalks(int graph[][V], int u, int v, int k)
{
// Base cases
if (k == 0 && u == v)
return 1;
if (k == 1 && graph[u][v]) return 1;
if (k <= 0)
return 0;
// Initialize result
int count = 0;
// Go to all adjacents of u and recur
for (int i = 0; i < V; i++)
if (graph[u][i] == 1) // Check if is adjacent of u
count += countwalks(graph, i, v, k-1);
return count;
}
// driver program to test above function
int main()
{
/* Let us create the graph shown in above diagram*/
int graph[V][V] = { {0, 1, 1, 1},
{0, 0, 0, 1},
{0, 0, 0, 1},
{0, 0, 0, 0}
};
int u = 0, v = 3, k = 2;
cout << countwalks(graph, u, v, k);
return 0;
}

Java
// Java program to count walks from u to v with exactly k edges
import java.util.*;
import java.lang.*;

import java.io.*;
class KPaths
{
static final int V = 4; //Number of vertices
// A naive recursive function to count walks from u
// to v with k edges
int countwalks(int graph[][], int u, int v, int k)
{
// Base cases
if (k == 0 && u == v)
return 1;
if (k == 1 && graph[u][v] == 1) return 1;
if (k <= 0)
return 0;
// Initialize result
int count = 0;
// Go to all adjacents of u and recur
for (int i = 0; i < V; i++)
if (graph[u][i] == 1) // Check if is adjacent of u
count += countwalks(graph, i, v, k-1);
return count;
}
// Driver method
public static void main (String[] args) throws java.lang.Exception
{
/* Let us create the graph shown in above diagram*/
int graph[][] =new int[][] { {0, 1, 1, 1},
{0, 0, 0, 1},
{0, 0, 0, 1},
{0, 0, 0, 0}
};
int u = 0, v = 3, k = 2;
KPaths p = new KPaths();
System.out.println(p.countwalks(graph, u, v, k));
}
}//Contributed by Aakash Hasija

The worst case time complexity of the above function is O(Vk) where V is the number of vertices in the given graph. We can simply analyze the
time complexity by drawing recursion tree. The worst occurs for a complete graph. In worst case, every internal node of recursion tree would have
exactly n children.
We can optimize the above solution using Dynamic Programming. The idea is to build a 3D table where first dimension is source, second
dimension is destination, third dimension is number of edges from source to destination, and the value is count of walks. Like other Dynamic
Programming problems, we fill the 3D table in bottom up manner.

C++
// C++ program to count walks from u to v with exactly k edges
#include <iostream>
using namespace std;
// Number of vertices in the graph
#define V 4
// A Dynamic programming based function to count
// to v with k edges
int countwalks(int graph[][V], int u, int v, int
{
// Table to be filled up using DP. The value
// store count of possible walks from i to j
int count[V][V][k+1];

walks from u
k)
count[i][j][e] will
with exactly k edges

// Loop for number of edges from 0 to k

for (int e = 0; e <= k; e++)
{
for (int i = 0; i < V; i++) // for source
{
for (int j = 0; j < V; j++) // for destination
{
// initialize value
count[i][j][e] = 0;

// from base cases

if (e == 0 && i == j)
count[i][j][e] = 1;
if (e == 1 && graph[i][j])
count[i][j][e] = 1;
// go to adjacent only when number of edges is more than 1
if (e > 1)
{
for (int a = 0; a < V; a++) // adjacent of source i
if (graph[i][a])
count[i][j][e] += count[a][j][e-1];
}
}
}
}
return count[u][v][k];
}
// driver program to test above function
int main()
{
/* Let us create the graph shown in above diagram*/
int graph[V][V] = { {0, 1, 1, 1},
{0, 0, 0, 1},
{0, 0, 0, 1},
{0, 0, 0, 0}
};
int u = 0, v = 3, k = 2;
cout << countwalks(graph, u, v, k);
return 0;
}

Java
// Java program to count walks from u to v with exactly k edges
import java.util.*;
import java.lang.*;
import java.io.*;
class KPaths
{
static final int V = 4; //Number of vertices
// A Dynamic programming based function to count walks from u
// to v with k edges
int countwalks(int graph[][], int u, int v, int k)
{
// Table to be filled up using DP. The value count[i][j][e]
// will/ store count of possible walks from i to j with
// exactly k edges
int count[][][] = new int[V][V][k+1];
// Loop for number of edges from 0 to k
for (int e = 0; e <= k; e++)
{
for (int i = 0; i < V; i++) // for source
{
for (int j = 0; j < V; j++) // for destination
{
// initialize value
count[i][j][e] = 0;
// from base cases
if (e == 0 && i == j)
count[i][j][e] = 1;
if (e == 1 && graph[i][j]!=0)
count[i][j][e] = 1;
// go to adjacent only when number of edges
// is more than 1
if (e > 1)
{
for (int a = 0; a < V; a++) // adjacent of i
if (graph[i][a]!=0)
count[i][j][e] += count[a][j][e-1];
}
}
}

}
return count[u][v][k];
}
// Driver method
public static void main (String[] args) throws java.lang.Exception
{
/* Let us create the graph shown in above diagram*/
int graph[][] =new int[][] { {0, 1, 1, 1},
{0, 0, 0, 1},
{0, 0, 0, 1},
{0, 0, 0, 0}
};
int u = 0, v = 3, k = 2;
KPaths p = new KPaths();
System.out.println(p.countwalks(graph, u, v, k));
}
}//Contributed by Aakash Hasija

Time complexity of the above DP based solution is O(V3K) which is much better than the naive solution.
We can also use Divide and Conquer to solve the above problem in O(V3Logk) time. The count of walks of length k from u to v is the [u][v]th
entry in (graph[V][V])k. We can calculate power of by doing O(Logk) multiplication by using the divide and conquer technique to calculate power.
A multiplication between two matrices of size V x V takes O(V3) time. Therefore overall time complexity of this method is O(V3Logk).

Euler Circuit in a Directed Graph

Eulerian Path is a path in graph that visits every edge exactly once. Eulerian Circuit is an Eulerian Path which starts and ends on the same vertex.
A graph is said to be eulerian if it has eulerian cycle. We have discussed eulerian circuit for an undirected graph. In this post, same is discussed for
a directed graph.
For example, the following graph has eulerian cycle as {1, 0, 3, 4, 0, 2, 1}

How to check if a directed graph is eulerian?

A directed graph has an eulerian cycle if following conditions are true (Source: Wiki)
1) All vertices with nonzero degree belong to a single strongly connected component.
2) In degree and out degree of every vertex is same.
We can detect singly connected component using Kosarajus DFS based simple algorithm.
To compare in degree and out degree, we need to store in degree an out degree of every vertex. Out degree can be obtained by size of adjacency
list. In degree can be stored by creating an array of size equal to number of vertices.
Following are C++ and Java implementations of above approach.

C++
// A C++ program to check if a given directed graph is Eulerian or not
#include<iostream>
#include <list>
#define CHARS 26
using namespace std;
// A class that represents an undirected graph
class Graph
{
int V;
// No. of vertices
list<int> *adj;
// A dynamic array of adjacency lists
int *in;
public:
// Constructor and destructor
Graph(int V);
~Graph() { delete [] adj; delete [] in; }
// function to add an edge to graph
void addEdge(int v, int w) { adj[v].push_back(w); (in[w])++; }
// Method to check if this graph is Eulerian or not
bool isEulerianCycle();
// Method to check if all non-zero degree vertices are connected
bool isSC();
// Function to do DFS starting from v. Used in isConnected();
void DFSUtil(int v, bool visited[]);
Graph getTranspose();
};
Graph::Graph(int V)
{
this->V = V;
adj = new list<int>[V];
in = new int[V];
for (int i = 0; i < V; i++)
in[i] = 0;
}
/* This function returns true if the directed graph has an eulerian
cycle, otherwise returns false */
bool Graph::isEulerianCycle()
{

// Check if all non-zero degree vertices are connected

if (isSC() == false)
return false;
// Check if in degree and out degree of every vertex is same
for (int i = 0; i < V; i++)
if (adj[i].size() != in[i])
return false;
return true;
}
// A recursive function to do DFS starting from v
void Graph::DFSUtil(int v, bool visited[])
{
// Mark the current node as visited and print it
visited[v] = true;
// Recur for all the vertices adjacent to this vertex
list<int>::iterator i;
for (i = adj[v].begin(); i != adj[v].end(); ++i)
if (!visited[*i])
DFSUtil(*i, visited);
}
// Function that returns reverse (or transpose) of this graph
// This function is needed in isSC()
Graph Graph::getTranspose()
{
Graph g(V);
for (int v = 0; v < V; v++)
{
// Recur for all the vertices adjacent to this vertex
list<int>::iterator i;
for(i = adj[v].begin(); i != adj[v].end(); ++i)
{
g.adj[*i].push_back(v);
(g.in[v])++;
}
}
return g;
}
// This function returns true if all non-zero degree vertices of
// graph are strongly connected (Please refer
// http://www.geeksforgeeks.org/connectivity-in-a-directed-graph/ )
bool Graph::isSC()
{
// Mark all the vertices as not visited (For first DFS)
bool visited[V];
for (int i = 0; i < V; i++)
visited[i] = false;
// Find the first vertex with non-zero degree
int n;
for (n = 0; n < V; n++)
if (adj[n].size() > 0)
break;
// Do DFS traversal starting from first non zero degree vertex.
DFSUtil(n, visited);
// If DFS traversal doesn't visit all vertices, then return false.
for (int i = 0; i < V; i++)
if (adj[i].size() > 0 && visited[i] == false)
return false;
// Create a reversed graph
Graph gr = getTranspose();
// Mark all the vertices as not visited (For second DFS)
for (int i = 0; i < V; i++)
visited[i] = false;
// Do DFS for reversed graph starting from first vertex.
// Staring Vertex must be same starting point of first DFS
gr.DFSUtil(n, visited);
// If all vertices are not visited in second DFS, then
// return false
for (int i = 0; i < V; i++)

if (adj[i].size() > 0 && visited[i] == false)

return false;
return true;
}
// Driver program to test above functions
int main()
{
// Create a graph given in the above diagram
Graph g(5);
g.addEdge(1, 0);
g.addEdge(0, 2);
g.addEdge(2, 1);
g.addEdge(0, 3);
g.addEdge(3, 4);
g.addEdge(4, 0);
if (g.isEulerianCycle())
cout << "Given directed graph is eulerian \n";
else
cout << "Given directed graph is NOT eulerian \n";
return 0;
}

Java
// A Java program to check if a given directed graph is Eulerian or not
// A class that represents an undirected graph
import java.io.*;
import java.util.*;
import java.util.LinkedList;
// This class represents a directed graph using adjacency list
class Graph
{
private int V; // No. of vertices
private LinkedList<Integer> adj[];//Adjacency List
private int in[]; //maintaining in degree
//Constructor
Graph(int v)
{
V = v;
adj = new LinkedList[v];
in = new int[V];
for (int i=0; i<v; ++i)
{
adj[i] = new LinkedList();
in[i] = 0;
}
}
//Function to add an edge into the graph
void addEdge(int v,int w)
{
adj[v].add(w);
in[w]++;
}
// A recursive function to print DFS starting from v
void DFSUtil(int v,Boolean visited[])
{
// Mark the current node as visited
visited[v] = true;
int n;
// Recur for all the vertices adjacent to this vertex
Iterator<Integer> i =adj[v].iterator();
while (i.hasNext())
{
n = i.next();
if (!visited[n])
DFSUtil(n,visited);
}
}
// Function that returns reverse (or transpose) of this graph

Graph getTranspose()
{
Graph g = new Graph(V);
for (int v = 0; v < V; v++)
{
// Recur for all the vertices adjacent to this vertex
Iterator<Integer> i = adj[v].listIterator();
while (i.hasNext())
{
g.adj[i.next()].add(v);
(g.in[v])++;
}
}
return g;
}
// The main function that returns true if graph is strongly
// connected
Boolean isSC()
{
// Step 1: Mark all the vertices as not visited (For
// first DFS)
Boolean visited[] = new Boolean[V];
for (int i = 0; i < V; i++)
visited[i] = false;
// Step 2: Do DFS traversal starting from first vertex.
DFSUtil(0, visited);
// If DFS traversal doesn't visit all vertices, then return false.
for (int i = 0; i < V; i++)
if (visited[i] == false)
return false;
// Step 3: Create a reversed graph
Graph gr = getTranspose();
// Step 4: Mark all the vertices as not visited (For second DFS)
for (int i = 0; i < V; i++)
visited[i] = false;
// Step 5: Do DFS for reversed graph starting from first vertex.
// Staring Vertex must be same starting point of first DFS
gr.DFSUtil(0, visited);
// If all vertices are not visited in second DFS, then
// return false
for (int i = 0; i < V; i++)
if (visited[i] == false)
return false;
return true;
}
/* This function returns true if the directed graph has an eulerian
cycle, otherwise returns false */
Boolean isEulerianCycle()
{
// Check if all non-zero degree vertices are connected
if (isSC() == false)
return false;
// Check if in degree and out degree of every vertex is same
for (int i = 0; i < V; i++)
if (adj[i].size() != in[i])
return false;
return true;
}
public static void main (String[] args) throws java.lang.Exception
{
Graph g = new Graph(5);
g.addEdge(1, 0);
g.addEdge(0, 2);
g.addEdge(2, 1);
g.addEdge(0, 3);
g.addEdge(3, 4);
g.addEdge(4, 0);
if (g.isEulerianCycle())

System.out.println("Given directed graph is eulerian ");

else
System.out.println("Given directed graph is NOT eulerian ");
}
}
//This code is contributed by Aakash Hasija

Given directed graph is eulerian

Time complexity of the above implementation is O(V + E) as Kosarajus algorithm takes O(V + E) time. After running Kosarajus algorithm we
traverse all vertices and compare in degree with out degree which takes O(V) time.
See following as an application of this.
Find if the given array of strings can be chained to form a circle.

Biconnected Components
A biconnected component is a maximal biconnected subgraph.
Biconnected Graph is already discussed here. In this article, we will see how to find biconnected component in a graph using algorithm by John
Hopcroft and Robert Tarjan.

In above graph, following are the biconnected components:

42 34 31 23 12
89
85 78 57
60 56 15 01
1011
Algorithm is based on Disc and Low Values discussed in Strongly Connected Components Article.
Idea is to store visited edges in a stack while DFS on a graph and keep looking for Articulation Points (highlighted in above figure). As soon as an
Articulation Point u is found, all edges visited while DFS from node u onwards will form one biconnected component. When DFS completes for
one connected component, all edges present in stack will form a biconnected component.
If there is no Articulation Point in graph, then graph is biconnected and so there will be one biconnected component which is the graph itself.

C++
// A C++ program to find biconnected components in a given undirected graph
#include<iostream>
#include <list>
#include <stack>
#define NIL -1
using namespace std;
int count = 0;
class Edge
{
public:
int u;
int v;
Edge(int u, int v);
};
Edge::Edge(int u, int v)
{
this->u = u;
this->v = v;
}
// A class that represents an directed graph
class Graph
{
int V;
// No. of vertices
int E;
// No. of edges
list<int> *adj;
// A dynamic array of adjacency lists
// A Recursive DFS based function used by BCC()
void BCCUtil(int u, int disc[], int low[],
list<Edge> *st, int parent[]);
public:
Graph(int V); // Constructor
void addEdge(int v, int w); // function to add an edge to graph
void BCC();
// prints strongly connected components
};
Graph::Graph(int V)
{
this->V = V;
this->E = 0;

adj = new list<int>[V];

}
void Graph::addEdge(int v, int w)
{
adj[v].push_back(w);
E++;
}
// A recursive function that finds and prints strongly connected
// components using DFS traversal
// u --> The vertex to be visited next
// disc[] --> Stores discovery times of visited vertices
// low[] -- >> earliest visited vertex (the vertex with minimum
//
discovery time) that can be reached from subtree
//
rooted with current vertex
// *st -- >> To store visited edges
void Graph::BCCUtil(int u, int disc[], int low[], list<Edge> *st,
int parent[])
{
// A static variable is used for simplicity, we can avoid use
// of static variable by passing a pointer.
static int time = 0;
// Initialize discovery time and low value
disc[u] = low[u] = ++time;
int children = 0;
// Go through all vertices adjacent to this
list<int>::iterator i;
for (i = adj[u].begin(); i != adj[u].end(); ++i)
{
int v = *i; // v is current adjacent of 'u'
// If v is not visited yet, then recur for it
if (disc[v] == -1)
{
children++;
parent[v] = u;
//store the edge in stack
st->push_back(Edge(u,v));
BCCUtil(v, disc, low, st, parent);
// Check if the subtree rooted with 'v' has a
// connection to one of the ancestors of 'u'
// Case 1 -- per Strongly Connected Components Article
low[u] = min(low[u], low[v]);
//If u is an articulation point,
//pop all edges from stack till u -- v
if( (disc[u] == 1 && children > 1) ||
(disc[u] > 1 && low[v] >= disc[u]) )
{
while(st->back().u != u || st->back().v != v)
{
cout << st->back().u << "--" << st->back().v << " ";
st->pop_back();
}
cout << st->back().u << "--" << st->back().v;
st->pop_back();
cout << endl;
count++;
}
}
// Update low value of 'u' only of 'v' is still in stack
// (i.e. it's a back edge, not cross edge).
// Case 2 -- per Strongly Connected Components Article
else if(v != parent[u] && disc[v] < low[u])
{
low[u] = min(low[u], disc[v]);
st->push_back(Edge(u,v));
}
}
}
// The function to do DFS traversal. It uses BCCUtil()
void Graph::BCC()
{
int *disc = new int[V];
int *low = new int[V];

int *parent = new int[V];

list<Edge> *st = new list<Edge>[E];
// Initialize disc and low, and parent arrays
for (int i = 0; i < V; i++)
{
disc[i] = NIL;
low[i] = NIL;
parent[i] = NIL;
}
for (int i = 0; i < V; i++)
{
if (disc[i] == NIL)
BCCUtil(i, disc, low, st, parent);
int j = 0;
//If stack is not empty, pop all edges from stack
while(st->size() > 0)
{
j = 1;
cout << st->back().u << "--" << st->back().v << " ";
st->pop_back();
}
if(j == 1)
{
cout << endl;
count++;
}
}
}
// Driver program to test above function
int main()
{
Graph g(12);
g.addEdge(0,1);g.addEdge(1,0);
g.addEdge(1,2);g.addEdge(2,1);
g.addEdge(1,3);g.addEdge(3,1);
g.addEdge(2,3);g.addEdge(3,2);
g.addEdge(2,4);g.addEdge(4,2);
g.addEdge(3,4);g.addEdge(4,3);
g.addEdge(1,5);g.addEdge(5,1);
g.addEdge(0,6);g.addEdge(6,0);
g.addEdge(5,6);g.addEdge(6,5);
g.addEdge(5,7);g.addEdge(7,5);
g.addEdge(5,8);g.addEdge(8,5);
g.addEdge(7,8);g.addEdge(8,7);
g.addEdge(8,9);g.addEdge(9,8);
g.addEdge(10,11);g.addEdge(11,10);
g.BCC();
cout << "Above are " << count << " biconnected components in graph";
return 0;
}

Java
// A Java program to find biconnected components in a given
// undirected graph
import java.io.*;
import java.util.*;
// This class represents a directed graph using adjacency
// list representation
class Graph
{
private int V, E; // No. of vertices & Edges respectively
private LinkedList<Integer> adj[]; // Adjacency List
// Count is number of biconnected components. time is
// used to find discovery times
static int count = 0, time = 0;
class Edge
{
int u;
int v;
Edge(int u, int v)
{
this.u = u;

this.v = v;
}
};
//Constructor
Graph(int v)
{
V = v;
E = 0;
adj = new LinkedList[v];
for (int i=0; i<v; ++i)
adj[i] = new LinkedList();
}
//Function to add an edge into the graph
void addEdge(int v,int w)
{
adj[v].add(w);
E++;
}
// A recursive function that finds and prints strongly connected
// components using DFS traversal
// u --> The vertex to be visited next
// disc[] --> Stores discovery times of visited vertices
// low[] -- >> earliest visited vertex (the vertex with minimum
//
discovery time) that can be reached from subtree
//
rooted with current vertex
// *st -- >> To store visited edges
void BCCUtil(int u, int disc[], int low[], LinkedList<Edge>st,
int parent[])
{
// Initialize discovery time and low value
disc[u] = low[u] = ++time;
int children = 0;
// Go through all vertices adjacent to this
Iterator<Integer> it = adj[u].iterator();
while (it.hasNext())
{
int v = it.next(); // v is current adjacent of 'u'
// If v is not visited yet, then recur for it
if (disc[v] == -1)
{
children++;
parent[v] = u;
// store the edge in stack
st.add(new Edge(u,v));
BCCUtil(v, disc, low, st, parent);
//
//
//
if

Check if the subtree rooted with 'v' has a

connection to one of the ancestors of 'u'
Case 1 -- per Strongly Connected Components Article
(low[u] > low[v])
low[u] = low[v];

// If u is an articulation point,
// pop all edges from stack till u -- v
if ( (disc[u] == 1 && children > 1) ||
(disc[u] > 1 && low[v] >= disc[u]) )
{
while (st.getLast().u != u || st.getLast().v != v)
{
System.out.print(st.getLast().u + "--" +
st.getLast().v + " ");
st.removeLast();
}
System.out.println(st.getLast().u + "--" +
st.getLast().v + " ");
st.removeLast();
count++;
}
}
// Update low value of 'u' only of 'v' is still in stack
// (i.e. it's a back edge, not cross edge).
// Case 2 -- per Strongly Connected Components Article

else if (v != parent[u] && disc[v] < low[u])

{
if (low[u]>disc[v])
low[u]=disc[v];
st.add(new Edge(u,v));
}
}
}
// The function to do DFS traversal. It uses BCCUtil()
void BCC()
{
int disc[] = new int[V];
int low[] = new int[V];
int parent[] = new int[V];
LinkedList<Edge> st = new LinkedList<Edge>();
// Initialize disc and low, and parent arrays
for (int i = 0; i < V; i++)
{
disc[i] = -1;
low[i] = -1;
parent[i] = -1;
}
for (int i = 0; i < V; i++)
{
if (disc[i] == -1)
BCCUtil(i, disc, low, st, parent);
int j = 0;
// If stack is not empty, pop all edges from stack
while (st.size() > 0)
{
j = 1;
System.out.print(st.getLast().u + "--" +
st.getLast().v + " ");
st.removeLast();
}
if (j == 1)
{
System.out.println();
count++;
}
}
}
public static void main(String args[])
{
Graph g = new Graph(12);
g.addEdge(0,1);
g.addEdge(1,0);
g.addEdge(1,2);
g.addEdge(2,1);
g.addEdge(1,3);
g.addEdge(3,1);
g.addEdge(2,3);
g.addEdge(3,2);
g.addEdge(2,4);
g.addEdge(4,2);
g.addEdge(3,4);
g.addEdge(4,3);
g.addEdge(1,5);
g.addEdge(5,1);
g.addEdge(0,6);
g.addEdge(6,0);
g.addEdge(5,6);
g.addEdge(6,5);
g.addEdge(5,7);
g.addEdge(7,5);
g.addEdge(5,8);
g.addEdge(8,5);
g.addEdge(7,8);
g.addEdge(8,7);
g.addEdge(8,9);
g.addEdge(9,8);
g.addEdge(10,11);
g.addEdge(11,10);
g.BCC();

System.out.println("Above are " + g.count +

" biconnected components in graph");
}
}
// This code is contributed by Aakash Hasija

Output:
4--2 3--4
8--9
8--5 7--8
6--0 5--6
10--11
Above are

3--1 2--3 1--2

5--7
1--5 0--1
5 biconnected components in graph

Tarjans Algorithm to find Strongly Connected Components

A directed graph is strongly connected if there is a path between all pairs of vertices. A strongly connected component (SCC) of a directed graph
is a maximal strongly connected subgraph. For example, there are 3 SCCs in the following graph.

We have discussed Kosarajus algorithm for strongly connected components. The previously discussed algorithm requires two DFS traversals of a
Graph. In this post, Tarjans algorithm is discussed that requires only one DFS traversal.
Tarjan Algorithm is based on following facts:
1. DFS search produces a DFS tree/forest
2. Strongly Connected Components form subtrees of the DFS tree.
3. If we can find head of such subtrees, we can print/store all the nodes in that subtree (including head) and that will be one SCC.
4. There is no back edge from one SCC to another (There can be cross edges, but cross edges will not be used while processing the graph).
To find head of a SCC, we calculate desc and low array (as done for articulation point, bridge, biconnected component). As discussed in the
previous posts, low[u] indicates earliest visited vertex (the vertex with minimum discovery time) that can be reached from subtree rooted with u. A
node u is head if disc[u] = low[u].
Disc and Low Values
(click on image to see it properly)

Strongly Connected Component relates to directed graph only, but Disc and Low values relate to both directed and undirected graph, so in above
pic we have taken an undirected graph.
In above Figure, we have shown a graph and its one of DFS tree (There could be different DFS trees on same graph depending on order in which
edges are traversed).
In DFS tree, continuous arrows are tree edges and dashed arrows are back edges (DFS Tree Edges
Disc and Low values are showin in Figure for every node as (Disc/Low).
Disc: This is the time when a node is visited 1st time while DFS traversal. For nodes A, B, C, .., J in DFS tree, Disc values are 1, 2, 3, .., 10.
Low: In DFS tree, Tree edges take us forward, from ancestor node to one of its descendants. For example, from node C, tree edges can take us
to node node G, node I etc. Back edges take us backward, from a descendant node to one of its ancestors. For example, from node G, Back
edges take us to E or C. If we look at both Tree and Back edge together, then we can see that if we start traversal from one node, we may go
down the tree via Tree edges and then go up via back edges. For example, from node E, we can go down to G and then go up to C. Similarly
from E, we can go down to I or J and then go up to F. Low value of a node tells the topmost reachable ancestor (with minimum possible Disc
value) via the subtree of that node. So for any node, Low value equal to its Disc value anyway (A node is ancestor of itself). Then we look into its
subtree and see if there is any node which can take us to any of its ancestor. If there are multiple back edges in subtree which take us to different

ancestors, then we take the one with minimum Disc value (i.e. the topmost one). If we look at node F, it has two subtrees. Subtree with node G,
takes us to E and C. The other subtree takes us back to F only. Here topmost ancestor is C where F can reach and so Low value of F is 3 (The
Disc value of C).
Based on above discussion, it should be clear that Low values of B, C, and D are 1 (As A is the topmost node where B, C and D can reach). In
same way, Low values of E, F, G are 3 and Low values of H, I, J are 6.
For any node u, when DFS starts, Low will be set to its Disc 1st.
Then later on DFS will be performed on each of its children v one by one, Low value of u can change it two case:
Case1 (Tree Edge): If node v is not visited already, then after DFS of v is complete, then minimum of low[u] and low[v] will be updated to
low[u].
low[u] = min(low[u], low[v]);
Case 2 (Back Edge): When child v is already visited, then minimum of low[u] and Disc[v] will be updated to low[u].
low[u] = min(low[u], disc[v]);
In case two, can we take low[v] instead of disc[v] ?? . Answer is NO. If you can think why answer is NO, you probably understood the Low and
Disc concept.
Same Low and Disc values help to solve other graph problems like articulation point, bridge and biconnected component.
To track the subtree rooted at head, we can use a stack (keep pushing node while visiting). When a head node found, pop all nodes from stack till
you get head out of stack.
To make sure, we dont consider cross edges, when we reach a node which is already visited, we should process the visited node only if it is
present in stack, else ignore the node.
Following is C++ implementation of Tarjans algorithm to print all SCCs.
// A C++ program to find strongly connected components in a given
// directed graph using Tarjan's algorithm (single DFS)
#include<iostream>
#include <list>
#include <stack>
#define NIL -1
using namespace std;
// A class that represents an directed graph
class Graph
{
int V;
// No. of vertices
list<int> *adj;
// A dynamic array of adjacency lists
// A Recursive DFS based function used by SCC()
void SCCUtil(int u, int disc[], int low[],
stack<int> *st, bool stackMember[]);
public:
Graph(int V); // Constructor
void addEdge(int v, int w); // function to add an edge to graph
void SCC();
// prints strongly connected components
};
Graph::Graph(int V)
{
this->V = V;
adj = new list<int>[V];
}
void Graph::addEdge(int v, int w)
{
adj[v].push_back(w);
}
// A recursive function that finds and prints strongly connected
// components using DFS traversal
// u --> The vertex to be visited next
// disc[] --> Stores discovery times of visited vertices
// low[] -- >> earliest visited vertex (the vertex with minimum
//
discovery time) that can be reached from subtree
//
rooted with current vertex
// *st -- >> To store all the connected ancestors (could be part
//
of SCC)
// stackMember[] --> bit/index array for faster check whether
//
a node is in stack
void Graph::SCCUtil(int u, int disc[], int low[], stack<int> *st,
bool stackMember[])
{
// A static variable is used for simplicity, we can avoid use
// of static variable by passing a pointer.

static int time = 0;

// Initialize discovery time and low value
disc[u] = low[u] = ++time;
st->push(u);
stackMember[u] = true;
// Go through all vertices adjacent to this
list<int>::iterator i;
for (i = adj[u].begin(); i != adj[u].end(); ++i)
{
int v = *i; // v is current adjacent of 'u'
// If v is not visited yet, then recur for it
if (disc[v] == -1)
{
SCCUtil(v, disc, low, st, stackMember);
// Check if the subtree rooted with 'v' has a
// connection to one of the ancestors of 'u'
// Case 1 (per above discussion on Disc and Low value)
low[u] = min(low[u], low[v]);
}
// Update low value of 'u' only of 'v' is still in stack
// (i.e. it's a back edge, not cross edge).
// Case 2 (per above discussion on Disc and Low value)
else if (stackMember[v] == true)
low[u] = min(low[u], disc[v]);
}
// head node found, pop the stack and print an SCC
int w = 0; // To store stack extracted vertices
if (low[u] == disc[u])
{
while (st->top() != u)
{
w = (int) st->top();
cout << w << " ";
stackMember[w] = false;
st->pop();
}
w = (int) st->top();
cout << w << "\n";
stackMember[w] = false;
st->pop();
}
}
// The function to do DFS traversal. It uses SCCUtil()
void Graph::SCC()
{
int *disc = new int[V];
int *low = new int[V];
bool *stackMember = new bool[V];
stack<int> *st = new stack<int>();
// Initialize disc and low, and stackMember arrays
for (int i = 0; i < V; i++)
{
disc[i] = NIL;
low[i] = NIL;
stackMember[i] = false;
}
// Call the recursive helper function to find strongly
// connected components in DFS tree with vertex 'i'
for (int i = 0; i < V; i++)
if (disc[i] == NIL)
SCCUtil(i, disc, low, st, stackMember);
}
// Driver program to test above function
int main()
{
cout << "\nSCCs in first graph \n";
Graph g1(5);
g1.addEdge(1, 0);
g1.addEdge(0, 2);
g1.addEdge(2, 1);
g1.addEdge(0, 3);

g1.addEdge(3, 4);
g1.SCC();
cout << "\nSCCs in second graph \n";
Graph g2(4);
g2.addEdge(0, 1);
g2.addEdge(1, 2);
g2.addEdge(2, 3);
g2.SCC();
cout << "\nSCCs in third graph \n";
Graph g3(7);
g3.addEdge(0, 1);
g3.addEdge(1, 2);
g3.addEdge(2, 0);
g3.addEdge(1, 3);
g3.addEdge(1, 4);
g3.addEdge(1, 6);
g3.addEdge(3, 5);
g3.addEdge(4, 5);
g3.SCC();
cout << "\nSCCs in fourth graph \n";
Graph g4(11);
g4.addEdge(0,1);g4.addEdge(0,3);
g4.addEdge(1,2);g4.addEdge(1,4);
g4.addEdge(2,0);g4.addEdge(2,6);
g4.addEdge(3,2);
g4.addEdge(4,5);g4.addEdge(4,6);
g4.addEdge(5,6);g4.addEdge(5,7);g4.addEdge(5,8);g4.addEdge(5,9);
g4.addEdge(6,4);
g4.addEdge(7,9);
g4.addEdge(8,9);
g4.addEdge(9,8);
g4.SCC();
cout << "\nSCCs in fifth graph \n";
Graph g5(5);
g5.addEdge(0,1);
g5.addEdge(1,2);
g5.addEdge(2,3);
g5.addEdge(2,4);
g5.addEdge(3,0);
g5.addEdge(4,2);
g5.SCC();
return 0;
}

Output:
SCCs in first graph
4
3
1 2 0
SCCs in second graph
3
2
1
0
SCCs in third graph
5
3
4
6
2 1 0
SCCs in fourth graph
8 9
7
5 4 6
3 2 1 0
10
SCCs in fifth graph
4 3 2 1 0

Time Complexity: The above algorithm mainly calls DFS, DFS takes O(V+E) for a graph represented using adjacency list.

References:
http://en.wikipedia.org/wiki/Tarjan%27s_strongly_connected_components_algorithm
http://www.ics.uci.edu/~eppstein/161/960220.html#sca

Graph Coloring | Set 1 (Introduction and Applications)

Graph coloring problem is to assign colors to certain elements of a graph subject to certain constraints.
Vertex coloring is the most common graph coloring problem. The problem is, given m colors, find a way of coloring the vertices of a graph such
that no two adjacent vertices are colored using same color. The other graph coloring problems like Edge Coloring (No vertex is incident to two
edges of same color) and Face Coloring (Geographical Map Coloring) can be transformed into vertex coloring.
Chromatic Number: The smallest number of colors needed to color a graph G is called its chromatic number. For example, the following can be
colored minimum 3 colors.

The problem to find chromatic number of a given graph is NP Complete.

Applications of Graph Coloring:

The graph coloring problem has huge number of applications.
1) Making Schedule or Time Table: Suppose we want to make am exam schedule for a university. We have list different subjects and students
enrolled in every subject. Many subjects would have common students (of same batch, some backlog students, etc). How do we schedule the
exam so that no two exams with a common student are scheduled at same time? How many minimum time slots are needed to schedule
all exams? This problem can be represented as a graph where every vertex is a subject and an edge between two vertices mean there is a
common student. So this is a graph coloring problem where minimum number of time slots is equal to the chromatic number of the graph.
2) Mobile Radio Frequency Assignment: When frequencies are assigned to towers, frequencies assigned to all towers at the same location
must be different. How to assign frequencies with this constraint? What is the minimum number of frequencies needed? This problem is also an
instance of graph coloring problem where every tower represents a vertex and an edge between two towers represents that they are in range of
each other.
3) Suduku: Suduku is also a variation of Graph coloring problem where every cell represents a vertex. There is an edge between two vertices if
they are in same row or same column or same block.
4) Register Allocation: In compiler optimization, register allocation is the process of assigning a large number of target program variables onto a
small number of CPU registers. This problem is also a graph coloring problem.
5) Bipartite Graphs: We can check if a graph is Bipartite or not by colowing the graph using two colors. If a given graph is 2-colorable, then it is
Bipartite, otherwise not. See this for more details.
6) Map Coloring: Geographical maps of countries or states where no two adjacent cities cannot be assigned same color. Four colors are
sufficient to color any map (See Four Color Theorem)
There can be many more applications: For example the below reference video lecture has a case study at 1:18.
Akamai runs a network of thousands of servers and the servers are used to distribute content on Internet. They install a new software or update
existing softwares pretty much every week. The update cannot be deployed on every server at the same time, because the server may have to be
taken down for the install. Also, the update should not be done one at a time, because it will take a lot of time. There are sets of servers that
cannot be taken down together, because they have certain critical functions. This is a typical scheduling application of graph coloring problem. It
turned out that 8 colors were good enough to color the graph of 75000 nodes. So they could install updates in 8 passes.
We will soon be discussing different ways to solve the graph coloring problem.
References:
Lec 6 | MIT 6.042J Mathematics for Computer Science, Fall 2010 | Video Lecture

Graph Coloring | Set 2 (Greedy Algorithm)

We introduced graph coloring and applications in previous post. As discussed in the previous post, graph coloring is widely used. Unfortunately,
there is no efficient algorithm available for coloring a graph with minimum number of colors as the problem is a known NP Complete problem.
There are approximate algorithms to solve the problem though. Following is the basic Greedy Algorithm to assign colors. It doesnt guarantee to
use minimum colors, but it guarantees an upper bound on the number of colors. The basic algorithm never uses more than d+1 colors where d is
the maximum degree of a vertex in the given graph.
Basic Greedy Coloring Algorithm:
1. Color first vertex with first color.
2. Do following for remaining V-1 vertices.
a) Consider the currently picked vertex and color it with the
lowest numbered color that has not been used on any previously
colored vertices adjacent to it. If all previously used colors
appear on vertices adjacent to v, assign a new color to it.

Following are C++ and Java implementations of the above Greedy Algorithm.

C++
// A C++ program to implement greedy algorithm for graph coloring
#include <iostream>
#include <list>
using namespace std;
// A class that represents an undirected graph
class Graph
{
int V;
// No. of vertices
list<int> *adj;
// A dynamic array of adjacency lists
public:
// Constructor and destructor
Graph(int V) { this->V = V; adj = new list<int>[V]; }
~Graph()
{ delete [] adj; }
// function to add an edge to graph
void addEdge(int v, int w);
// Prints greedy coloring of the vertices
void greedyColoring();
};
void Graph::addEdge(int v, int w)
{
adj[v].push_back(w);
adj[w].push_back(v); // Note: the graph is undirected
}
// Assigns colors (starting from 0) to all vertices and prints
// the assignment of colors
void Graph::greedyColoring()
{
int result[V];
// Assign the first color to first vertex
result[0] = 0;
// Initialize remaining V-1 vertices as unassigned
for (int u = 1; u < V; u++)
result[u] = -1; // no color is assigned to u
// A temporary array to store the available colors. True
// value of available[cr] would mean that the color cr is
// assigned to one of its adjacent vertices
bool available[V];
for (int cr = 0; cr < V; cr++)
available[cr] = false;
// Assign colors to remaining V-1 vertices
for (int u = 1; u < V; u++)
{
// Process all adjacent vertices and flag their colors
// as unavailable
list<int>::iterator i;
for (i = adj[u].begin(); i != adj[u].end(); ++i)
if (result[*i] != -1)
available[result[*i]] = true;

// Find the first available color

int cr;
for (cr = 0; cr < V; cr++)
if (available[cr] == false)
break;
result[u] = cr; // Assign the found color
// Reset the values back to false for the next iteration
for (i = adj[u].begin(); i != adj[u].end(); ++i)
if (result[*i] != -1)
available[result[*i]] = false;
}
// print the result
for (int u = 0; u < V; u++)
cout << "Vertex " << u << " ---> Color "
<< result[u] << endl;
}
// Driver program to test above function
int main()
{
Graph g1(5);
g1.addEdge(0, 1);
g1.addEdge(0, 2);
g1.addEdge(1, 2);
g1.addEdge(1, 3);
g1.addEdge(2, 3);
g1.addEdge(3, 4);
cout << "Coloring of graph 1 \n";
g1.greedyColoring();
Graph g2(5);
g2.addEdge(0, 1);
g2.addEdge(0, 2);
g2.addEdge(1, 2);
g2.addEdge(1, 4);
g2.addEdge(2, 4);
g2.addEdge(4, 3);
cout << "\nColoring of graph 2 \n";
g2.greedyColoring();
return 0;
}

Java
// A Java program to implement greedy algorithm for graph coloring
import java.io.*;
import java.util.*;
import java.util.LinkedList;
// This class represents an undirected graph using adjacency list
class Graph
{
private int V; // No. of vertices
private LinkedList<Integer> adj[]; //Adjacency List
//Constructor
Graph(int v)
{
V = v;
adj = new LinkedList[v];
for (int i=0; i<v; ++i)
adj[i] = new LinkedList();
}
//Function to add an edge into the graph
void addEdge(int v,int w)
{
adj[v].add(w);
adj[w].add(v); //Graph is undirected
}
// Assigns colors (starting from 0) to all vertices and
// prints the assignment of colors
void greedyColoring()
{

int result[] = new int[V];

// Assign the first color to first vertex
result[0] = 0;
// Initialize remaining V-1 vertices as unassigned
for (int u = 1; u < V; u++)
result[u] = -1; // no color is assigned to u
// A temporary array to store the available colors. True
// value of available[cr] would mean that the color cr is
// assigned to one of its adjacent vertices
boolean available[] = new boolean[V];
for (int cr = 0; cr < V; cr++)
available[cr] = false;
// Assign colors to remaining V-1 vertices
for (int u = 1; u < V; u++)
{
// Process all adjacent vertices and flag their colors
// as unavailable
Iterator<Integer> it = adj[u].iterator() ;
while (it.hasNext())
{
int i = it.next();
if (result[i] != -1)
available[result[i]] = true;
}
// Find the first available color
int cr;
for (cr = 0; cr < V; cr++)
if (available[cr] == false)
break;
result[u] = cr; // Assign the found color
// Reset the values back to false for the next iteration
it = adj[u].iterator() ;
while (it.hasNext())
{
int i = it.next();
if (result[i] != -1)
available[result[i]] = false;
}
}
// print the result
for (int u = 0; u < V; u++)
System.out.println("Vertex " + u + " ---> Color "
+ result[u]);
}
// Driver method
public static void main(String args[])
{
Graph g1 = new Graph(5);
g1.addEdge(0, 1);
g1.addEdge(0, 2);
g1.addEdge(1, 2);
g1.addEdge(1, 3);
g1.addEdge(2, 3);
g1.addEdge(3, 4);
System.out.println("Coloring of graph 1");
g1.greedyColoring();
System.out.println();
Graph g2 = new Graph(5);
g2.addEdge(0, 1);
g2.addEdge(0, 2);
g2.addEdge(1, 2);
g2.addEdge(1, 4);
g2.addEdge(2, 4);
g2.addEdge(4, 3);
System.out.println("Coloring of graph 2 ");
g2.greedyColoring();
}
}
// This code is contributed by Aakash Hasija

Coloring
Vertex 0
Vertex 1
Vertex 2
Vertex 3
Vertex 4

of graph 1
---> Color
---> Color
---> Color
---> Color
---> Color

0
1
2
0
1

Coloring
Vertex 0
Vertex 1
Vertex 2
Vertex 3
Vertex 4

of graph 2
---> Color
---> Color
---> Color
---> Color
---> Color

0
1
2
0
3

Time Complexity: O(V^2 + E) in worst case.

Analysis of Basic Algorithm
The above algorithm doesnt always use minimum number of colors. Also, the number of colors used sometime depend on the order in which
vertices are processed. For example, consider the following two graphs. Note that in graph on right side, vertices 3 and 4 are swapped. If we
consider the vertices 0, 1, 2, 3, 4 in left graph, we can color the graph using 3 colors. But if we consider the vertices 0, 1, 2, 3, 4 in right graph, we
need 4 colors.

So the order in which the vertices are picked is important. Many people have suggested different ways to find an ordering that work better than the
basic algorithm on average. The most common is WelshPowell Algorithm which considers vertices in descending order of degrees.
How does the basic algorithm guarantee an upper bound of d+1?
Here d is the maximum degree in the given graph. Since d is maximum degree, a vertex cannot be attached to more than d vertices. When we
color a vertex, at most d colors could have already been used by its adjacent. To color this vertex, we need to pick the smallest numbered color
that is not used by the adjacent vertices. If colors are numbered like 1, 2, ., then the value of such smallest number must be between 1 to d+1
(Note that d numbers are already picked by adjacent vertices).
This can also be proved using induction. See this video lecture for proof.
We will soon be discussing some interesting facts about chromatic number and graph coloring.

Travelling Salesman Problem | Set 1 (Naive and Dynamic Programming)

Travelling Salesman Problem (TSP): Given a set of cities and distance between every pair of cities, the problem is to find the shortest possible
route that visits every city exactly once and returns to the starting point.
Note the difference between Hamiltonian Cycle and TSP. The Hamiltoninan cycle problem is to find if there exist a tour that visits every city
exactly once. Here we know that Hamiltonian Tour exists (because the graph is complete) and in fact many such tours exist, the problem is to find
a minimum weight Hamiltonian Cycle.

For example, consider the graph shown in figure on right side. A TSP tour in the graph is 1-2-4-3-1. The cost of the tour is 10+25+30+15 which
is 80.
The problem is a famous NP hard problem. There is no polynomial time know solution for this problem.
Following are different solutions for the traveling salesman problem.
Naive Solution:
1) Consider city 1 as the starting and ending point.
2) Generate all (n-1)! Permutations of cities.
3) Calculate cost of every permutation and keep track of minimum cost permutation.
4) Return the permutation with minimum cost.
Time Complexity: ?(n!)
Dynamic Programming:
Let the given set of vertices be {1, 2, 3, 4,.n}. Let us consider 1 as starting and ending point of output. For every other vertex i (other than 1), we
find the minimum cost path with 1 as the starting point, i as the ending point and all vertices appearing exactly once. Let the cost of this path be
cost(i), the cost of corresponding Cycle would be cost(i) + dist(i, 1) where dist(i, 1) is the distance from i to 1. Finally, we return the minimum of
all [cost(i) + dist(i, 1)] values. This looks simple so far. Now the question is how to get cost(i)?
To calculate cost(i) using Dynamic Programming, we need to have some recursive relation in terms of sub-problems. Let us define a term C(S, i)
be the cost of the minimum cost path visiting each vertex in set S exactly once, starting at 1 and ending at i.
We start with all subsets of size 2 and calculate C(S, i) for all subsets where S is the subset, then we calculate C(S, i) for all subsets S of size 3 and
so on. Note that 1 must be present in every subset.
If size of S is 2, then S must be {1, i},
C(S, i) = dist(1, i)
Else if size of S is greater than 2.
C(S, i) = min { C(S-{i}, j) + dis(j, i)} where j belongs to S, j != i and j != 1.

For a set of size n, we consider n-2 subsets each of size n-1 such that all subsets dont have nth in them.
Using the above recurrence relation, we can write dynamic programming based solution. There are at most O(n*2n) subproblems, and each one
takes linear time to solve. The total running time is therefore O(n2*2n). The time complexity is much less than O(n!), but still exponential. Space
required is also exponential. So this approach is also infeasible even for slightly higher number of vertices.
We will soon be discussing approximate algorithms for travelling salesman problem.
References:
http://www.lsi.upc.edu/~mjserna/docencia/algofib/P07/dynprog.pdf
http://www.cs.berkeley.edu/~vazirani/algorithms/chap6.pdf

Travelling Salesman Problem | Set 2 (Approximate using MST)

We introduced Travelling Salesman Problem and discussed Naive and Dynamic Programming Solutions for the problem in the previous post,.
Both of the solutions are infeasible. In fact, there is no polynomial time solution available for this problem as the problem is a known NP-Hard
problem. There are approximate algorithms to solve the problem though. The approximate algorithms work only if the problem instance satisfies
Triangle-Inequality.
Triangle-Inequality: The least distant path to reach a vertex j from i is always to reach j directly from i, rather than through some other vertex k
(or vertices), i.e., dis(i, j) is always less than or equal to dis(i, k) + dist(k, j). The Triangle-Inequality holds in many practical situations.
When the cost function satisfies the triangle inequality, we can design an approximate algorithm for TSP that returns a tour whose cost is never
more than twice the cost of an optimal tour. The idea is to use Minimum Spanning Tree (MST). Following is the MST based algorithm.
Algorithm:
1) Let 1 be the starting and ending point for salesman.
2) Construct MST from with 1 as root using Prims Algorithm.
3) List vertices visited in preorder walk of the constructed MST and add 1 at the end.
Let us consider the following example. The first diagram is the given graph. The second diagram shows MST constructed with 1 as root. The
preorder traversal of MST is 1-2-4-3. Adding 1 at the end gives 1-2-4-3-1 which is the output of this algorithm.

In this case, the approximate algorithm produces the optimal tour, but it may not produce optimal tour in all cases.
How is this algorithm 2-approximate? The cost of the output produced by the above algorithm is never more than twice the cost of best
possible output. Let us see how is this guaranteed by the above algorithm.
Let us define a term full walk to understand this. A full walk is lists all vertices when they are first visited in preorder, it also list vertices when they
are returned after a subtree is visited in preorder. The full walk of above tree would be 1-2-1-4-1-3-1.
Following are some important facts that prove the 2-approximateness.
1) The cost of best possible Travelling Salesman tour is never less than the cost of MST. (The definition of MST says, it is a minimum cost tree that
connects all vertices).
2) The total cost of full walk is at most twice the cost of MST (Every edge of MST is visited at-most twice)
3) The output of the above algorithm is less than the cost of full walk. In above algorithm, we print preorder walk as output. In prreorder walk,
two or more edges of full walk are replaced with a single edge. For example, 2-1 and 1-4 are replaced by 1 edge 2-4. So if the graph follows
triangle inequality, then this is always true.
From the above three statements, we can conclude that the cost of output produced by the approximate algorithm is never more than twice the
cost of best possible solution.
We have discussed a very simple 2-approximate algorithm for the travelling salesman problem. There are other better approximate algorithms for
the problem. For example Christofides algorithm is 1.5 approximate algorithm. We will soon be discussing these algorithms as separate posts.
References:
Introduction to Algorithms 3rd Edition by Clifford Stein, Thomas H. Cormen, Charles E. Leiserson, Ronald L. Rivest
http://www.personal.kent.edu/~rmuhamma/Algorithms/MyAlgorithms/AproxAlgor/TSP/tsp.htm

Backtracking | Set 6 (Hamiltonian Cycle)

Hamiltonian Path in an undirected graph is a path that visits each vertex exactly once. A Hamiltonian cycle (or Hamiltonian circuit) is a Hamiltonian
Path such that there is an edge (in graph) from the last vertex to the first vertex of the Hamiltonian Path. Determine whether a given graph contains
Hamiltonian Cycle or not. If it contains, then print the path. Following are the input and output of the required function.
Input:
A 2D array graph[V][V] where V is the number of vertices in graph and graph[V][V] is adjacency matrix representation of the graph. A value
graph[i][j] is 1 if there is a direct edge from i to j, otherwise graph[i][j] is 0.
Output:
An array path[V] that should contain the Hamiltonian Path. path[i] should represent the ith vertex in the Hamiltonian Path. The code should also
return false if there is no Hamiltonian Cycle in the graph.
For example, a Hamiltonian Cycle in the following graph is {0, 1, 2, 4, 3, 0}. There are more Hamiltonian Cycles in the graph like {0, 3, 4, 2, 1,
0}
(0)--(1)--(2)
| / \ |
| / \ |
| /
\ |
(3)-------(4)

And the following graph doesnt contain any Hamiltonian Cycle.

(0)--(1)--(2)
| / \ |
| / \ |
| /
\ |
(3)
(4)

Naive Algorithm
Generate all possible configurations of vertices and print a configuration that satisfies the given constraints. There will be n! (n factorial)
configurations.
while there are untried conflagrations
{
generate the next configuration
if ( there are edges between two consecutive vertices of this
configuration and there is an edge from the last vertex to
the first ).
{
print this configuration;
break;
}
}

Backtracking Algorithm
Create an empty path array and add vertex 0 to it. Add other vertices, starting from the vertex 1. Before adding a vertex, check for whether it is
adjacent to the previously added vertex and not already added. If we find such a vertex, we add the vertex as part of the solution. If we do not
find a vertex then we return false.
Implementation of Backtracking solution
Following are implementations of the Backtracking solution.

C/C++
/* C/C++ program for solution of Hamiltonian Cycle problem
using backtracking */
#include<stdio.h>
// Number of vertices in the graph
#define V 5
void printSolution(int path[]);
/* A utility function to check if the vertex v can be added at
index 'pos' in the Hamiltonian Cycle constructed so far (stored
in 'path[]') */
bool isSafe(int v, bool graph[V][V], int path[], int pos)
{
/* Check if this vertex is an adjacent vertex of the previously
added vertex. */
if (graph [ path[pos-1] ][ v ] == 0)

return false;
/* Check if the vertex has already been included.
This step can be optimized by creating an array of size V */
for (int i = 0; i < pos; i++)
if (path[i] == v)
return false;
return true;
}
/* A recursive utility function to solve hamiltonian cycle problem */
bool hamCycleUtil(bool graph[V][V], int path[], int pos)
{
/* base case: If all vertices are included in Hamiltonian Cycle */
if (pos == V)
{
// And if there is an edge from the last included vertex to the
// first vertex
if ( graph[ path[pos-1] ][ path[0] ] == 1 )
return true;
else
return false;
}
// Try different vertices as a next candidate in Hamiltonian Cycle.
// We don't try for 0 as we included 0 as starting point in in hamCycle()
for (int v = 1; v < V; v++)
{
/* Check if this vertex can be added to Hamiltonian Cycle */
if (isSafe(v, graph, path, pos))
{
path[pos] = v;
/* recur to construct rest of the path */
if (hamCycleUtil (graph, path, pos+1) == true)
return true;
/* If adding vertex v doesn't lead to a solution,
then remove it */
path[pos] = -1;
}
}
/* If no vertex can be added to Hamiltonian Cycle constructed so far,
then return false */
return false;
}
/* This function solves the Hamiltonian Cycle problem using Backtracking.
It mainly uses hamCycleUtil() to solve the problem. It returns false
if there is no Hamiltonian Cycle possible, otherwise return true and
prints the path. Please note that there may be more than one solutions,
this function prints one of the feasible solutions. */
bool hamCycle(bool graph[V][V])
{
int *path = new int[V];
for (int i = 0; i < V; i++)
path[i] = -1;
/* Let us put vertex 0 as the first vertex in the path. If there is
a Hamiltonian Cycle, then the path can be started from any point
of the cycle as the graph is undirected */
path[0] = 0;
if ( hamCycleUtil(graph, path, 1) == false )
{
printf("\nSolution does not exist");
return false;
}
printSolution(path);
return true;
}
/* A utility function to print solution */
void printSolution(int path[])
{
printf ("Solution Exists:"
" Following is one Hamiltonian Cycle \n");
for (int i = 0; i < V; i++)
printf(" %d ", path[i]);

// Let us print the first vertex again to show the complete cycle
printf(" %d ", path[0]);
printf("\n");
}
// driver program to test above function
int main()
{
/* Let us create the following graph
(0)--(1)--(2)
| / \ |
| / \ |
| /
\ |
(3)-------(4)
*/
bool graph1[V][V] = {{0, 1, 0, 1, 0},
{1, 0, 1, 1, 1},
{0, 1, 0, 0, 1},
{1, 1, 0, 0, 1},
{0, 1, 1, 1, 0},
};
// Print the solution
hamCycle(graph1);
/* Let us create the following graph
(0)--(1)--(2)
| / \ |
| / \ |
| /
\ |
(3)
(4)
*/
bool graph2[V][V] = {{0, 1, 0, 1, 0},
{1, 0, 1, 1, 1},
{0, 1, 0, 0, 1},
{1, 1, 0, 0, 0},
{0, 1, 1, 0, 0},
};
// Print the solution
hamCycle(graph2);
return 0;
}

Java
/* Java program for solution of Hamiltonian Cycle problem
using backtracking */
class HamiltonianCycle
{
final int V = 5;
int path[];
/* A utility function to check if the vertex v can be
added at index 'pos'in the Hamiltonian Cycle
constructed so far (stored in 'path[]') */
boolean isSafe(int v, int graph[][], int path[], int pos)
{
/* Check if this vertex is an adjacent vertex of
the previously added vertex. */
if (graph[path[pos - 1]][v] == 0)
return false;
/* Check if the vertex has already been included.
This step can be optimized by creating an array
of size V */
for (int i = 0; i < pos; i++)
if (path[i] == v)
return false;
return true;
}
/* A recursive utility function to solve hamiltonian
cycle problem */
boolean hamCycleUtil(int graph[][], int path[], int pos)
{
/* base case: If all vertices are included in
Hamiltonian Cycle */
if (pos == V)

{
// And if there is an edge from the last included
// vertex to the first vertex
if (graph[path[pos - 1]][path[0]] == 1)
return true;
else
return false;
}
// Try different vertices as a next candidate in
// Hamiltonian Cycle. We don't try for 0 as we
// included 0 as starting point in in hamCycle()
for (int v = 1; v < V; v++)
{
/* Check if this vertex can be added to Hamiltonian
Cycle */
if (isSafe(v, graph, path, pos))
{
path[pos] = v;
/* recur to construct rest of the path */
if (hamCycleUtil(graph, path, pos + 1) == true)
return true;
/* If adding vertex v doesn't lead to a solution,
then remove it */
path[pos] = -1;
}
}
/* If no vertex can be added to Hamiltonian Cycle
constructed so far, then return false */
return false;
}
/* This function solves the Hamiltonian Cycle problem using
Backtracking. It mainly uses hamCycleUtil() to solve the
problem. It returns false if there is no Hamiltonian Cycle
possible, otherwise return true and prints the path.
Please note that there may be more than one solutions,
this function prints one of the feasible solutions. */
int hamCycle(int graph[][])
{
path = new int[V];
for (int i = 0; i < V; i++)
path[i] = -1;
/* Let us put vertex 0 as the first vertex in the path.
If there is a Hamiltonian Cycle, then the path can be
started from any point of the cycle as the graph is
undirected */
path[0] = 0;
if (hamCycleUtil(graph, path, 1) == false)
{
System.out.println("\nSolution does not exist");
return 0;
}
printSolution(path);
return 1;
}
/* A utility function to print solution */
void printSolution(int path[])
{
System.out.println("Solution Exists: Following" +
" is one Hamiltonian Cycle");
for (int i = 0; i < V; i++)
System.out.print(" " + path[i] + " ");
// Let us print the first vertex again to show the
// complete cycle
System.out.println(" " + path[0] + " ");
}
// driver program to test above function
public static void main(String args[])
{
HamiltonianCycle hamiltonian =
new HamiltonianCycle();
/* Let us create the following graph

(0)--(1)--(2)
| / \ |
| / \ |
| /
\ |
(3)-------(4)
int graph1[][] =
{1, 0, 1, 1,
{0, 1, 0, 0,
{1, 1, 0, 0,
{0, 1, 1, 1,
};

*/
{{0, 1, 0, 1, 0},
1},
1},
1},
0},

// Print the solution

hamiltonian.hamCycle(graph1);
/* Let us create
(0)--(1)--(2)
| / \ |
| / \ |
| /
\ |
(3)
(4)
int graph2[][] =
{1, 0, 1, 1,
{0, 1, 0, 0,
{1, 1, 0, 0,
{0, 1, 1, 0,
};

the following graph

*/
{{0, 1, 0, 1, 0},
1},
1},
0},
0},

// Print the solution

hamiltonian.hamCycle(graph2);
}
}
// This code is contributed by Abhishek Shankhadhar

Solution Exists: Following is one Hamiltonian Cycle

0 1 2 4 3 0
Solution does not exist

Vertex Cover Problem | Set 1 (Introduction and Approximate Algorithm)

A vertex cover of an undirected graph is a subset of its vertices such that for every edge (u, v) of the graph, either u or v is in vertex cover.
Although the name is Vertex Cover, the set covers all edges of the given graph. Given an undirected graph, the vertex cover problem is to
find minimum size vertex cover.
Following are some examples.

Vertex Cover Problem is a known NP Complete problem, i.e., there is no polynomial time solution for this unless P = NP. There are approximate
polynomial time algorithms to solve the problem though. Following is a simple approximate algorithm adapted from CLRS book.
Approximate Algorithm for Vertex Cover:
1) Initialize the result as {}
2) Consider a set of all edges in given graph. Let the set be E.
3) Do following while E is not empty
...a) Pick an arbitrary edge (u, v) from set E and add 'u' and 'v' to result
...b) Remove all edges from E which are either incident on u or v.
4) Return result

Following diagram taken from CLRS book shows execution of above approximate algorithm.

How well the above algorithm perform?

It can be proved that the above approximate algorithm never finds a vertex cover whose size is more than twice the size of minimum possible
vertex cover (Refer this for proof)
Implementation:
Following are C++ and Java implementations of above approximate algorithm.

C++
// Program to print Vertex Cover of a given undirected graph
#include<iostream>

#include <list>
using namespace std;
// This class represents a undirected graph using adjacency list
class Graph
{
int V;
// No. of vertices
list<int> *adj; // Pointer to an array containing adjacency lists
public:
Graph(int V); // Constructor
void addEdge(int v, int w); // function to add an edge to graph
void printVertexCover(); // prints vertex cover
};
Graph::Graph(int V)
{
this->V = V;
adj = new list<int>[V];
}
void Graph::addEdge(int v, int w)
{
adj[v].push_back(w); // Add w to vs list.
adj[w].push_back(v); // Since the graph is undirected
}
// The function to print vertex cover
void Graph::printVertexCover()
{
// Initialize all vertices as not visited.
bool visited[V];
for (int i=0; i<V; i++)
visited[i] = false;
list<int>::iterator i;
// Consider all edges one by one
for (int u=0; u<V; u++)
{
// An edge is only picked when both visited[u] and visited[v]
// are false
if (visited[u] == false)
{
// Go through all adjacents of u and pick the first not
// yet visited vertex (We are basically picking an edge
// (u, v) from remaining edges.
for (i= adj[u].begin(); i != adj[u].end(); ++i)
{
int v = *i;
if (visited[v] == false)
{
// Add the vertices (u, v) to the result set.
// We make the vertex u and v visited so that
// all edges from/to them would be ignored
visited[v] = true;
visited[u] = true;
break;
}
}
}
}
// Print the vertex cover
for (int i=0; i<V; i++)
if (visited[i])
cout << i << " ";
}
// Driver program to test methods of graph class
int main()
{
// Create a graph given in the above diagram
Graph g(7);
g.addEdge(0, 1);
g.addEdge(0, 2);
g.addEdge(1, 3);
g.addEdge(3, 4);
g.addEdge(4, 5);
g.addEdge(5, 6);
g.printVertexCover();

return 0;
}

Java
// Java Program to print Vertex Cover of a given undirected graph
import java.io.*;
import java.util.*;
import java.util.LinkedList;
// This class represents an undirected graph using adjacency list
class Graph
{
private int V; // No. of vertices
// Array of lists for Adjacency List Representation
private LinkedList<Integer> adj[];
// Constructor
Graph(int v)
{
V = v;
adj = new LinkedList[v];
for (int i=0; i<v; ++i)
adj[i] = new LinkedList();
}
//Function to add an edge into the graph
void addEdge(int v, int w)
{
adj[v].add(w); // Add w to v's list.
adj[w].add(v); //Graph is undirected
}
// The function to print vertex cover
void printVertexCover()
{
// Initialize all vertices as not visited.
boolean visited[] = new boolean[V];
for (int i=0; i<V; i++)
visited[i] = false;
Iterator<Integer> i;
// Consider all edges one by one
for (int u=0; u<V; u++)
{
// An edge is only picked when both visited[u]
// and visited[v] are false
if (visited[u] == false)
{
// Go through all adjacents of u and pick the
// first not yet visited vertex (We are basically
// picking an edge (u, v) from remaining edges.
i = adj[u].iterator();
while (i.hasNext())
{
int v = i.next();
if (visited[v] == false)
{
// Add the vertices (u, v) to the result
// set. We make the vertex u and v visited
// so that all edges from/to them would
// be ignored
visited[v] = true;
visited[u] = true;
break;
}
}
}
}
// Print the vertex cover
for (int j=0; j<V; j++)
if (visited[j])
System.out.print(j+" ");
}
// Driver method

public static void main(String args[])

{
// Create a graph given in the above diagram
Graph g = new Graph(7);
g.addEdge(0, 1);
g.addEdge(0, 2);
g.addEdge(1, 3);
g.addEdge(3, 4);
g.addEdge(4, 5);
g.addEdge(5, 6);
g.printVertexCover();
}
}
// This code is contributed by Aakash Hasija

0 1 3 4 5 6

Time Complexity of above algorithm is O(V + E).

Exact Algorithms:
Although the problem is NP complete, it can be solved in polynomial time for following types of graphs.
1) Bipartite Graph
2) Tree Graph
The problem to check whether there is a vertex cover of size smaller than or equal to a given number k can also be solved in polynomial time if k is
bounded by O(LogV) (Refer this)
We will soon be discussing exact algorithms for vertex cover.

K Centers Problem | Set 1 (Greedy Approximate Algorithm)

Given n cities and distances between every pair of cities, select k cities to place warehouses (or ATMs) such that the maximum distance of a city
to a warehouse (or ATM) is minimized.
For example consider the following four cities, 0, 1, 2 and 3 and distances between them, how do place 2 ATMs among these 4 cities so that the
maximum distance of a city to an ATM is minimized.

There is no polynomial time solution available for this problem as the problem is a known NP-Hard problem. There is a polynomial time Greedy
approximate algorithm, the greedy algorithm provides a solution which is never worse that twice the optimal solution. The greedy solution works
only if the distances between cities follow Triangular Inequality (Distance between two points is always smaller than sum of distances through a
third point).
The 2-Approximate Greedy Algorithm:
1) Choose the first center arbitrarily.
2) Choose remaining k-1 centers using the following criteria.
Let c1, c2, c3, ci be the already chosen centers. Choose
(i+1)th center by picking the city which is farthest from already
selected centers, i.e, the point p which has following value as maximum
Min[dist(p, c1), dist(p, c2), dist(p, c3), . dist(p, ci)]
The following diagram taken from here illustrates above algorithm.

Example (k = 3 in the above shown Graph)

a) Let the first arbitrarily picked vertex be 0.
b) The next vertex is 1 because 1 is the farthest vertex from 0.
c) Remaining cities are 2 and 3. Calculate their distances from already selected centers (0 and 1). The greedy algorithm basically calculates
following values.
Minimum of all distanced from 2 to already considered centers
Min[dist(2, 0), dist(2, 1)] = Min[7, 8] = 7
Minimum of all distanced from 3 to already considered centers
Min[dist(3, 0), dist(3, 1)] = Min[6, 5] = 5
After computing the above values, the city 2 is picked as the value corresponding to 2 is maximum.
Note that the greedy algorithm doesnt give best solution for k = 2 as this is just an approximate algorithm with bound as twice of optimal.
Proof that the above greedy algorithm is 2 approximate.
Let OPT be the maximum distance of a city from a center in the Optimal solution. We need to show that the maximum distance obtained from
Greedy algorithm is 2*OPT.
The proof can be done using contradiction.
a) Assume that the distance from the furthest point to all centers is > 2OPT.

b) This means that distances between all centers are also > 2OPT.
c) We have k + 1 points with distances > 2OPT between every pair.
d) Each point has a center of the optimal solution with distance ? OPT to it.
e) There exists a pair of points with the same center X in the optimal solution (pigeonhole principle: k optimal centers, k+1 points)
f) The distance between them is at most 2OPT (triangle inequality) which is a contradiction.
Source:
http://algo2.iti.kit.edu/vanstee/courses/kcenter.pdf

Ford-Fulkerson Algorithm for Maximum Flow Problem

Given a graph which represents a flow network where every edge has a capacity. Also given two vertices source s and sink t in the graph, find the
maximum possible flow from s to t with following constraints:
a) Flow on an edge doesnt exceed the given capacity of the edge.
b) Incoming flow is equal to outgoing flow for every vertex except s and t.
For example, consider the following graph from CLRS book.

The maximum possible flow in the above graph is 23.

Ford-Fulkerson Algorithm
The following is simple idea of Ford-Fulkerson algorithm:
1) Start with initial flow as 0.
2) While there is a augmenting path from source to sink.
Add this path-flow to flow.
3) Return flow.

Time Complexity: Time complexity of the above algorithm is O(max_flow * E). We run a loop while there is an augmenting path. In worst case,
we may add 1 unit flow in every iteration. Therefore the time complexity becomes O(max_flow * E).
How to implement the above simple algorithm?
Let us first define the concept of Residual Graph which is needed for understanding the implementation.
Residual Graph of a flow network is a graph which indicates additional possible flow. If there is a path from source to sink in residual graph, then
it is possible to add flow. Every edge of a residual graph has a value called residual capacity which is equal to original capacity of the edge minus
current flow. Residual capacity is basically the current capacity of the edge.
Let us now talk about implementation details. Residual capacity is 0 if there is no edge between to vertices of residual graph. We can initialize the
residual graph as original graph as there is no initial flow and initially residual capacity is equal to original capacity. To find an augmenting path, we
can either do a BFS or DFS of the residual graph. We have used BFS in below implementation. Using BFS, we can find out if there is a path from
source to sink. BFS also builds parent[] array. Using the parent[] array, we traverse through the found path and find possible flow through this path
by finding minimum residual capacity along the path. We later add the found path flow to overall flow.
The important thing is, we need to update residual capacities in the residual graph. We subtract path flow from all edges along the path and we add
path flow along the reverse edges We need to add path flow along reverse edges because may later need to send flow in reverse direction (See
following video for example).
http://www.youtube.com/watch?v=-8MwfgB-lyM
Following are C++ and Java implementations of Ford-Fulkerson algorithm. To keep things simple, graph is represented as a 2D matrix.

C++
// C++ program for implementation of Ford Fulkerson algorithm
#include <iostream>
#include <limits.h>
#include <string.h>
#include <queue>
using namespace std;
// Number of vertices in given graph

#define V 6
/* Returns true if there is a path from source 's' to sink 't' in
residual graph. Also fills parent[] to store the path */
bool bfs(int rGraph[V][V], int s, int t, int parent[])
{
// Create a visited array and mark all vertices as not visited
bool visited[V];
memset(visited, 0, sizeof(visited));
// Create a queue, enqueue source vertex and mark source vertex
// as visited
queue <int> q;
q.push(s);
visited[s] = true;
parent[s] = -1;
// Standard BFS Loop
while (!q.empty())
{
int u = q.front();
q.pop();
for (int v=0; v<V; v++)
{
if (visited[v]==false && rGraph[u][v] > 0)
{
q.push(v);
parent[v] = u;
visited[v] = true;
}
}
}
// If we reached sink in BFS starting from source, then return
// true, else false
return (visited[t] == true);
}
// Returns tne maximum flow from s to t in the given graph
int fordFulkerson(int graph[V][V], int s, int t)
{
int u, v;
// Create a residual graph and fill the residual graph with
// given capacities in the original graph as residual capacities
// in residual graph
int rGraph[V][V]; // Residual graph where rGraph[i][j] indicates
// residual capacity of edge from i to j (if there
// is an edge. If rGraph[i][j] is 0, then there is not)
for (u = 0; u < V; u++)
for (v = 0; v < V; v++)
rGraph[u][v] = graph[u][v];
int parent[V]; // This array is filled by BFS and to store path
int max_flow = 0; // There is no flow initially
// Augment the flow while tere is path from source to sink
while (bfs(rGraph, s, t, parent))
{
// Find minimum residual capacity of the edhes along the
// path filled by BFS. Or we can say find the maximum flow
// through the path found.
int path_flow = INT_MAX;
for (v=t; v!=s; v=parent[v])
{
u = parent[v];
path_flow = min(path_flow, rGraph[u][v]);
}
// update residual capacities of the edges and reverse edges
// along the path
for (v=t; v != s; v=parent[v])
{
u = parent[v];
rGraph[u][v] -= path_flow;
rGraph[v][u] += path_flow;
}
// Add path flow to overall flow

max_flow += path_flow;
}
// Return the overall flow
return max_flow;
}
// Driver program to test above functions
int main()
{
// Let us create a graph shown in the above example
int graph[V][V] = { {0, 16, 13, 0, 0, 0},
{0, 0, 10, 12, 0, 0},
{0, 4, 0, 0, 14, 0},
{0, 0, 9, 0, 0, 20},
{0, 0, 0, 7, 0, 4},
{0, 0, 0, 0, 0, 0}
};
cout << "The maximum possible flow is " << fordFulkerson(graph, 0, 5);
return 0;
}

Java
// Java program for implementation of Ford Fulkerson algorithm
import java.util.*;
import java.lang.*;
import java.io.*;
import java.util.LinkedList;
class MaxFlow
{
static final int V = 6; //Number of vertices in graph
/* Returns true if there is a path from source 's' to sink
't' in residual graph. Also fills parent[] to store the
path */
boolean bfs(int rGraph[][], int s, int t, int parent[])
{
// Create a visited array and mark all vertices as not
// visited
boolean visited[] = new boolean[V];
for(int i=0; i<V; ++i)
visited[i]=false;
// Create a queue, enqueue source vertex and mark
// source vertex as visited
LinkedList<Integer> queue = new LinkedList<Integer>();
queue.add(s);
visited[s] = true;
parent[s]=-1;
// Standard BFS Loop
while (queue.size()!=0)
{
int u = queue.poll();
for (int v=0; v<V; v++)
{
if (visited[v]==false && rGraph[u][v] > 0)
{
queue.add(v);
parent[v] = u;
visited[v] = true;
}
}
}
// If we reached sink in BFS starting from source, then
// return true, else false
return (visited[t] == true);
}
// Returns tne maximum flow from s to t in the given graph
int fordFulkerson(int graph[][], int s, int t)
{
int u, v;

// Create a residual graph and fill the residual graph

// with given capacities in the original graph as
// residual capacities in residual graph
// Residual graph where rGraph[i][j] indicates
// residual capacity of edge from i to j (if there
// is an edge. If rGraph[i][j] is 0, then there is
// not)
int rGraph[][] = new int[V][V];
for (u = 0; u < V; u++)
for (v = 0; v < V; v++)
rGraph[u][v] = graph[u][v];
// This array is filled by BFS and to store path
int parent[] = new int[V];
int max_flow = 0; // There is no flow initially
// Augment the flow while tere is path from source
// to sink
while (bfs(rGraph, s, t, parent))
{
// Find minimum residual capacity of the edhes
// along the path filled by BFS. Or we can say
// find the maximum flow through the path found.
int path_flow = Integer.MAX_VALUE;
for (v=t; v!=s; v=parent[v])
{
u = parent[v];
path_flow = Math.min(path_flow, rGraph[u][v]);
}
// update residual capacities of the edges and
// reverse edges along the path
for (v=t; v != s; v=parent[v])
{
u = parent[v];
rGraph[u][v] -= path_flow;
rGraph[v][u] += path_flow;
}
// Add path flow to overall flow
max_flow += path_flow;
}
// Return the overall flow
return max_flow;
}
// Driver program to test above functions
public static void main (String[] args) throws java.lang.Exception
{
// Let us create a graph shown in the above example
int graph[][] =new int[][] { {0, 16, 13, 0, 0, 0},
{0, 0, 10, 12, 0, 0},
{0, 4, 0, 0, 14, 0},
{0, 0, 9, 0, 0, 20},
{0, 0, 0, 7, 0, 4},
{0, 0, 0, 0, 0, 0}
};
MaxFlow m = new MaxFlow();
System.out.println("The maximum possible flow is " +
m.fordFulkerson(graph, 0, 5));
}
}

The maximum possible flow is 23

The above implementation of Ford Fulkerson Algorithm is called Edmonds-Karp Algorithm. The idea of Edmonds-Karp is to use BFS in Ford
Fulkerson implementation as BFS always picks a path with minimum number of edges. When BFS is used, the worst case time complexity can be
reduced to O(VE2). The above implementation uses adjacency matrix representation though where BFS takes O(V2) time, the time complexity of
the above implementation is O(EV3) (Refer CLRS book for proof of time complexity)
This is an important problem as it arises in many practical situations. Examples include, maximizing the transportation with given traffic limits,

maximizing packet flow in computer networks.

Exercise:
Modify the above implementation so that it that runs in O(VE2) time.
References:
http://www.stanford.edu/class/cs97si/08-network-flow-problems.pdf
Introduction to Algorithms 3rd Edition by Clifford Stein, Thomas H. Cormen, Charles E. Leiserson, Ronald L. Rivest

Find maximum number of edge disjoint paths between two vertices

Given a directed graph and two vertices in it, source s and destination t, find out the maximum number of edge disjoint paths from s to t. Two paths
are said edge disjoint if they dont share any edge.

There can be maximum two edge disjoint paths from source 0 to destination 7 in the above graph. Two edge disjoint paths are highlighted below in
red and blue colors are 0-2-6-7 and 0-3-6-5-7.

Note that the paths may be different, but the maximum number is same. For example, in the above diagram, another possible set of paths is 0-1-26-7 and 0-3-6-5-7 respectively.
This problem can be solved by reducing it to maximum flow problem. Following are steps.
1) Consider the given source and destination as source and sink in flow network. Assign unit capacity to each edge.
2) Run Ford-Fulkerson algorithm to find the maximum flow from source to sink.
3) The maximum flow is equal to the maximum number of edge-disjoint paths.
When we run Ford-Fulkerson, we reduce the capacity by a unit. Therefore, the edge can not be used again. So the maximum flow is equal to the
maximum number of edge-disjoint paths.
Following is C++ implementation of the above algorithm. Most of the code is taken from here.
// C++ program to find maximum number of edge disjoint paths
#include <iostream>
#include <limits.h>
#include <string.h>
#include <queue>
using namespace std;
// Number of vertices in given graph
#define V 8
/* Returns true if there is a path from source 's' to sink 't' in
residual graph. Also fills parent[] to store the path */
bool bfs(int rGraph[V][V], int s, int t, int parent[])
{
// Create a visited array and mark all vertices as not visited
bool visited[V];
memset(visited, 0, sizeof(visited));
// Create a queue, enqueue source vertex and mark source vertex
// as visited
queue <int> q;
q.push(s);
visited[s] = true;
parent[s] = -1;
// Standard BFS Loop
while (!q.empty())
{
int u = q.front();
q.pop();
for (int v=0; v<V; v++)

{
if (visited[v]==false && rGraph[u][v] > 0)
{
q.push(v);
parent[v] = u;
visited[v] = true;
}
}
}
// If we reached sink in BFS starting from source, then return
// true, else false
return (visited[t] == true);
}
// Returns tne maximum number of edge-disjoint paths from s to t.
// This function is copy of forFulkerson() discussed at http://goo.gl/wtQ4Ks
int findDisjointPaths(int graph[V][V], int s, int t)
{
int u, v;
// Create a residual graph and fill the residual graph with
// given capacities in the original graph as residual capacities
// in residual graph
int rGraph[V][V]; // Residual graph where rGraph[i][j] indicates
// residual capacity of edge from i to j (if there
// is an edge. If rGraph[i][j] is 0, then there is not)
for (u = 0; u < V; u++)
for (v = 0; v < V; v++)
rGraph[u][v] = graph[u][v];
int parent[V]; // This array is filled by BFS and to store path
int max_flow = 0; // There is no flow initially
// Augment the flow while tere is path from source to sink
while (bfs(rGraph, s, t, parent))
{
// Find minimum residual capacity of the edges along the
// path filled by BFS. Or we can say find the maximum flow
// through the path found.
int path_flow = INT_MAX;
for (v=t; v!=s; v=parent[v])
{
u = parent[v];
path_flow = min(path_flow, rGraph[u][v]);
}
// update residual capacities of the edges and reverse edges
// along the path
for (v=t; v != s; v=parent[v])
{
u = parent[v];
rGraph[u][v] -= path_flow;
rGraph[v][u] += path_flow;
}
// Add path flow to overall flow
max_flow += path_flow;
}
// Return the overall flow (max_flow is equal to maximum
// number of edge-disjoint paths)
return max_flow;
}
// Driver program to test above functions
int main()
{
// Let us create a graph shown in the above example
int graph[V][V] = { {0, 1, 1, 1, 0, 0, 0, 0},
{0, 0, 1, 0, 0, 0, 0, 0},
{0, 0, 0, 1, 0, 0, 1, 0},
{0, 0, 0, 0, 0, 0, 1, 0},
{0, 0, 1, 0, 0, 0, 0, 1},
{0, 1, 0, 0, 0, 0, 0, 1},
{0, 0, 0, 0, 0, 1, 0, 1},
{0, 0, 0, 0, 0, 0, 0, 0}
};

int s =
int t =
cout <<
<<

0;
7;
"There can be maximum " << findDisjointPaths(graph, s, t)
" edge-disjoint paths from " << s <<" to "<< t ;

return 0;
}

Output:
There can be maximum 2 edge-disjoint paths from 0 to 7

Time Complexity: Same as time complexity of Edmonds-Karp implementation of Ford-Fulkerson (See time complexity discussed here)
References:
http://www.win.tue.nl/~nikhil/courses/2012/2WO08/max-flow-applications-4up.pdf

Find minimum s-t cut in a flow network

In a flow network, an s-t cut is a cut that requires the source s and the sink t to be in different subsets, and it consists of edges going from the
sources side to the sinks side. The capacity of an s-t cut is defined by the sum of capacity of each edge in the cut-set. (Source: Wiki)
The problem discussed here is to find minimum capacity s-t cut of the given network. Expected output is all edges of the minimum cut.
For example, in the following flow network, example s-t cuts are {{0 ,1}, {0, 2}}, {{0, 2}, {1, 2}, {1, 3}}, etc. The minimum s-t cut is {{1, 3},
{4, 3}, {4 5}} which has capacity as 12+7+4 = 23.

We strongly recommend to read the below post first.

Ford-Fulkerson Algorithm for Maximum Flow Problem
Minimum Cut and Maximum Flow
Like Maximum Bipartite Matching, this is another problem which can solved using Ford-Fulkerson Algorithm. This is based on max-flow min-cut
theorem.
The max-flow min-cut theorem states that in a flow network, the amount of maximum flow is equal to capacity of the minimum cut. See CLRS
book for proof of this theorem.
From Ford-Fulkerson, we get capacity of minimum cut. How to print all edges that form the minimum cut? The idea is to use residual graph.
Following are steps to print all edges of minimum cut.
1) Run Ford-Fulkerson algorithm and consider the final residual graph.
2) Find the set of vertices that are reachable from source in the residual graph.
3) All edges which are from a reachable vertex to non-reachable vertex are minimum cut edges. Print all such edges.
Following is C++ implementation of the above approach.

C++
// C++ program for finding minimum cut using Ford-Fulkerson
#include <iostream>
#include <limits.h>
#include <string.h>
#include <queue>
using namespace std;
// Number of vertices in given graph
#define V 6
/* Returns true if there is a path from source 's' to sink 't' in
residual graph. Also fills parent[] to store the path */
int bfs(int rGraph[V][V], int s, int t, int parent[])
{
// Create a visited array and mark all vertices as not visited
bool visited[V];
memset(visited, 0, sizeof(visited));
// Create a queue, enqueue source vertex and mark source vertex
// as visited
queue <int> q;
q.push(s);
visited[s] = true;
parent[s] = -1;
// Standard BFS Loop
while (!q.empty())
{
int u = q.front();

q.pop();
for (int v=0; v<V; v++)
{
if (visited[v]==false && rGraph[u][v] > 0)
{
q.push(v);
parent[v] = u;
visited[v] = true;
}
}
}
// If we reached sink in BFS starting from source, then return
// true, else false
return (visited[t] == true);
}
// A DFS based function to find all reachable vertices from s. The function
// marks visited[i] as true if i is reachable from s. The initial values in
// visited[] must be false. We can also use BFS to find reachable vertices
void dfs(int rGraph[V][V], int s, bool visited[])
{
visited[s] = true;
for (int i = 0; i < V; i++)
if (rGraph[s][i] && !visited[i])
dfs(rGraph, i, visited);
}
// Prints the minimum s-t cut
void minCut(int graph[V][V], int s, int t)
{
int u, v;
// Create a residual graph and fill the residual graph with
// given capacities in the original graph as residual capacities
// in residual graph
int rGraph[V][V]; // rGraph[i][j] indicates residual capacity of edge i-j
for (u = 0; u < V; u++)
for (v = 0; v < V; v++)
rGraph[u][v] = graph[u][v];
int parent[V]; // This array is filled by BFS and to store path
// Augment the flow while tere is path from source to sink
while (bfs(rGraph, s, t, parent))
{
// Find minimum residual capacity of the edhes along the
// path filled by BFS. Or we can say find the maximum flow
// through the path found.
int path_flow = INT_MAX;
for (v=t; v!=s; v=parent[v])
{
u = parent[v];
path_flow = min(path_flow, rGraph[u][v]);
}
// update residual capacities of the edges and reverse edges
// along the path
for (v=t; v != s; v=parent[v])
{
u = parent[v];
rGraph[u][v] -= path_flow;
rGraph[v][u] += path_flow;
}
}
// Flow is maximum now, find vertices reachable from s
bool visited[V];
memset(visited, false, sizeof(visited));
dfs(rGraph, s, visited);
// Print all edges that are from a reachable vertex to
// non-reachable vertex in the original graph
for (int i = 0; i < V; i++)
for (int j = 0; j < V; j++)
if (visited[i] && !visited[j] && graph[i][j])
cout << i << " - " << j << endl;
return;
}

// Driver program to test above functions

int main()
{
// Let us create a graph shown in the above example
int graph[V][V] = { {0, 16, 13, 0, 0, 0},
{0, 0, 10, 12, 0, 0},
{0, 4, 0, 0, 14, 0},
{0, 0, 9, 0, 0, 20},
{0, 0, 0, 7, 0, 4},
{0, 0, 0, 0, 0, 0}
};
minCut(graph, 0, 5);
return 0;
}

Java
// Java program for implementation of Ford Fulkerson algorithm
import java.util.*;
import java.lang.*;
import java.io.*;
import java.util.LinkedList;
class MaxFlow
{
static final int V=6;

//Number of vertices in graph

/* Returns true if there is a path from source 's' to sink 't'

in residual graph. Also fills parent[] to store the path */
boolean bfs(int rGraph[][], int s, int t, int parent[])
{
// Create a visited array and mark all vertices as not visited
boolean visited[] = new boolean[V];
for(int i=0;i<V;++i)
visited[i]=false;
// Create a queue, enqueue source vertex and mark source vertex
// as visited
LinkedList<Integer> queue = new LinkedList<Integer>();
queue.add(s);
visited[s] = true;
parent[s]=-1;
// Standard BFS Loop
while (queue.size()!=0)
{
int u = queue.poll();
for (int v=0; v<V; v++)
{
if (visited[v]==false && rGraph[u][v] > 0)
{
queue.add(v);
parent[v] = u;
visited[v] = true;
}
}
}
// If we reached sink in BFS starting from source, then return
// true, else false
return (visited[t] == true);
}
// Returns tne maximum flow from s to t in the given graph
int fordFulkerson(int graph[][], int s, int t)
{
int u, v;
// Create a residual graph and fill the residual graph with
// given capacities in the original graph as residual capacities
// in residual graph
// Residual graph where rGraph[i][j] indicates residual capacity
// of edge from i to j (if there is an edge. If rGraph[i][j] is 0,
// then there is not)
int rGraph[][] = new int[V][V];

for (u = 0; u < V; u++)

for (v = 0; v < V; v++)
rGraph[u][v] = graph[u][v];
int parent[] = new int[V]; // filled by BFS and to store path
int max_flow = 0; // There is no flow initially
// Augment the flow while tere is path from source to sink
while (bfs(rGraph, s, t, parent))
{
// Find minimum residual capacity of the edhes along the
// path filled by BFS. Or we can say find the maximum flow
// through the path found.
int path_flow = Integer.MAX_VALUE;
for (v=t; v!=s; v=parent[v])
{
u = parent[v];
path_flow = Math.min(path_flow, rGraph[u][v]);
}
// update residual capacities of the edges and reverse edges
// along the path
for (v=t; v != s; v=parent[v])
{
u = parent[v];
rGraph[u][v] -= path_flow;
rGraph[v][u] += path_flow;
}
// Add path flow to overall flow
max_flow += path_flow;
}
// Return the overall flow
return max_flow;
}
// Driver program to test above functions
public static void main (String[] args) throws java.lang.Exception
{
// Let us create a graph shown in the above example
int graph[][] =new int[][] { {0, 16, 13, 0, 0, 0},
{0, 0, 10, 12, 0, 0},
{0, 4, 0, 0, 14, 0},
{0, 0, 9, 0, 0, 20},
{0, 0, 0, 7, 0, 4},
{0, 0, 0, 0, 0, 0}
};
MaxFlow m = new MaxFlow();
System.out.println("The maximum possible flow is " +
m.fordFulkerson(graph, 0, 5));
}
}

1 - 3
4 - 3
4 - 5

References:
http://www.stanford.edu/class/cs97si/08-network-flow-problems.pdf
http://www.cs.princeton.edu/courses/archive/spring06/cos226/lectures/maxflow.pdf

Maximum Bipartite Matching

A matching in a Bipartite Graph is a set of the edges chosen in such a way that no two edges share an endpoint. A maximum matching is a
matching of maximum size (maximum number of edges). In a maximum matching, if any edge is added to it, it is no longer a matching. There can be
more than one maximum matchings for a given Bipartite Graph.
Why do we care?
There are many real world problems that can be formed as Bipartite Matching. For example, consider the following problem:
There are M job applicants and N jobs. Each applicant has a subset of jobs that he/she is interested in. Each job opening can only accept
one applicant and a job applicant can be appointed for only one job. Find an assignment of jobs to applicants in such that as many
applicants as possible get jobs.

We strongly recommend to read the following post first.

Ford-Fulkerson Algorithm for Maximum Flow Problem
Maximum Bipartite Matching and Max Flow Problem
Maximum Bipartite Matching (MBP) problem can be solved by converting it into a flow network (See this video to know how did we arrive this
conclusion). Following are the steps.

1) Build a Flow Network

There must be a source and sink in a flow network. So we add a source and add edges from source to all applicants. Similarly, add edges from all
jobs to sink. The capacity of every edge is marked as 1 unit.

2) Find the maximum flow.

We use Ford-Fulkerson algorithm to find the maximum flow in the flow network built in step 1. The maximum flow is actually the MBP we are
looking for.

How to implement the above approach?

Let us first define input and output forms. Input is in the form of Edmonds matrix which is a 2D array bpGraph[M][N] with M rows (for M job
applicants) and N columns (for N jobs). The value bpGraph[i][j] is 1 if ith applicant is interested in jth job, otherwise 0.
Output is number maximum number of people that can get jobs.
A simple way to implement this is to create a matrix that represents adjacency matrix representation of a directed graph with M+N+2 vertices.
Call the fordFulkerson() for the matrix. This implementation requires O((M+N)*(M+N)) extra space.
Extra space can be be reduced and code can be simplified using the fact that the graph is bipartite and capacity of every edge is either 0 or 1. The
idea is to use DFS traversal to find a job for an applicant (similar to augmenting path in Ford-Fulkerson). We call bpm() for every applicant, bpm()
is the DFS based function that tries all possibilities to assign a job to the applicant.
In bpm(), we one by one try all jobs that an applicant u is interested in until we find a job, or all jobs are tried without luck. For every job we try,
we do following.
If a job is not assigned to anybody, we simply assign it to the applicant and return true. If a job is assigned to somebody else say x, then we
recursively check whether x can be assigned some other job. To make sure that x doesnt get the same job again, we mark the job v as seen before
we make recursive call for x. If x can get other job, we change the applicant for job v and return true. We use an array maxR[0..N-1] that stores
the applicants assigned to different jobs.
If bmp() returns true, then it means that there is an augmenting path in flow network and 1 unit of flow is added to the result in maxBPM().

C++
// A C++ program to find maximal Bipartite matching.
#include <iostream>
#include <string.h>
using namespace std;
// M is number of applicants and N is number of jobs
#define M 6
#define N 6
// A DFS based recursive function that returns true if a
// matching for vertex u is possible
bool bpm(bool bpGraph[M][N], int u, bool seen[], int matchR[])
{
// Try every job one by one
for (int v = 0; v < N; v++)
{
// If applicant u is interested in job v and v is
// not visited
if (bpGraph[u][v] && !seen[v])
{
seen[v] = true; // Mark v as visited
//
//
//
//
//
if
{

If job 'v' is not assigned to an applicant OR

previously assigned applicant for job v (which is matchR[v])
has an alternate job available.
Since v is marked as visited in the above line, matchR[v]
in the following recursive call will not get job 'v' again
(matchR[v] < 0 || bpm(bpGraph, matchR[v], seen, matchR))
matchR[v] = u;
return true;

}
}
}
return false;
}
// Returns maximum number of matching from M to N
int maxBPM(bool bpGraph[M][N])
{
// An array to keep track of the applicants assigned to
// jobs. The value of matchR[i] is the applicant number
// assigned to job i, the value -1 indicates nobody is
// assigned.
int matchR[N];
// Initially all jobs are available
memset(matchR, -1, sizeof(matchR));

int result = 0; // Count of jobs assigned to applicants

for (int u = 0; u < M; u++)
{
// Mark all jobs as not seen for next applicant.
bool seen[N];
memset(seen, 0, sizeof(seen));
// Find if the applicant 'u' can get a job
if (bpm(bpGraph, u, seen, matchR))
result++;
}
return result;
}
// Driver program to test above functions
int main()
{
// Let us create a bpGraph shown in the above example
bool bpGraph[M][N] = { {0, 1, 1, 0, 0, 0},
{1, 0, 0, 1, 0, 0},
{0, 0, 1, 0, 0, 0},
{0, 0, 1, 1, 0, 0},
{0, 0, 0, 0, 0, 0},
{0, 0, 0, 0, 0, 1}
};
cout << "Maximum number of applicants that can get job is "
<< maxBPM(bpGraph);
return 0;
}

Java
// A Java program to find maximal Bipartite matching.
import java.util.*;
import java.lang.*;
import java.io.*;
class MaxBipartite
{
// M is number of applicants and N is number of jobs
static final int M = 6;
static final int N = 6;
// A DFS based recursive function that returns true if a
// matching for vertex u is possible
boolean bpm(boolean bpGraph[][], int u, boolean seen[],
int matchR[])
{
// Try every job one by one
for (int v = 0; v < N; v++)
{
// If applicant u is interested in job v and v
// is not visited
if (bpGraph[u][v] && !seen[v])
{
seen[v] = true; // Mark v as visited
//
//
//
//
//
//
if

If job 'v' is not assigned to an applicant OR

previously assigned applicant for job v (which
is matchR[v]) has an alternate job available.
Since v is marked as visited in the above line,
matchR[v] in the following recursive call will
not get job 'v' again
(matchR[v] < 0 || bpm(bpGraph, matchR[v],
seen, matchR))

{
matchR[v] = u;
return true;
}
}
}
return false;
}
// Returns maximum number of matching from M to N
int maxBPM(boolean bpGraph[][])
{

// An array to keep track of the applicants assigned to

// jobs. The value of matchR[i] is the applicant number
// assigned to job i, the value -1 indicates nobody is
// assigned.
int matchR[] = new int[N];
// Initially all jobs are available
for(int i=0; i<N; ++i)
matchR[i] = -1;
int result = 0; // Count of jobs assigned to applicants
for (int u = 0; u < M; u++)
{
// Mark all jobs as not seen for next applicant.
boolean seen[] =new boolean[N] ;
for(int i=0; i<N; ++i)
seen[i] = false;
// Find if the applicant 'u' can get a job
if (bpm(bpGraph, u, seen, matchR))
result++;
}
return result;
}
// Driver method
public static void main (String[] args) throws java.lang.Exception
{
// Let us create a bpGraph shown in the above example
boolean bpGraph[][] = new boolean[][]{
{false, true, true, false, false, false},
{true, false, false, true, false, false},
{false, false, true, false, false, false},
{false, false, true, true, false, false},
{false, false, false, false, false, false},
{false, false, false, false, false, true}
};
MaxBipartite m = new MaxBipartite();
System.out.println( "Maximum number of applicants that can"+
" get job is "+m.maxBPM(bpGraph));
}
}

Maximum number of applicants that can get job is 5

References:
http://www.cs.cornell.edu/~wdtseng/icpc/notes/graph_part5.pdf
http://www.youtube.com/watch?v=NlQqmEXuiC8
http://en.wikipedia.org/wiki/Maximum_matching
http://www.stanford.edu/class/cs97si/08-network-flow-problems.pdf
http://www.cs.princeton.edu/courses/archive/spring13/cos423/lectures/07NetworkFlowII-22.pdf
http://www.ise.ncsu.edu/fangroup/or766.dir/or766_ch7.pdf

Channel Assignment Problem

There are M transmitter and N receiver stations. Given a matrix that keeps track of the number of packets to be transmitted from a given
transmitter to a receiver. If the (i; j)-th entry of the matrix is k, it means at that time the station i has k packets for transmission to station j.
During a time slot, a transmitter can send only one packet and a receiver can receive only one packet. Find the channel assignments so that
maximum number of packets are transferred from transmitters to receivers during the next time slot.
Example:
0 2 0
3 0 1
2 4 0

The above is the input format. We call the above matrix M. Each value M[i; j] represents the number of packets Transmitter i has to send to
Receiver j. The output should be:
The number of maximum packets sent in the time slot is 3
T1 -> R2
T2 -> R3
T3 -> R1

Note that the maximum number of packets that can be transferred in any slot is min(M, N).
Algorithm:
The channel assignment problem between sender and receiver can be easily transformed into Maximum Bipartite Matching(MBP) problem that
can be solved by converting it into a flow network.
Step 1: Build a Flow Network
There must be a source and sink in a flow network. So we add a dummy source and add edges from source to all senders. Similarly, add edges
from all receivers to dummy sink. The capacity of all added edges is marked as 1 unit.
Step 2: Find the maximum flow.
We use Ford-Fulkerson algorithm to find the maximum flow in the flow network built in step 1. The maximum flow is actually the maximum number
of packets that can be transmitted without bandwidth interference in a time slot.
Implementation:
Let us first define input and output forms. Input is in the form of Edmonds matrix which is a 2D array table[M][N] with M rows (for M senders)
and N columns (for N receivers). The value table[i][j] is the number of packets that has to be sent from transmitter i to receiver j. Output is the
maximum number of packets that can be transmitted without bandwidth interference in a time slot.
A simple way to implement this is to create a matrix that represents adjacency matrix representation of a directed graph with M+N+2 vertices.
Call the fordFulkerson() for the matrix. This implementation requires O((M+N)*(M+N)) extra space.
Extra space can be reduced and code can be simplified using the fact that the graph is bipartite. The idea is to use DFS traversal to find a receiver
for a transmitter (similar to augmenting path in Ford-Fulkerson). We call bpm() for every applicant, bpm() is the DFS based function that tries all
possibilities to assign a receiver to the sender. In bpm(), we one by one try all receivers that a sender u is interested in until we find a receiver, or all
receivers are tried without luck.
For every receiver we try, we do following:
If a receiver is not assigned to anybody, we simply assign it to the sender and return true. If a receiver is assigned to somebody else say x, then we
recursively check whether x can be assigned some other receiver. To make sure that x doesnt get the same receiver again, we mark the receiver v
as seen before we make recursive call for x. If x can get other receiver, we change the sender for receiver v and return true. We use an array
maxR[0..N-1] that stores the senders assigned to different receivers.
If bmp() returns true, then it means that there is an augmenting path in flow network and 1 unit of flow is added to the result in maxBPM().
Time and space complexity analysis:
In case of bipartite matching problem, F ? |V| since there can be only |V| possible edges coming out from source node. So the total running time is
O(mn) = O((m + n)n). The space complexity is also substantially reduces from O ((M+N)*(M+N)) to just a single dimensional array of size M
thus storing the mapping between M and N.
#include <iostream>
#include <string.h>
#include <vector>
#define M 3
#define N 4
using namespace std;
// A Depth First Search based recursive function that returns true
// if a matching for vertex u is possible
bool bpm(int table[M][N], int u, bool seen[], int matchR[])
{
// Try every receiver one by one
for (int v = 0; v < N; v++)
{
// If sender u has packets to send to receiver v and

//
//
//
if
{

receiver v is not already mapped to any other sender

just check if the number of packets is greater than '0'
because only one packet can be sent in a time frame anyways
(table[u][v]>0 && !seen[v])
seen[v] = true; // Mark v as visited
//
//
//
//
//
if
{

If receiver 'v' is not assigned to any sender OR

previously assigned sender for receiver v (which is
matchR[v]) has an alternate receiver available. Since
v is marked as visited in the above line, matchR[v] in
the following recursive call will not get receiver 'v' again
(matchR[v] < 0 || bpm(table, matchR[v], seen, matchR))
matchR[v] = u;
return true;

}
}
}
return false;
}
// Returns maximum number of packets
// time slot from sender to receiver
int maxBPM(int table[M][N])
{
// An array to keep track of the
// The value of matchR[i] is the
// the value -1 indicates nobody
int matchR[N];

that can be sent parallely in 1

receivers assigned to the senders.

sender ID assigned to receiver i.
is assigned.

// Initially all receivers are not mapped to any senders

memset(matchR, -1, sizeof(matchR));
int result = 0; // Count of receivers assigned to senders
for (int u = 0; u < M; u++)
{
// Mark all receivers as not seen for next sender
bool seen[N];
memset(seen, 0, sizeof(seen));
// Find if the sender 'u' can be assigned to the receiver
if (bpm(table, u, seen, matchR))
result++;
}
cout << "The number of maximum packets sent in the time slot is "
<< result << "\n";
for (int x=0; x<N; x++)
if (matchR[x]+1!=0)
cout << "T" << matchR[x]+1 << "-> R" << x+1 << "\n";
return result;
}
// Driver program to test above function
int main()
{
int table[M][N] = {{0, 2, 0}, {3, 0, 1}, {2, 4, 0}};
int max_flow = maxBPM(table);
return 0;
}

Output:
The number of maximum packets sent in the time slot is 3
T3-> R1
T1-> R2
T2-> R3

Given an array of strings, find if the strings can be chained to form a circle
Given an array of strings, find if the given strings can be chained to form a circle. A string X can be put before another string Y in circle if the last
character of X is same as first character of Y.
Examples:
Input: arr[] = {"geek", "king"}
Output: Yes, the given strings can be chained.
Note that the last character of first string is same
as first character of second string and vice versa is
also true.
Input: arr[] = {"for", "geek", "rig", "kaf"}
Output: Yes, the given strings can be chained.
The strings can be chained as "for", "rig", "geek"
and "kaf"
Input: arr[] = {"aab", "bac", "aaa", "cda"}
Output: Yes, the given strings can be chained.
The strings can be chained as "aaa", "aab", "bac"
and "cda"
Input: arr[] = {"aaa", "bbb", "baa", "aab"};
Output: Yes, the given strings can be chained.
The strings can be chained as "aaa", "aab", "bbb"
and "baa"
Input: arr[] = {"aaa"};
Output: Yes
Input: arr[] = {"aaa", "bbb"};
Output: No

The idea is to create a directed graph of all characters and then find if their is an eulerian circuit in the graph or not. If there is an eulerian circuit,
then chain can be formed, otherwise not.
Note that a directed graph has eulerian circuit only if in degree and out degree of every vertex is same, and all non-zero degree vertices form a
single strongly connected component.
Following are detailed steps of the algorithm.
1) Create a directed graph g with number of vertices equal to the size of alphabet. We have created a graph with 26 vertices in the below
program.
2) Do following for every string in the given array of strings.
..a) Add an edge from first character to last character of the given graph.
3) If the created graph has eulerian circuit, then return true, else return false.
Following is C++ implementation of the above algorithm.
// A C++ program to check if a given directed graph is Eulerian or not
#include<iostream>
#include <list>
#define CHARS 26
using namespace std;
// A class that represents an undirected graph
class Graph
{
int V;
// No. of vertices
list<int> *adj;
// A dynamic array of adjacency lists
int *in;
public:
// Constructor and destructor
Graph(int V);
~Graph() { delete [] adj; delete [] in; }
// function to add an edge to graph
void addEdge(int v, int w) { adj[v].push_back(w); (in[w])++; }
// Method to check if this graph is Eulerian or not
bool isEulerianCycle();
// Method to check if all non-zero degree vertices are connected
bool isSC();

// Function to do DFS starting from v. Used in isConnected();

void DFSUtil(int v, bool visited[]);
Graph getTranspose();
};
Graph::Graph(int V)
{
this->V = V;
adj = new list<int>[V];
in = new int[V];
for (int i = 0; i < V; i++)
in[i] = 0;
}
/* This function returns true if the directed graph has an eulerian
cycle, otherwise returns false */
bool Graph::isEulerianCycle()
{
// Check if all non-zero degree vertices are connected
if (isSC() == false)
return false;
// Check if in degree and out degree of every vertex is same
for (int i = 0; i < V; i++)
if (adj[i].size() != in[i])
return false;
return true;
}
// A recursive function to do DFS starting from v
void Graph::DFSUtil(int v, bool visited[])
{
// Mark the current node as visited and print it
visited[v] = true;
// Recur for all the vertices adjacent to this vertex
list<int>::iterator i;
for (i = adj[v].begin(); i != adj[v].end(); ++i)
if (!visited[*i])
DFSUtil(*i, visited);
}
// Function that returns reverse (or transpose) of this graph
// This function is needed in isSC()
Graph Graph::getTranspose()
{
Graph g(V);
for (int v = 0; v < V; v++)
{
// Recur for all the vertices adjacent to this vertex
list<int>::iterator i;
for(i = adj[v].begin(); i != adj[v].end(); ++i)
{
g.adj[*i].push_back(v);
(g.in[v])++;
}
}
return g;
}
// This function returns true if all non-zero degree vertices of
// graph are strongly connected. Please refer
// http://www.geeksforgeeks.org/connectivity-in-a-directed-graph/
bool Graph::isSC()
{
// Mark all the vertices as not visited (For first DFS)
bool visited[V];
for (int i = 0; i < V; i++)
visited[i] = false;
// Find the first vertex with non-zero degree
int n;
for (n = 0; n < V; n++)
if (adj[n].size() > 0)
break;
// Do DFS traversal starting from first non zero degree vertex.
DFSUtil(n, visited);

// If DFS traversal doesnt visit all vertices, then return false.

for (int i = 0; i < V; i++)
if (adj[i].size() > 0 && visited[i] == false)
return false;
// Create a reversed graph
Graph gr = getTranspose();
// Mark all the vertices as not visited (For second DFS)
for (int i = 0; i < V; i++)
visited[i] = false;
// Do DFS for reversed graph starting from first vertex.
// Staring Vertex must be same starting point of first DFS
gr.DFSUtil(n, visited);
// If all vertices are not visited in second DFS, then
// return false
for (int i = 0; i < V; i++)
if (adj[i].size() > 0 && visited[i] == false)
return false;
return true;
}
// This function takes an of strings and returns true
// if the given array of strings can be chained to
// form cycle
bool canBeChained(string arr[], int n)
{
// Create a graph with 'aplha' edges
Graph g(CHARS);
// Create an edge from first character to last character
// of every string
for (int i = 0; i < n; i++)
{
string s = arr[i];
g.addEdge(s[0]-'a', s[s.length()-1]-'a');
}
// The given array of strings can be chained if there
// is an eulerian cycle in the created graph
return g.isEulerianCycle();
}
// Driver program to test above functions
int main()
{
string arr1[] = {"for", "geek", "rig", "kaf"};
int n1 = sizeof(arr1)/sizeof(arr1[0]);
canBeChained(arr1, n1)? cout << "Can be chained \n" :
cout << "Can't be chained \n";
string arr2[] = {"aab", "abb"};
int n2 = sizeof(arr2)/sizeof(arr2[0]);
canBeChained(arr2, n2)? cout << "Can be chained \n" :
cout << "Can't be chained \n";
return 0;
}

Output:
Can be chained
Can't be chained

Given a sorted dictionary of an alien language, find order of characters

Given a sorted dictionary (array of words) of an alien language, find order of characters in the language.
Examples:
Input: words[] = {"baa", "abcd", "abca", "cab", "cad"}
Output: Order of characters is 'b', 'd', 'a', 'c'
Note that words are sorted and in the given language "baa"
comes before "abcd", therefore 'b' is before 'a' in output.
Similarly we can find other orders.
Input: words[] = {"caa", "aaa", "aab"}
Output: Order of characters is 'c', 'a', 'b'

The idea is to create a graph of characters and then find topological sorting of the created graph. Following are the detailed steps.
1) Create a graph g with number of vertices equal to the size of alphabet in the given alien language. For example, if the alphabet size is 5, then
there can be 5 characters in words. Initially there are no edges in graph.
2) Do following for every pair of adjacent words in given sorted array.
..a) Let the current pair of words be word1 and word2. One by one compare characters of both words and find the first mismatching characters.
..b) Create an edge in g from mismatching character of word1 to that of word2.
3) Print topological sorting of the above created graph.
Following is C++ implementation of the above algorithm.
// A C++ program to order of characters in an alien language
#include<iostream>
#include <list>
#include <stack>
#include <cstring>
using namespace std;
// Class to represent a graph
class Graph
{
int V;
// No. of vertices'
// Pointer to an array containing adjacency listsList
list<int> *adj;
// A function used by topologicalSort
void topologicalSortUtil(int v, bool visited[], stack<int> &Stack);
public:
Graph(int V); // Constructor
// function to add an edge to graph
void addEdge(int v, int w);
// prints a Topological Sort of the complete graph
void topologicalSort();
};
Graph::Graph(int V)
{
this->V = V;
adj = new list<int>[V];
}
void Graph::addEdge(int v, int w)
{
adj[v].push_back(w); // Add w to vs list.
}
// A recursive function used by topologicalSort
void Graph::topologicalSortUtil(int v, bool visited[], stack<int> &Stack)
{
// Mark the current node as visited.
visited[v] = true;
// Recur for all the vertices adjacent to this vertex
list<int>::iterator i;
for (i = adj[v].begin(); i != adj[v].end(); ++i)
if (!visited[*i])
topologicalSortUtil(*i, visited, Stack);

// Push current vertex to stack which stores result

Stack.push(v);
}
// The function to do Topological Sort. It uses recursive topologicalSortUtil()
void Graph::topologicalSort()
{
stack<int> Stack;
// Mark all the vertices as not visited
bool *visited = new bool[V];
for (int i = 0; i < V; i++)
visited[i] = false;
// Call the recursive helper function to store Topological Sort
// starting from all vertices one by one
for (int i = 0; i < V; i++)
if (visited[i] == false)
topologicalSortUtil(i, visited, Stack);
// Print contents of stack
while (Stack.empty() == false)
{
cout << (char) ('a' + Stack.top()) << " ";
Stack.pop();
}
}
int min(int x, int y)
{
return (x < y)? x : y;
}
// This function fidns and prints order of characer from a sorted
// array of words. n is size of words[]. alpha is set of possible
// alphabets.
// For simplicity, this function is written in a way that only
// first 'alpha' characters can be there in words array. For
// example if alpha is 7, then words[] should have only 'a', 'b',
// 'c' 'd', 'e', 'f', 'g'
void printOrder(string words[], int n, int alpha)
{
// Create a graph with 'aplha' edges
Graph g(alpha);
// Process all adjacent pairs of words and create a graph
for (int i = 0; i < n-1; i++)
{
// Take the current two words and find the first mismatching
// character
string word1 = words[i], word2 = words[i+1];
for (int j = 0; j < min(word1.length(), word2.length()); j++)
{
// If we find a mismatching character, then add an edge
// from character of word1 to that of word2
if (word1[j] != word2[j])
{
g.addEdge(word1[j]-'a', word2[j]-'a');
break;
}
}
}
// Print topological sort of the above created graph
g.topologicalSort();
}
// Driver program to test above functions
int main()
{
string words[] = {"caa", "aaa", "aab"};
printOrder(words, 3, 3);
return 0;
}

Output:
c a b

Time Complexity: The first step to create a graph takes O(n + alhpa) time where n is number of given words and alpha is number of characters
in given alphabet. The second step is also topological sorting. Note that there would be alpha vertices and at-most (n-1) edges in the graph. The

time complexity of topological sorting is O(V+E) which is O(n + aplha) here. So overall time complexity is O(n + aplha) + O(n + aplha) which is
O(n + aplha).
Exercise:
The above code doesnt work when the input is not valid. For example {aba, bba, aaa} is not valid, because from first two words, we can deduce
a should appear before b, but from last two words, we can deduce b should appear before a which is not possible. Extend the above program to
handle invalid inputs and generate the output as Not valid.

Kargers algorithm for Minimum Cut | Set 1 (Introduction and Implementation)

Given an undirected and unweighted graph, find the smallest cut (smallest number of edges that disconnects the graph into two components).
The input graph may have parallel edges.
For example consider the following example, the smallest cut has 2 edges.

A Simple Solution use Max-Flow based s-t cut algorithm to find minimum cut. Consider every pair of vertices as source s and sink t, and call
minimum s-t cut algorithm to find the s-t cut. Return minimum of all s-t cuts. Best possible time complexity of this algorithm is O(V5) for a graph.
[How? there are total possible V2 pairs and s-t cut algorithm for one pair takes O(V*E) time and E = O(V2)].
Below is simple Kargers Algorithm for this purpose. Below Kargers algorithm can be implemented in O(E) = O(V2) time.
1) Initialize contracted graph CG as copy of original graph
2) While there are more than 2 vertices.
a) Pick a random edge (u, v) in the contracted graph.
b) Merge (or contract) u and v into a single vertex (update
the contracted graph).
c) Remove self-loops
3) Return cut represented by two vertices.

Let us understand above algorithm through the example given.

Let the first randomly picked vertex be a which connects vertices 0 and 1. We remove this edge and contract the graph (combine vertices 0 and
1). We get the following graph.

Let the next randomly picked edge be d. We remove this edge and combine vertices (0,1) and 3.

We need to remove self-loops in the graph. So we remove edge c

Now graph has two vertices, so we stop. The number of edges in the resultant graph is the cut produced by Kargers algorithm.
Kargers algorithm is a Monte Carlo algorithm and cut produced by it may not be minimum. For example, the following diagram shows
that a different order of picking random edges produces a min-cut of size 3.

Below is C++ implementation of above algorithm. The input graph is represented as a collection of edges and union-find data structure is used to
keep track of components.
// Karger's algorithm to find Minimum Cut in an
// undirected, unweighted and connected graph.
#include <stdio.h>
#include <stdlib.h>
#include <time.h>
// a structure to represent a unweighted edge in graph
struct Edge
{
int src, dest;
};
// a structure to represent a connected, undirected
// and unweighted graph as a collection of edges.
struct Graph
{
// V-> Number of vertices, E-> Number of edges
int V, E;
// graph is represented as an array of edges.
// Since the graph is undirected, the edge
// from src to dest is also edge from dest
// to src. Both are counted as 1 edge here.
Edge* edge;
};
// A structure to represent a subset for union-find
struct subset
{
int parent;
int rank;
};
// Function prototypes for union-find (These functions are defined
// after kargerMinCut() )
int find(struct subset subsets[], int i);
void Union(struct subset subsets[], int x, int y);
// A very basic implementation of Karger's randomized
// algorithm for finding the minimum cut. Please note
// that Karger's algorithm is a Monte Carlo Randomized algo
// and the cut returned by the algorithm may not be
// minimum always
int kargerMinCut(struct Graph* graph)
{
// Get data of given graph
int V = graph->V, E = graph->E;
Edge *edge = graph->edge;
// Allocate memory for creating V subsets.
struct subset *subsets = new subset[V];
// Create V subsets with single elements
for (int v = 0; v < V; ++v)
{
subsets[v].parent = v;

subsets[v].rank = 0;
}
// Initially there are V vertices in
// contracted graph
int vertices = V;
// Keep contracting vertices until there are
// 2 vertices.
while (vertices > 2)
{
// Pick a random edge
int i = rand() % E;
// Find vertices (or sets) of two corners
// of current edge
int subset1 = find(subsets, edge[i].src);
int subset2 = find(subsets, edge[i].dest);
// If two corners belong to same subset,
// then no point considering this edge
if (subset1 == subset2)
continue;
// Else contract the edge (or combine the
// corners of edge into one vertex)
else
{
printf("Contracting edge %d-%d\n",
edge[i].src, edge[i].dest);
vertices--;
Union(subsets, subset1, subset2);
}
}
// Now we have two vertices (or subsets) left in
// the contracted graph, so count the edges between
// two components and return the count.
int cutedges = 0;
for (int i=0; i<E; i++)
{
int subset1 = find(subsets, edge[i].src);
int subset2 = find(subsets, edge[i].dest);
if (subset1 != subset2)
cutedges++;
}
return cutedges;
}
// A utility function to find set of an element i
// (uses path compression technique)
int find(struct subset subsets[], int i)
{
// find root and make root as parent of i
// (path compression)
if (subsets[i].parent != i)
subsets[i].parent =
find(subsets, subsets[i].parent);
return subsets[i].parent;
}
// A function that does union of two sets of x and y
// (uses union by rank)
void Union(struct subset subsets[], int x, int y)
{
int xroot = find(subsets, x);
int yroot = find(subsets, y);
// Attach smaller rank tree under root of high
// rank tree (Union by Rank)
if (subsets[xroot].rank < subsets[yroot].rank)
subsets[xroot].parent = yroot;
else if (subsets[xroot].rank > subsets[yroot].rank)
subsets[yroot].parent = xroot;
// If ranks are same, then make one as root and
// increment its rank by one
else
{

subsets[yroot].parent = xroot;
subsets[xroot].rank++;
}
}
// Creates a graph with V vertices and E edges
struct Graph* createGraph(int V, int E)
{
Graph* graph = new Graph;
graph->V = V;
graph->E = E;
graph->edge = new Edge[E];
return graph;
}
// Driver program to test above functions
int main()
{
/* Let us create following unweighted graph
0------1
| \
|
| \ |
|
\|
2------3 */
int V = 4; // Number of vertices in graph
int E = 5; // Number of edges in graph
struct Graph* graph = createGraph(V, E);
// add edge 0-1
graph->edge[0].src = 0;
graph->edge[0].dest = 1;
// add edge 0-2
graph->edge[1].src = 0;
graph->edge[1].dest = 2;
// add edge 0-3
graph->edge[2].src = 0;
graph->edge[2].dest = 3;
// add edge 1-3
graph->edge[3].src = 1;
graph->edge[3].dest = 3;
// add edge 2-3
graph->edge[4].src = 2;
graph->edge[4].dest = 3;
// Use a different seed value for every run.
srand(time(NULL));
printf("\nCut found by Karger's randomized algo is %d\n",
kargerMinCut(graph));
return 0;
}

Output:
Contracting edge 0-2
Contracting edge 0-3
Cut found by Karger's randomized algo is 2

Note that the above program is based on outcome of a random function and may produce different output.
In this post, we have discussed simple Kargers algorithm and have seen that the algorithm doesnt always produce min-cut. The above algorithm
produces min-cut with probability greater or equal to that 1/(n2). See next post on Analysis and Applications of Kargers Algortihm, applications,
proof of this probability and improvements are discussed.
References:
http://en.wikipedia.org/wiki/Karger%27s_algorithm
https://www.youtube.com/watch?v=P0l8jMDQTEQ
https://www.cs.princeton.edu/courses/archive/fall13/cos521/lecnotes/lec2final.pdf
http://web.stanford.edu/class/archive/cs/cs161/cs161.1138/lectures/11/Small11.pdf

Kargers algorithm for Minimum Cut | Set 2 (Analysis and Applications)

We have introduced and discussed below Kargers algorithm in set 1.
1) Initialize contracted graph CG as copy of original graph
2) While there are more than 2 vertices.
a) Pick a random edge (u, v) in the contracted graph.
b) Merge (or contract) u and v into a single vertex (update
the contracted graph).
c) Remove self-loops
3) Return cut represented by two vertices.

As discussed in the previous post, Kargers algorithm doesnt always find min cut. In this post, the probability of finding min-cut is discussed.
Probability that the cut produced by Kargers Algorithm is Min-Cut is at greater than or equal to 1/(n2)
Proof:
Let there be a unique Min-Cut of given graph and let there be C edges in the Min-Cut and the edges be {e1, e2, e3, .. ec}. The Kargers algorithm
would produce this Min-Cut if and only if none of the edges in set {e1, e2, e3, .. ec} is removed in iterations in the main while loop of above
algorithm.

c is number of edges in min-cut

m is total number of edges
n is total number of vertices
S1 = Event that
e3, .. ec}
S2 = Event that
e3, .. ec}
S3 = Event that
e3, .. ec}

one of the edges in {e1, e2,

is chosen in 1st iteration.
one of the edges in {e1, e2,
is chosen in 2nd iteration.
one of the edges in {e1, e2,
is chosen in 3rd iteration.

..................
..................
The cut produced by Karger's algorithm would be a min-cut if none of the above
events happen.
So the required probability is P[S1' ? S2' ? S3' ? ............]

Probability that a min-cut edge is chosen in first iteration:

Let us calculate P[S1']
P[S1] = c/m
P[S1'] = (1 - c/m)
Above value is in terms of m (or edges), let us convert
it in terms of n (or vertices) using below 2 facts..
1) Since size of min-cut is c, degree of all vertices must be greater
than or equal to c.
2) As per Handshaking Lemma, sum of degrees of all vertices = 2m
From above two facts, we can conclude below.
n*c <= 2m
m >= nc/2
P[S1] <= c / (cn/2)
<= 2/n

P[S1] <= c / (cn/2)

<= 2/n
P[S1'] >= (1-2/n) ------------(1)

Probability that a min-cut edge is chosen in second iteration:

P[S1' ? S2'] = P[S2' | S1' ] * P[S1']
In the above expression, we know value of P[S1'] >= (1-2/n)
P[S2' | S1'] is conditional probability that is, a min cut is
not chosen in second iteration given that it is not chosen in first iteration
Since there are total (n-1) edges left now and number of cut edges is still c,
we can replace n by n-1 in inequality (1). So we get.
P[S2' | S1' ] >= (1 - 2/(n-1))
P[S1' ? S2'] >= (1-2/n) x (1-2/(n-1))

Probability that a min-cut edge is chosen in all iterations:

P[S1' ? S2' ? S3' ?.......... ? Sn-2']
>= [1 - 2/n] * [1 - 2/(n-1)] * [1 - 2/(n-2)] * [1 - 2/(n-3)] *...
... * [1 - 2/(n - (n-4)] * [1 - 2/(n - (n-3)]
>= [(n-2)/n] * [(n-3)/(n-1)] * [(n-4)/(n-2)] * .... 2/4 * 2/3
>= 2/(n * (n-1))
>= 1/n2

How to increase probability of success?

The above probability of success of basic algorithm is very less. For example, for a graph with 10 nodes, the probability of finding the min-cut is
greater than or equal to 1/100. The probability can be increased by repeated runs of basic algorithm and return minimum of all cuts found.
Applications:
1) In war situation, a party would be interested in finding minimum number of links that break communication network of enemy.
2) The min-cut problem can be used to study reliability of a network (smallest number of edges that can fail).
3) Study of network optimization (find a maximum flow).
4) Clustering problems (edges like associations rules) Matching problems (an NC algorithm for min-cut in directed graphs would result in an NC
algorithm for maximum matching in bipartite graphs)
5) Matching problems (an NC algorithm for min-cut in directed graphs would result in an NC algorithm for maximum matching in bipartite graphs)
Sources:
https://www.youtube.com/watch?v=-UuivvyHPas
http://disi.unal.edu.co/~gjhernandezp/psc/lectures/02/MinCut.pdf

HopcroftKarp Algorithm for Maximum Matching | Set 1 (Introduction)

A matching in a Bipartite Graph is a set of the edges chosen in such a way that no two edges share an endpoint. A maximum matching is a
matching of maximum size (maximum number of edges). In a maximum matching, if any edge is added to it, it is no longer a matching. There can be
more than one maximum matching for a given Bipartite Graph.
We have discussed importance of maximum matching and Ford Fulkerson Based approach for maximal Bipartite Matching in previous post. Time
complexity of the Ford Fulkerson based algorithm is O(V x E).
Hopcroft Karp algorithm is an improvement that runs in O(?V x E) time. Let us define few terms before we discuss the algorithm
Free Node or Vertex: Given a matching M, a node that is not part of matching is called free node. Initially all vertices as free (See first graph of
below diagram). In second graph, u2 and v2 are free. In third graph, no vertex is free.
Matching and Not-Matching edges: Given a matching M, edges that are part of matching are called Matching edges and edges that are not part
of M (or connect free nodes) are called Not-Matching edges. In first graph, all edges are non-matching. In second graph, (u0, v1), (u1, v0) and
(u3, v3) are matching and others not-matching.
Alternating Paths: Given a matching M, an alternating path is a path in which the edges belong alternatively to the matching and not matching. All
single edges paths are alternating paths. Examples of alternating paths in middle graph are u0-v1-u2 and u2-v1-u0-v2.
Augmenting path: Given a matching M, an augmenting path is an alternating path that starts from and ends on free vertices. All single edge paths
that start and end with free vertices are augmenting paths. In below diagram, augmenting paths are highlighted with blue color. Note that the
augmenting path always has one extra matching edge.
The Hopcroft Karp algorithm is based on below concept.
A matching M is not maximum if there exists an augmenting path. It is also true other way, i.e, a matching is maximum if no augmenting
path exists
So the idea is to one by one look for augmenting paths. And add the found paths to current matching.
Hopcroft Karp Algorithm
1) Initialize Maximal Matching M as empty.
2) While there exists an Augmenting Path p
Remove matching edges of p from M and add not-matching edges of p to M
(This increases size of M by 1 as p starts and ends with a free vertex)
3) Return M.

Below diagram shows working of the algorithm.

In the initial graph all single edges are augmenting paths and we can pick in any order. In the middle stage, there is only one augmenting path. We
remove matching edges of this path from M and add not-matching edges. In final matching, there are no augmenting paths so the matching is
maximum.
Implementation of Hopcroft Karp algorithm is discussed in set 2.
HopcroftKarp Algorithm for Maximum Matching | Set 2 (Implementation)
References:
https://en.wikipedia.org/wiki/Hopcroft%E2%80%93Karp_algorithm
http://www.dis.uniroma1.it/~leon/tcs/lecture2.pdf

HopcroftKarp Algorithm for Maximum Matching | Set 1 (Introduction)

A matching in a Bipartite Graph is a set of the edges chosen in such a way that no two edges share an endpoint. A maximum matching is a
matching of maximum size (maximum number of edges). In a maximum matching, if any edge is added to it, it is no longer a matching. There can be
more than one maximum matching for a given Bipartite Graph.
We have discussed importance of maximum matching and Ford Fulkerson Based approach for maximal Bipartite Matching in previous post. Time
complexity of the Ford Fulkerson based algorithm is O(V x E).
Hopcroft Karp algorithm is an improvement that runs in O(?V x E) time. Let us define few terms before we discuss the algorithm
Free Node or Vertex: Given a matching M, a node that is not part of matching is called free node. Initially all vertices as free (See first graph of
below diagram). In second graph, u2 and v2 are free. In third graph, no vertex is free.
Matching and Not-Matching edges: Given a matching M, edges that are part of matching are called Matching edges and edges that are not part
of M (or connect free nodes) are called Not-Matching edges. In first graph, all edges are non-matching. In second graph, (u0, v1), (u1, v0) and
(u3, v3) are matching and others not-matching.
Alternating Paths: Given a matching M, an alternating path is a path in which the edges belong alternatively to the matching and not matching. All
single edges paths are alternating paths. Examples of alternating paths in middle graph are u0-v1-u2 and u2-v1-u0-v2.
Augmenting path: Given a matching M, an augmenting path is an alternating path that starts from and ends on free vertices. All single edge paths
that start and end with free vertices are augmenting paths. In below diagram, augmenting paths are highlighted with blue color. Note that the
augmenting path always has one extra matching edge.
The Hopcroft Karp algorithm is based on below concept.
A matching M is not maximum if there exists an augmenting path. It is also true other way, i.e, a matching is maximum if no augmenting
path exists
So the idea is to one by one look for augmenting paths. And add the found paths to current matching.
Hopcroft Karp Algorithm
1) Initialize Maximal Matching M as empty.
2) While there exists an Augmenting Path p
Remove matching edges of p from M and add not-matching edges of p to M
(This increases size of M by 1 as p starts and ends with a free vertex)
3) Return M.

Below diagram shows working of the algorithm.

In the initial graph all single edges are augmenting paths and we can pick in any order. In the middle stage, there is only one augmenting path. We
remove matching edges of this path from M and add not-matching edges. In final matching, there are no augmenting paths so the matching is
maximum.
Implementation of Hopcroft Karp algorithm is discussed in set 2.
HopcroftKarp Algorithm for Maximum Matching | Set 2 (Implementation)
References:
https://en.wikipedia.org/wiki/Hopcroft%E2%80%93Karp_algorithm
http://www.dis.uniroma1.it/~leon/tcs/lecture2.pdf

Linearity of Expectation
This post is about a mathematical concept, but it covers one of the required topics to understand Randomized Algorithms.
Let us consider the following simple problem.
Given a fair dice with 6 faces, the dice is thrown n times, find expected value of sum of all results.
For example, if n = 2, there are total 36 possible outcomes.
(1, 1), (1, 2), .... (1, 6)
(2, 1), (2, 2), .... (2, 6)
................
................
(6, 1), (6, 2), ..... (6, 6)

Expected value of a discrete random variable is R defined as following. Suppose R can take value r1 with probability p1, value r2 with probability
p2, and so on, up to value rk with probability pk. Then the expectation of this random variable R is defined as
E[R] = r1*p1 + r2*p2 + ... rk*pk

Let us calculate expected value for the above example.

Expected Value of sum = 2*1/36 + 3*1/6 + .... + 7*1/36 +
of two dice throws
3*1/36 + 4*1/6 + .... + 8*1/36 +
........................
.........................
7*1/36 + 8*1/6 + .... + 12*1/36
= 7

The above way to solve the problem becomes difficult when there are more dice throws.
If we know about linearity of expectation, then we can quickly solve the above problem for any number of throws.
Linearity of Expectation: Let R1 and R2 be two discrete random variables on some probability space, then
E[R1 + R2] = E[R1] + E[R2]

Using the above formula, we can quickly solve the dice problem.
Expected Value of sum of 2 dice throws =
=
=
=

2*(Expected value of one dice throw)

2*(1/6 + 2/6 + .... 6/6)
2*7/2
7

Expected value of sum for n dice throws is = n * 7/2 = 3.5 * n

Some interesting facts about Linearly of Expectation:

1) Linearity of expectation holds for both dependent and independent events. On the other hand the rule E[R1R2] = E[R1]*E[R2] is true only for
independent events.
2) Linearity of expectation holds for any number of random variables on some probability space. Let R1, R2, R3, Rk be k random variables, then
E[R1 + R2 + R3 + + Rk] = E[R1] + E[R2] + E[R3] + + E[Rk]
Another example that can be easily solved with linearity of expectation:
Hat-Check Problem: Let there be group of n men where every man has one hat. The hats are redistributed and every man gets a random hat back.
What is the expected number of men that get their original hat back.
Solution: Let Ri be a random variable, the value of random variable is 1 if ith man gets the same hat back, otherwise 0.
So the expected number of men to get the right hat back is
= E[R1] + E[R2] + .. + E[Rn]
= P(R1 = 1) + P(R2 = 1) + .... + P(Rn = 1)
[Here P(Ri = 1) indicates probability that Ri is 1]
= 1/n + 1/n + ... + 1/n
= 1

So on average 1 person gets the right hat back.

Exercise:
1) Given a fair coin, what is the expected number of heads when coin is tossed n times.

2) Balls and Bins: Suppose we have m balls, labeled i = 1, , m and n bins, labeled j = 1, .. ,n. Each ball is thrown into one of the bin independently
and uniformly at random.
a) What is the expected number of balls in every bin
b) What is the expected number of empty bins.
3) Coupon Collector: Suppose there are n types of coupons in a lottery and each lot contains one coupon (with probability 1 = n each). How
many lots have to be bought (in expectation) until we have at least one coupon of each type.
See following for solution of Coupon Collector.
Expected Number of Trials until Success
Linearity of expectation is useful in algorithms. For example, expected time complexity of random algorithms like randomized quick sort is
evaluated using linearity of expectation (See this for reference).
References:
http://www.cse.iitd.ac.in/~mohanty/col106/Resources/linearity_expectation.pdf
http://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-042j-mathematics-for-computer-science-fall-2010/videolectures/lecture-22-expectation-i/

Expected Number of Trials until Success

Consider the following famous puzzle.
In a country, all families want a boy. They keep having babies till a boy is born. What is the expected ratio of boys and girls in the
country?
This puzzle can be easily solved if we know following interesting result in probability and expectation.
If probability of success is p in every trial, then expected number of trials until success is 1/p
Proof: Let R be a random variable that indicates number of trials until success.
The expected value of R is sum of following infinite series
E[R] = 1*p + 2*(1-p)*p + 3*(1-p)2*p + 4*(1-p)3*p + ........
Taking 'p' out
E[R] = p[1 + 2*(1-p) + 3*(1-p)2 + 4*(1-p)3 + .......] ---->(1)
Multiplying both sides with '(1-p)' and subtracting
(1-p)*E[R] = p[1*(1-p) + 2*(1-p)2 + 3*(1-p)3 + .......] --->(2)
Subtracting (2) from (1), we get
p*E[R] = p[1 + (1-p) + (1-p)2 + (1-p)3 + ........]
Canceling p from both sides
E[R] = [1 + (1-p) + (1-p)2 + (1-p)3 + ........]
Above is an infinite geometric progression with ratio (1-p).
Since (1-p) is less than, we can apply sum formula.
E[R] = 1/[1 - (1-p)]
= 1/p

Solution of Boys/Girls ratio puzzle:

Let us use the above result to solve the puzzle. In the given puzzle, probability of success in every trial is 1/2 (assuming that girls and boys are
equally likely).
Let p be probability of having a baby boy.
Number of kids until a baby boy is born = 1/p
= 1/(1/2)
= 2
Since expected number of kids in a family is 2,
ratio of boys and girls is 50:50.

Let us discuss another problem that uses above result.

Coupon Collector Problem:
Suppose there are n types of coupons in a lottery and each lot contains one coupon (with probability 1 = n each). How many lots have to
be bought (in expectation) until we have at least one coupon of each type.
The solution of this problem is also based on above result.
Let Xi be the number of lots bought before ith new coupon is collected.
Note that X1 is 1 as the first coupon is always a new coupon (not collected before).
Let p be probability that 2nd coupon is collected in next buy. The value of p is (n-1)/n. So the number of trials needed before 2nd new coupon is
picked is 1/p which means n/(n-1). [This is where we use above result]
Similarly, the number of trials needed before 3rd new coupon is collected is n/(n-2)
Using Linearity of expectation,
we can say that the total number of expected trials =
1 + n/(n-1) + n/(n-2) + n/(n-3) + .... + n/2 + n/1
= n[1/n + 1/(n-1) + 1/(n-2) + 1/(n-3) + ....+ 1/2 + 1/1]
= n * Hn
Here Hn is n-th Harmonic number
Since Logn <= Hn <= Logn + 1, we need to buy around
nLogn lots to collect all n coupons.

Exercise:

1) A 6 faced fair dice is thrown until a '5' is seen as result of dice throw. What is the expected number of throws?
2) What is the ratio of boys and girls in above puzzle if probability of a baby boy is 1/3?
Reference:
http://www.cse.iitd.ac.in/~mohanty/col106/Resources/linearity_expectation.pdf
http://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-042j-mathematics-for-computer-science-fall-2010/videolectures/lecture-22-expectation-i/

Randomized Algorithms | Set 1 (Introduction and Analysis)

What is a Randomized Algorithm?
An algorithm that uses random numbers to decide what to do next anywhere in its logic is called Randomized Algorithm.. For example, in
Randomized Quick Sort, we use random number to pick the next pivot (or we randomly shuffle the array). And in Kargers algorithm, we randomly
pick an edge.

How to analyse Randomized Algorithms?

Some randomized algorithms have deterministic time complexity. For example, this implementation of Kargers algorithm has time complexity as
O(E). Such algorithms are called Monte Carlo Algorithms and are easier to analyse for worst case.
On the other hand, time complexity of other randomized algorithms (other than Las Vegas) is dependent on value of random variable. Such
Randomized algorithms are called Las Vegas Algorithms. These algorithms are typically analysed for expected worst case. To compute expected
time taken in worst case, all possible values of the used random variable needs to be considered in worst case and time taken by every possible
value needs to be evaluated. Average of all evaluated times is the expected worst case time complexity. Below facts are generally helpful in
analysis os such algorithms.
Linearity of Expectation
Expected Number of Trials until Success.
For example consider below a randomized version of QuickSort.
A Central Pivot is a pivot that divides the array in such a way that one side has at-least 1/4 elements.
// Sorts an array arr[low..high]
randQuickSort(arr[], low, high)
1. If low >= high, then EXIT.
2. While pivot 'x' is not a Central Pivot.
(i) Choose uniformly at random a number from [low..high].
Let the randomly picked number number be x.
(ii) Count elements in arr[low..high] that are smaller
than arr[x]. Let this count be sc.
(iii) Count elements in arr[low..high] that are greater
than arr[x]. Let this count be gc.
(iv) Let n = (high-low+1). If sc >= n/4 and
gc >= n/4, then x is a central pivot.
3. Partition arr[low..high] around the pivot x.
4. // Recur for smaller elements
randQuickSort(arr, low, sc-1)
5. // Recur for greater elements
randQuickSort(arr, high-gc+1, high)

The important thing in our analysis is, time taken by step 2 is O(n).
How many times while loop runs before finding a central pivot?
The probability that the randomly chosen element is central pivot is 1/2.
Therefore, expected number of times the while loop runs is 2 (See this for details)
Thus, the expected time complexity of step 2 is O(n).
What is overall Time Complexity in Worst Case?
In worst case, each partition divides array such that one side has n/4 elements and other side has 3n/4 elements. The worst case height of recursion
tree is Log 3/4 n which is O(Log n).
T(n) < T(n/4) + T(3n/4) + O(n)
T(n) < 2T(3n/4) + O(n)
Solution of above recurrence is O(n Log n)

Note that the above randomized algorithm is not the best way to implement randomized Quick Sort. The idea here is to simplify the analysis as it is
simple to analyse.
Typically, randomized Quick Sort is implemented by randomly picking a pivot (no loop). Or by shuffling array elements. Expected worst case time
complexity of this algorithm is also O(n Log n), but analysis is complex, the MIT prof himself mentions same in his lecture here.
Randomized Algorithms | Set 2 (Classification and Applications)

References:
http://www.tcs.tifr.res.in/~workshop/nitrkl_igga/randomized-lecture.pdf

Randomized Algorithms | Set 2 (Classification and Applications)

We strongly recommend to refer below post as a prerequisite of this.
Randomized Algorithms | Set 1 (Introduction and Analysis)

Classification
Randomized algorithms are classified in two categories.
Las Vegas: These algorithms always produce correct or optimum result. Time complexity of these algorithms is based on a random value and
time complexity is evaluated as expected value. For example, Randomized QuickSort always sorts an input array and expected worst case time
complexity of QuickSort is O(nLogn).
Monte Carlo: Produce correct or optimum result with some probability. These algorithms have deterministic running time and it is generally easier
to find out worst case time complexity. For example this implementation of Kargers Algorithm produces minimum cut with probability greater than
or equal to 1/n2 (n is number of vertices) and has worst case time complexity as O(E).
Example to Understand Classification:
Consider a binary array where exactly half elements are 0
and half are 1. The task is to find index of any 1.

A Las Vegas algorithm for this task is to keep picking a random element until we find a 1. A Monte Carlo algorithm for the same is to keep picking
a random element until we either find 1 or we have tried maximum allowed times say k.
The Las Vegas algorithm always finds an index of 1, but time complexity is determined as expect value. The expected number of trials before
success is 2, therefore expected time complexity is O(1).
The Monte Carlo Algorithm finds a 1 with probability [1 (1/2)k]. Time complexity of Monte Carlo is O(k) which is deterministic

Applications and Scope:

Consider a tool that basically does sorting. Let the tool be used by many users and there are few users who always use tool for already
sorted array. If the tool uses simple (not randomized) QuickSort, then those few users are always going to face worst case situation. On the
other hand if the tool uses Randomized QuickSort, then there is no user that always gets worst case. Everybody gets expected O(n Log n)
time.
Randomized algorithms have huge applications in Cryptography.
Load Balancing.
Number-Theoretic Applications: Primality Testing
Data Structures: Hashing, Sorting, Searching, Order Statistics and Computational Geometry.
Algebraic identities: Polynomial and matrix identity verification. Interactive proof systems.
Mathematical programming: Faster algorithms for linear programming, Rounding linear program solutions to integer program solutions
Graph algorithms: Minimum spanning trees, shortest paths, minimum cuts.
Counting and enumeration: Matrix permanent Counting combinatorial structures.
Parallel and distributed computing: Deadlock avoidance distributed consensus.
Probabilistic existence proofs: Show that a combinatorial object arises with non-zero probability among objects drawn from a suitable
probability space.
Derandomization: First devise a randomized algorithm then argue that it can be derandomized to yield a deterministic algorithm.
Sources:
http://theory.stanford.edu/people/pragh/amstalk.pdf
https://en.wikipedia.org/wiki/Randomized_algorithm

Kargers algorithm for Minimum Cut | Set 1 (Introduction and Implementation)

Given an undirected and unweighted graph, find the smallest cut (smallest number of edges that disconnects the graph into two components).
The input graph may have parallel edges.
For example consider the following example, the smallest cut has 2 edges.

A Simple Solution use Max-Flow based s-t cut algorithm to find minimum cut. Consider every pair of vertices as source s and sink t, and call
minimum s-t cut algorithm to find the s-t cut. Return minimum of all s-t cuts. Best possible time complexity of this algorithm is O(V5) for a graph.
[How? there are total possible V2 pairs and s-t cut algorithm for one pair takes O(V*E) time and E = O(V2)].
Below is simple Kargers Algorithm for this purpose. Below Kargers algorithm can be implemented in O(E) = O(V2) time.
1) Initialize contracted graph CG as copy of original graph
2) While there are more than 2 vertices.
a) Pick a random edge (u, v) in the contracted graph.
b) Merge (or contract) u and v into a single vertex (update
the contracted graph).
c) Remove self-loops
3) Return cut represented by two vertices.

Let us understand above algorithm through the example given.

Let the first randomly picked vertex be a which connects vertices 0 and 1. We remove this edge and contract the graph (combine vertices 0 and
1). We get the following graph.

Let the next randomly picked edge be d. We remove this edge and combine vertices (0,1) and 3.

We need to remove self-loops in the graph. So we remove edge c

Now graph has two vertices, so we stop. The number of edges in the resultant graph is the cut produced by Kargers algorithm.
Kargers algorithm is a Monte Carlo algorithm and cut produced by it may not be minimum. For example, the following diagram shows
that a different order of picking random edges produces a min-cut of size 3.

Below is C++ implementation of above algorithm. The input graph is represented as a collection of edges and union-find data structure is used to
keep track of components.
// Karger's algorithm to find Minimum Cut in an
// undirected, unweighted and connected graph.
#include <stdio.h>
#include <stdlib.h>
#include <time.h>
// a structure to represent a unweighted edge in graph
struct Edge
{
int src, dest;
};
// a structure to represent a connected, undirected
// and unweighted graph as a collection of edges.
struct Graph
{
// V-> Number of vertices, E-> Number of edges
int V, E;
// graph is represented as an array of edges.
// Since the graph is undirected, the edge
// from src to dest is also edge from dest
// to src. Both are counted as 1 edge here.
Edge* edge;
};
// A structure to represent a subset for union-find
struct subset
{
int parent;
int rank;
};
// Function prototypes for union-find (These functions are defined
// after kargerMinCut() )
int find(struct subset subsets[], int i);
void Union(struct subset subsets[], int x, int y);
// A very basic implementation of Karger's randomized
// algorithm for finding the minimum cut. Please note
// that Karger's algorithm is a Monte Carlo Randomized algo
// and the cut returned by the algorithm may not be
// minimum always
int kargerMinCut(struct Graph* graph)
{
// Get data of given graph
int V = graph->V, E = graph->E;
Edge *edge = graph->edge;
// Allocate memory for creating V subsets.
struct subset *subsets = new subset[V];
// Create V subsets with single elements
for (int v = 0; v < V; ++v)
{
subsets[v].parent = v;

subsets[v].rank = 0;
}
// Initially there are V vertices in
// contracted graph
int vertices = V;
// Keep contracting vertices until there are
// 2 vertices.
while (vertices > 2)
{
// Pick a random edge
int i = rand() % E;
// Find vertices (or sets) of two corners
// of current edge
int subset1 = find(subsets, edge[i].src);
int subset2 = find(subsets, edge[i].dest);
// If two corners belong to same subset,
// then no point considering this edge
if (subset1 == subset2)
continue;
// Else contract the edge (or combine the
// corners of edge into one vertex)
else
{
printf("Contracting edge %d-%d\n",
edge[i].src, edge[i].dest);
vertices--;
Union(subsets, subset1, subset2);
}
}
// Now we have two vertices (or subsets) left in
// the contracted graph, so count the edges between
// two components and return the count.
int cutedges = 0;
for (int i=0; i<E; i++)
{
int subset1 = find(subsets, edge[i].src);
int subset2 = find(subsets, edge[i].dest);
if (subset1 != subset2)
cutedges++;
}
return cutedges;
}
// A utility function to find set of an element i
// (uses path compression technique)
int find(struct subset subsets[], int i)
{
// find root and make root as parent of i
// (path compression)
if (subsets[i].parent != i)
subsets[i].parent =
find(subsets, subsets[i].parent);
return subsets[i].parent;
}
// A function that does union of two sets of x and y
// (uses union by rank)
void Union(struct subset subsets[], int x, int y)
{
int xroot = find(subsets, x);
int yroot = find(subsets, y);
// Attach smaller rank tree under root of high
// rank tree (Union by Rank)
if (subsets[xroot].rank < subsets[yroot].rank)
subsets[xroot].parent = yroot;
else if (subsets[xroot].rank > subsets[yroot].rank)
subsets[yroot].parent = xroot;
// If ranks are same, then make one as root and
// increment its rank by one
else
{

subsets[yroot].parent = xroot;
subsets[xroot].rank++;
}
}
// Creates a graph with V vertices and E edges
struct Graph* createGraph(int V, int E)
{
Graph* graph = new Graph;
graph->V = V;
graph->E = E;
graph->edge = new Edge[E];
return graph;
}
// Driver program to test above functions
int main()
{
/* Let us create following unweighted graph
0------1
| \
|
| \ |
|
\|
2------3 */
int V = 4; // Number of vertices in graph
int E = 5; // Number of edges in graph
struct Graph* graph = createGraph(V, E);
// add edge 0-1
graph->edge[0].src = 0;
graph->edge[0].dest = 1;
// add edge 0-2
graph->edge[1].src = 0;
graph->edge[1].dest = 2;
// add edge 0-3
graph->edge[2].src = 0;
graph->edge[2].dest = 3;
// add edge 1-3
graph->edge[3].src = 1;
graph->edge[3].dest = 3;
// add edge 2-3
graph->edge[4].src = 2;
graph->edge[4].dest = 3;
// Use a different seed value for every run.
srand(time(NULL));
printf("\nCut found by Karger's randomized algo is %d\n",
kargerMinCut(graph));
return 0;
}

Output:
Contracting edge 0-2
Contracting edge 0-3
Cut found by Karger's randomized algo is 2

Note that the above program is based on outcome of a random function and may produce different output.
In this post, we have discussed simple Kargers algorithm and have seen that the algorithm doesnt always produce min-cut. The above algorithm
produces min-cut with probability greater or equal to that 1/(n2). See next post on Analysis and Applications of Kargers Algortihm, applications,
proof of this probability and improvements are discussed.
References:
http://en.wikipedia.org/wiki/Karger%27s_algorithm
https://www.youtube.com/watch?v=P0l8jMDQTEQ
https://www.cs.princeton.edu/courses/archive/fall13/cos521/lecnotes/lec2final.pdf
http://web.stanford.edu/class/archive/cs/cs161/cs161.1138/lectures/11/Small11.pdf

Kth Smallest/Largest Element in Unsorted Array | Set 2 (Expected Linear Time)

We recommend to read following post as a prerequisite of this post.
Kth Smallest/Largest Element in Unsorted Array | Set 1
Given an array and a number k where k is smaller than size of array, we need to find the kth smallest element in the given array. It is given that ll
array elements are distinct.
Examples:
Input: arr[] = {7, 10, 4, 3, 20, 15}
k = 3
Output: 7
Input: arr[] = {7, 10, 4, 3, 20, 15}
k = 4
Output: 10

We have discussed three different solutions here.

In this post method 4 is discussed which is mainly an extension of method 3 (QuickSelect) discussed in the previous post. The idea is to randomly
pick a pivot element. To implement randomized partition, we use a random function, rand() to generate index between l and r, swap the element at
randomly generated index with the last element, and finally call the standard partition process which uses last element as pivot.
Following is C++ implementation of above Randomized QuickSelect.
// C++ implementation of randomized quickSelect
#include<iostream>
#include<climits>
#include<cstdlib>
using namespace std;
int randomPartition(int arr[], int l, int r);
// This function returns k'th smallest element in arr[l..r] using
// QuickSort based method. ASSUMPTION: ALL ELEMENTS IN ARR[] ARE DISTINCT
int kthSmallest(int arr[], int l, int r, int k)
{
// If k is smaller than number of elements in array
if (k > 0 && k <= r - l + 1)
{
// Partition the array around a random element and
// get position of pivot element in sorted array
int pos = randomPartition(arr, l, r);
// If position is same as k
if (pos-l == k-1)
return arr[pos];
if (pos-l > k-1) // If position is more, recur for left subarray
return kthSmallest(arr, l, pos-1, k);
// Else recur for right subarray
return kthSmallest(arr, pos+1, r, k-pos+l-1);
}
// If k is more than number of elements in array
return INT_MAX;
}
void swap(int *a, int *b)
{
int temp = *a;
*a = *b;
*b = temp;
}
// Standard partition process of QuickSort(). It considers the last
// element as pivot and moves all smaller element to left of it and
// greater elements to right. This function is used by randomPartition()
int partition(int arr[], int l, int r)
{
int x = arr[r], i = l;
for (int j = l; j <= r - 1; j++)
{
if (arr[j] <= x)
{
swap(&arr[i], &arr[j]);

i++;
}
}
swap(&arr[i], &arr[r]);
return i;
}
// Picks a random pivot element between l and r and partitions
// arr[l..r] arount the randomly picked element using partition()
int randomPartition(int arr[], int l, int r)
{
int n = r-l+1;
int pivot = rand() % n;
swap(&arr[l + pivot], &arr[r]);
return partition(arr, l, r);
}
// Driver program to test above methods
int main()
{
int arr[] = {12, 3, 5, 7, 4, 19, 26};
int n = sizeof(arr)/sizeof(arr[0]), k = 3;
cout << "K'th smallest element is " << kthSmallest(arr, 0, n-1, k);
return 0;
}

Output:
K'th smallest element is 5

Time Complexity:
The worst case time complexity of the above solution is still O(n2). In worst case, the randomized function may always pick a corner element. The
expected time complexity of above randomized QuickSelect is ?(n), see CLRS book or MIT video lecture for proof. The assumption in the
analysis is, random number generator is equally likely to generate any number in the input range.
Sources:
MIT Video Lecture on Order Statistics, Median
Introduction to Algorithms by Clifford Stein, Thomas H. Cormen, Charles E. Leiserson, Ronald L.

Reservoir Sampling
Reservoir sampling is a family of randomized algorithms for randomly choosing k samples from a list of n items, where n is either a very large or
unknown number. Typically n is large enough that the list doesnt fit into main memory. For example, a list of search queries in Google and
Facebook.
So we are given a big array (or stream) of numbers (to simplify), and we need to write an efficient function to randomly select k numbers where 1
<= k <= n. Let the input array be stream[].
A simple solution is to create an array reservoir[] of maximum size k. One by one randomly select an item from stream[0..n-1]. If the selected
item is not previously selected, then put it in reservoir[]. To check if an item is previously selected or not, we need to search the item in
reservoir[]. The time complexity of this algorithm will be O(k^2). This can be costly if k is big. Also, this is not efficient if the input is in the form of
a stream.
It can be solved in O(n) time. The solution also suits well for input in the form of stream. The idea is similar to this post. Following are the steps.
1) Create an array reservoir[0..k-1] and copy first k items of stream[] to it.
2) Now one by one consider all items from (k+1)th item to nth item.
a) Generate a random number from 0 to i where i is index of current item in stream[]. Let the generated random number is j.
b) If j is in range 0 to k-1, replace reservoir[j] with arr[i]
Following is C implementation of the above algorithm.
// An efficient program to randomly select k items from a stream of items
#include <stdio.h>
#include <stdlib.h>
#include <time.h>
// A utility function to print an array
void printArray(int stream[], int n)
{
for (int i = 0; i < n; i++)
printf("%d ", stream[i]);
printf("\n");
}
// A function to randomly select k items from stream[0..n-1].
void selectKItems(int stream[], int n, int k)
{
int i; // index for elements in stream[]
// reservoir[] is the output array. Initialize it with
// first k elements from stream[]
int reservoir[k];
for (i = 0; i < k; i++)
reservoir[i] = stream[i];
// Use a different seed value so that we don't get
// same result each time we run this program
srand(time(NULL));
// Iterate from the (k+1)th element to nth element
for (; i < n; i++)
{
// Pick a random index from 0 to i.
int j = rand() % (i+1);
// If the randomly picked index is smaller than k, then replace
// the element present at the index with new element from stream
if (j < k)
reservoir[j] = stream[i];
}
printf("Following are k randomly selected items \n");
printArray(reservoir, k);
}
// Driver program to test above function.
int main()
{
int stream[] = {1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12};
int n = sizeof(stream)/sizeof(stream[0]);
int k = 5;
selectKItems(stream, n, k);
return 0;
}

Output:
Following are k randomly selected items
6 2 11 8 12

Time Complexity: O(n)

How does this work?
To prove that this solution works perfectly, we must prove that the probability that any item stream[i] where 0 <= i < n will be in final
reservoir[] is k/n. Let us divide the proof in two cases as first k items are treated differently.
Case 1: For last n-k stream items, i.e., for stream[i] where k <= i < n
For every such stream item stream[i], we pick a random index from 0 to i and if the picked index is one of the first k indexes, we replace the
element at picked index with stream[i]
To simplify the proof, let us first consider the last item. The probability that the last item is in final reservoir = The probability that one of the first k
indexes is picked for last item = k/n (the probability of picking one of the k items from a list of size n)
Let us now consider the second last item. The probability that the second last item is in final reservoir[] = [Probability that one of the first k
indexes is picked in iteration for stream[n-2]] X [Probability that the index picked in iteration for stream[n-1] is not same as index picked for
stream[n-2] ] = [k/(n-1)]*[(n-1)/n] = k/n.
Similarly, we can consider other items for all stream items from stream[n-1] to stream[k] and generalize the proof.
Case 2: For first k stream items, i.e., for stream[i] where 0 <= i < k
The first k items are initially copied to reservoir[] and may be removed later in iterations for stream[k] to stream[n].
The probability that an item from stream[0..k-1] is in final array = Probability that the item is not picked when items stream[k], stream[k+1], .
stream[n-1] are considered = [k/(k+1)] x [(k+1)/(k+2)] x [(k+2)/(k+3)] x x [(n-1)/n] = k/n
References:
http://en.wikipedia.org/wiki/Reservoir_sampling

Shuffle a given array

Given an array, write a program to generate a random permutation of array elements. This question is also asked as shuffle a deck of cards or
randomize a given array.

Let the given array be arr[]. A simple solution is to create an auxiliary array temp[] which is initially a copy of arr[]. Randomly select an element
from temp[], copy the randomly selected element to arr[0] and remove the selected element from temp[]. Repeat the same process n times and
keep copying elements to arr[1], arr[2], . The time complexity of this solution will be O(n^2).
FisherYates shuffle Algorithm works in O(n) time complexity. The assumption here is, we are given a function rand() that generates random
number in O(1) time.
The idea is to start from the last element, swap it with a randomly selected element from the whole array (including last). Now consider the array
from 0 to n-2 (size reduced by 1), and repeat the process till we hit the first element.
Following is the detailed algorithm
To shuffle an array a of n elements (indices 0..n-1):
for i from n - 1 downto 1 do
j = random integer with 0 <= j <= i
exchange a[j] and a[i]

Following is C++ implementation of this algorithm.

// C Program to shuffle a given array
#include <stdio.h>
#include <stdlib.h>
#include <time.h>
// A utility function to swap to integers
void swap (int *a, int *b)
{
int temp = *a;
*a = *b;
*b = temp;
}
// A utility function to print an array
void printArray (int arr[], int n)
{
for (int i = 0; i < n; i++)
printf("%d ", arr[i]);
printf("\n");
}
// A function to generate a random permutation of arr[]
void randomize ( int arr[], int n )
{
// Use a different seed value so that we don't get same
// result each time we run this program
srand ( time(NULL) );
// Start from the last element and swap one by one. We don't
// need to run for the first element that's why i > 0
for (int i = n-1; i > 0; i--)
{
// Pick a random index from 0 to i
int j = rand() % (i+1);
// Swap arr[i] with the element at random index
swap(&arr[i], &arr[j]);
}
}
// Driver program to test above function.
int main()
{
int arr[] = {1, 2, 3, 4, 5, 6, 7, 8};
int n = sizeof(arr)/ sizeof(arr[0]);

randomize (arr, n);

printArray(arr, n);
return 0;
}

Output:
7 8 4 6 3 1 2 5

The above function assumes that rand() generates a random number.

Time Complexity: O(n), assuming that the function rand() takes O(1) time.
How does this work?
The probability that ith element (including the last one) goes to last position is 1/n, because we randomly pick an element in first iteration.
The probability that ith element goes to second last position can be proved to be 1/n by dividing it in two cases.
Case 1: i = n-1 (index of last element):
The probability of last element going to second last position is = (probability that last element doesn't stay at its original position) x (probability that
the index picked in previous step is picked again so that the last element is swapped)
So the probability = ((n-1)/n) x (1/(n-1)) = 1/n
Case 2: 0 < i < n-1 (index of non-last):
The probability of ith element going to second position = (probability that ith element is not picked in previous iteration) x (probability that ith
element is picked in this iteration)
So the probability = ((n-1)/n) x (1/(n-1)) = 1/n
We can easily generalize above proof for any other position.

Select a Random Node from a Singly Linked List

Given a singly linked list, select a random node from linked list (the probability of picking a node should be 1/N if there are N nodes in list). You
are given a random number generator.
Below is a Simple Solution
1) Count number of nodes by traversing the list.
2) Traverse the list again and select every node with probability 1/N. The selection can be done by generating a random number from 0 to N-i for
ith node, and selecting the ith node node only if generated number is equal to 0 (or any other fixed number from 0 to N-i).
We get uniform probabilities with above schemes.
i = 1, probability of selecting first node = 1/N
i = 2, probability of selecting second node =
[probability that first node is not selected] *
[probability that second node is selected]
= ((N-1)/N)* 1/(N-1)
= 1/N

Similarly, probabilities of other selecting other nodes is 1/N

The above solution requires two traversals of linked list.
How to select a random node with only one traversal allowed?
The idea is to use Reservoir Sampling. Following are the steps. This is a simpler version of Reservoir Sampling as we need to select only one key
instead of k keys.
(1) Initialize result as first node
result = head->key
(2) Initialize n = 2
(3) Now one by one consider all nodes from 2nd node onward.
(3.a) Generate a random number from 0 to n-1.
Let the generated random number is j.
(3.b) If j is equal to 0 (we could choose other fixed number
between 0 to n-1), then replace result with current node.
(3.c) n = n+1
(3.d) current = current->next

Below is C implementation of above algorithm.

/* C program to randomly select a node from a singly
linked list */
#include<stdio.h>
#include<stdlib.h>
#include <time.h>
/* Link list node */
struct node
{
int key;
struct node* next;
};
// A reservoir sampling based function to print a
// random node from a linked list
void printRandom(struct node *head)
{
// IF list is empty
if (head == NULL)
return;
// Use a different seed value so that we don't get
// same result each time we run this program
srand(time(NULL));
// Initialize result as first node
int result = head->key;
// Iterate from the (k+1)th element to nth element
struct node *current = head;
int n;
for (n=2; current!=NULL; n++)
{
// change result with probability 1/n
if (rand() % n == 0)
result = current->key;

// Move to next node

current = current->next;
}
printf("Randomly selected key is %d\n", result);
}
/* BELOW FUNCTIONS ARE JUST UTILITY TO TEST */
/* A utility function to create a new node */
struct node *newNode(int new_key)
{
/* allocate node */
struct node* new_node =
(struct node*) malloc(sizeof(struct node));
/* put in the key */
new_node->key = new_key;
new_node->next = NULL;
return new_node;
}
/* A utility function to insert a node at the beginning
of linked list */
void push(struct node** head_ref, int new_key)
{
/* allocate node */
struct node* new_node = new node;
/* put in the key */
new_node->key = new_key;
/* link the old list off the new node */
new_node->next = (*head_ref);
/* move the head to point to the new node */
(*head_ref)
= new_node;
}
// Driver program to test above functions
int main()
{
struct node *head = NULL;
push(&head, 5);
push(&head, 20);
push(&head, 4);
push(&head, 3);
push(&head, 30);
printRandom(head);
return 0;
}
Randomly selected key is 4

Note that the above program is based on outcome of a random function and may produce different output.
How does this work?
Let there be total N nodes in list. It is easier to understand from last node.
The probability that last node is result simply 1/N [For last or Nth node, we generate a random number between 0 to N-1 and make last node as
result if the generated number is 0 (or any other fixed number]
The probability that second last node is result should also be 1/N.
The probability that the second last node is result
= [Probability that the second last node replaces result] X
[Probability that the last node doesn't replace the result]
= [1 / (N-1)] * [(N-1)/N]
= 1/N

Similarly we can show probability for 3rd last node and other nodes.

Commonly Asked Algorithm Interview Questions | Set 1

What is an algorithm?
Informally, an algorithm is any well-defined computational procedure that takes some value, or set of values, as input and produces some value, or
set of values, as output. An algorithm is thus a sequence of computational steps that transform the input into the output. (Source: Introduction to
Algorithms 3rd Edition by CLRS)
What is time complexity of Binary Search?
Time complexity of binary search is O(Logn). See Binary Search for more details.
Can Binary Search be used for linked lists?
Since random access is not allowed in linked list, we cannot reach the middle element in O(1) time. Therefore Binary Search is not possible for
linked lists. There are other ways though, refer Skip List for example.
How to find if two given rectangles overlap?
Two rectangles do not overlap if one of the following conditions is true.
1) One rectangle is above top edge of other rectangle.
2) One rectangle is on left side of left edge of other rectangle.
See Find if two rectangles overlap for more details.
How to find angle between hour and minute hands at a given time?
The idea is to take a reference point as 12. Find the angle moved by hour and minute hands, subtract the two angles to find the angle between
them. See angle between hour hand and minute hand for more details
When does the worst case of QuickSort occur?
In quickSort, we select a pivot element, then partition the given array around the pivot element by placing pivot element at its correct position in
sorted array.
The worst case of quickSort occurs when one part after partition contains all elements and other part is empty. For example, if the input array is
sorted and if last or first element is chosen as a pivot, then the worst occurs. See http://geeksquiz.com/quick-sort/ for more details.
A sorted array is rotated at some unknown point, how to efficiently search an element in it.
A simple approach is linear search, but we can search in O(Logn) time using Binary Search. See Search an element in a sorted and pivoted array
for more details.
Other variations of this problem like find the minimum element or maximum element in a sorted and rotated array.
Given a big string of characters, how to efficiently find the first unique character in it?
The efficient solution is to use character as an index in a count array. Traverse the given string and store index of first occurrence of every
character, also store count of occurrences. Then traverse the count array and find the smallest index with count as 1. See find the first unique
character for more details.
How to count inversions in a sorted array?
Two elements arr[i] and arr[j] in an array arr[] form an inversion if a[i] > a[j] and i < j. How to count all inversions in an unsorted array. See Count
Inversions in an array for all approaches.
Given a big array, how to efficiently find kth largest element in it?
There can be many solutions for this. The best solution is to use min heap. We Build a Min Heap MH of the first k elements. For each element,
after the kth element (arr[k] to arr[n-1]), compare it with root of MH, if the element is greater than the root then make it root and call heapify for
MH, Else ignore it. Finally, MH has k largest elements and root of the MH is the kth largest element. See k largest(or smallest) elements for more
details.
Given an array of size n with range of numbers from 1 to n+1. The array doesnt contain any duplicate, one number is missing, find
the missing number.
There can be many ways to solve it. The best among is to use XOR. See Find the missing number for details. There are many variations of this
problem like find the two repeating numbers, find a missing and a repeating number, etc.
How to write an efficient method to calculate x raise to the power n?
The idea is to use divide an conquer here to do it in O(Logn) time. See Write a C program to calculate pow(x,n) for more details.
Given an input string and a dictionary of words, find out if the input string can be segmented into a space-separated sequence of
dictionary words.
The idea is to use Dynamic Programming. See Word Break Problem for more details.
Given a row of n coins of values v1 . . . vn, where n is even. We play a game against an opponent by alternating turns. In each turn,
a player selects either the first or last coin from the row, removes it from the row permanently, and receives the value of the coin.
Determine the maximum possible amount of money we can definitely win if we move first.
This is also a Dynamic Programming Question. See Optimal Strategy for a Game for more details.

You are given an array of sorted words in an arbitrary language, you need to find order (or precedence of characters) in the
language. For example if the given arrays is {baa, abcd, abca, cab, cad}, then order of characters is b, d, a, c. Note that words are sorted and in
the given language baa comes before abcd, therefore b is before a in output. Similarly we can find other orders.
This can be solved using two steps: First create a graph by processing given set of words, then do topological sorting of the created graph, See this
for more details.
We will soon be adding more questions.
You may also like to see Commonly Asked Data Structure Interview Questions | Set 1

Given a matrix of O and X, find the largest subsquare surrounded by X

Given a matrix where every element is either O or X, find the largest subsquare surrounded by X.
In the below article, it is assumed that the given matrix is also square matrix. The code given below can be easily extended for rectangular matrices.
Examples:
Input: mat[N][N] = { {'X', 'O', 'X',
{'X', 'X', 'X',
{'X', 'X', 'O',
{'X', 'X', 'X',
{'X', 'X', 'X',
};
Output: 3
The square submatrix starting at (1,
submatrix surrounded by 'X'

'X',
'X',
'X',
'X',
'O',

Input: mat[M][N] = { {'X', 'O', 'X',

{'X', 'O', 'X',
{'X', 'X', 'X',
{'X', 'X', 'X',
{'X', 'X', 'X',
};
Output: 4
The square submatrix starting at (0,
submatrix surrounded by 'X'

'X',
'X',
'O',
'X',
'O',

'X'},
'X'},
'O'},
'X'},
'O'},

1) is the largest
'X',
'O',
'O',
'X',
'X',

'X'},
'X'},
'X'},
'X'},
'O'},

2) is the largest

A Simple Solution is to consider every square submatrix and check whether is has all corner edges filled with X. The time complexity of this
solution is O(N4).
We can solve this problem in O(N3) time using extra space. The idea is to create two auxiliary arrays hor[N][N] and ver[N][N]. The value
stored in hor[i][j] is the number of horizontal continuous X characters till mat[i][j] in mat[][]. Similarly, the value stored in ver[i][j] is the number of
vertical continuous X characters till mat[i][j] in mat[][]. Following is an example.
mat[6][6] = X
X
X
O
X
O

O
O
X
X
X
O

X
X
X
X
X
X

X
X
O
X
O
O

X
O
O
X
X
O

X
X
X
X
O
O

hor[6][6] = 1
1
1
0
1
0

0
0
2
1
2
0

1
1
3
2
3
1

2
2
0
3
0
0

3
0
0
4
1
0

4
1
1
5
0
0

ver[6][6] = 1
2
3
0
1
0

0
0
1
2
3
0

1
2
3
4
5
6

1
2
0
1
0
0

1
0
0
1
2
0

1
2
3
4
0
0

Once we have filled values in hor[][] and ver[][], we start from the bottommost-rightmost corner of matrix and move toward the leftmost-topmost
in row by row manner. For every visited entry mat[i][j], we compare the values of hor[i][j] and ver[i][j], and pick the smaller of two as we need a
square. Let the smaller of two be small. After picking smaller of two, we check if both ver[][] and hor[][] for left and up edges respectively. If they
have entries for the same, then we found a subsquare. Otherwise we try for small-1.
Below is C++ implementation of the above idea.
// A C++ program to find the largest subsquare
// surrounded by 'X' in a given matrix of 'O' and 'X'
#include<iostream>
using namespace std;
// Size of given matrix is N X N
#define N 6
// A utility function to find minimum of two numbers
int getMin(int x, int y) { return (x<y)? x: y; }
// Returns size of maximum size subsquare matrix
// surrounded by 'X'
int findSubSquare(int mat[][N])

{
int max = 1; // Initialize result
// Initialize the left-top value in hor[][] and ver[][]
int hor[N][N], ver[N][N];
hor[0][0] = ver[0][0] = (mat[0][0] == 'X');
// Fill values in hor[][] and ver[][]
for (int i=0; i<N; i++)
{
for (int j=0; j<N; j++)
{
if (mat[i][j] == 'O')
ver[i][j] = hor[i][j] = 0;
else
{
hor[i][j] = (j==0)? 1: hor[i][j-1] + 1;
ver[i][j] = (i==0)? 1: ver[i-1][j] + 1;
}
}
}
// Start from the rightmost-bottommost corner element and find
// the largest ssubsquare with the help of hor[][] and ver[][]
for (int i = N-1; i>=1; i--)
{
for (int j = N-1; j>=1; j--)
{
// Find smaller of values in hor[][] and ver[][]
// A Square can only be made by taking smaller
// value
int small = getMin(hor[i][j], ver[i][j]);
// At this point, we are sure that there is a right
// vertical line and bottom horizontal line of length
// at least 'small'.
// We found a bigger square if following conditions
// are met:
// 1)If side of square is greater than max.
// 2)There is a left vertical line of length >= 'small'
// 3)There is a top horizontal line of length >= 'small'
while (small > max)
{
if (ver[i][j-small+1] >= small &&
hor[i-small+1][j] >= small)
{
max = small;
}
small--;
}
}
}
return max;
}
// Driver program to test above
int main()
{
int mat[][N] = {{'X', 'O',
{'X', 'O',
{'X', 'X',
{'O', 'X',
{'X', 'X',
{'O', 'O',
};
cout << findSubSquare(mat);
return 0;
}

Output:
4

function
'X',
'X',
'X',
'X',
'X',
'X',

'X',
'X',
'O',
'X',
'O',
'O',

'X',
'O',
'O',
'X',
'X',
'O',

'X'},
'X'},
'X'},
'X'},
'O'},
'O'},

Nuts & Bolts Problem (Lock & Key problem)

Given a set of n nuts of different sizes and n bolts of different sizes. There is a one-one mapping between nuts and bolts. Match nuts and bolts
efficiently.
Constraint: Comparison of a nut to another nut or a bolt to another bolt is not allowed. It means nut can only be compared with bolt and bolt can
only be compared with nut to see which one is bigger/smaller.
Other way of asking this problem is, given a box with locks and keys where one lock can be opened by one key in the box. We need to match the
pair.
Brute force Way: Start with the first bolt and compare it with each nut until we find a match. In the worst case we require n comparisons. Doing
this for all bolts gives us O(n^2) complexity.
Quick Sort Way: We can use quick sort technique to solve this. We represent nuts and bolts in character array for understanding the logic.
Nuts represented as array of character
char nuts[] = {@, #, $, %, ^, &}
Bolts represented as array of character
char bolts[] = {$, %, &, ^, @, #}
This algorithm first performs a partition by picking last element of bolts array as pivot, rearrange the array of nuts and returns the partition index i
such that all nuts smaller than nuts[i] are on the left side and all nuts greater than nuts[i] are on the right side. Next using the nuts[i] we can partition
the array of bolts. Partitioning operations can easily be implemented in O(n). This operation also makes nuts and bolts array nicely partitioned.
Now we apply this partitioning recursively on the left and right sub-array of nuts and bolts.
As we apply partitioning on nuts and bolts both so the total time complexity will be ?(2*nlogn) = ?(nlogn) on average.
Here for the sake of simplicity we have chosen last element always as pivot. We can do randomized quick sort too.
A Java based implementation of idea is below:
// Java program to solve nut and bolt problem using Quick Sort
public class NutsAndBoltsMatch
{
//Driver method
public static void main(String[] args)
{
// Nuts and bolts are represented as array of characters
char nuts[] = {'@', '#', '$', '%', '^', '&'};
char bolts[] = {'$', '%', '&', '^', '@', '#'};
// Method based on quick sort which matches nuts and bolts
matchPairs(nuts, bolts, 0, 5);
System.out.println("Matched nuts and bolts are : ");
printArray(nuts);
printArray(bolts);
}
// Method to print the array
private static void printArray(char[] arr) {
for (char ch : arr){
System.out.print(ch + " ");
}
System.out.print("\n");
}
// Method which works just like quick sort
private static void matchPairs(char[] nuts, char[] bolts, int low,
int high)
{
if (low < high)
{
// Choose last character of bolts array for nuts partition.
int pivot = partition(nuts, low, high, bolts[high]);
// Now using the partition of nuts choose that for bolts
// partition.
partition(bolts, low, high, nuts[pivot]);
// Recur for [low...pivot-1] & [pivot+1...high] for nuts and
// bolts array.
matchPairs(nuts, bolts, low, pivot-1);
matchPairs(nuts, bolts, pivot+1, high);

}
}
// Similar to standard partition method. Here we pass the pivot element
// too instead of choosing it inside the method.
private static int partition(char[] arr, int low, int high, char pivot)
{
int i = low;
char temp1, temp2;
for (int j = low; j < high; j++)
{
if (arr[j] < pivot){
temp1 = arr[i];
arr[i] = arr[j];
arr[j] = temp1;
i++;
} else if(arr[j] == pivot){
temp1 = arr[j];
arr[j] = arr[high];
arr[high] = temp1;
j--;
}
}
temp2 = arr[i];
arr[i] = arr[high];
arr[high] = temp2;
// Return the partition index of an array based on the pivot
// element of other array.
return i;
}
}

Output:
Matched nuts and bolts are :
# $ % & @ ^
# $ % & @ ^

Flood fill Algorithm how to implement fill() in paint?

In MS-Paint, when we take the brush to a pixel and click, the color of the region of that pixel is replaced with a new selected color. Following is
the problem statement to do this task.
Given a 2D screen, location of a pixel in the screen and a color, replace color of the given pixel and all adjacent same colored pixels with the given
color.
Example:
Input:
screen[M][N] = {{1, 1, 1, 1, 1, 1, 1, 1},
{1, 1, 1, 1, 1, 1, 0, 0},
{1, 0, 0, 1, 1, 0, 1, 1},
{1, 2, 2, 2, 2, 0, 1, 0},
{1, 1, 1, 2, 2, 0, 1, 0},
{1, 1, 1, 2, 2, 2, 2, 0},
{1, 1, 1, 1, 1, 2, 1, 1},
{1, 1, 1, 1, 1, 2, 2, 1},
};
x = 4, y = 4, newColor = 3
The values in the given 2D screen indicate colors of the pixels.
x and y are coordinates of the brush, newColor is the color that
should replace the previous color on screen[x][y] and all surrounding
pixels with same color.
Output:
Screen should be changed to following.
screen[M][N] = {{1, 1, 1, 1, 1, 1, 1, 1},
{1, 1, 1, 1, 1, 1, 0, 0},
{1, 0, 0, 1, 1, 0, 1, 1},
{1, 3, 3, 3, 3, 0, 1, 0},
{1, 1, 1, 3, 3, 0, 1, 0},
{1, 1, 1, 3, 3, 3, 3, 0},
{1, 1, 1, 1, 1, 3, 1, 1},
{1, 1, 1, 1, 1, 3, 3, 1},
};

Flood Fill Algorithm:

The idea is simple, we first replace the color of current pixel, then recur for 4 surrounding points. The following is detailed algorithm.
// A recursive function to replace previous color 'prevC' at '(x, y)'
// and all surrounding pixels of (x, y) with new color 'newC' and
floodFil(screen[M][N], x, y, prevC, newC)
1) If x or y is outside the screen, then return.
2) If color of screen[x][y] is not same as prevC, then return
3) Recur for north, south, east and west.
floodFillUtil(screen, x+1, y, prevC, newC);
floodFillUtil(screen, x-1, y, prevC, newC);
floodFillUtil(screen, x, y+1, prevC, newC);
floodFillUtil(screen, x, y-1, prevC, newC);

The following is C++ implementation of above algorithm.

// A C++ program to implement flood fill algorithm
#include<iostream>
using namespace std;
// Dimentions of paint screen
#define M 8
#define N 8
// A recursive function to replace previous color 'prevC' at '(x, y)'
// and all surrounding pixels of (x, y) with new color 'newC' and
void floodFillUtil(int screen[][N], int x, int y, int prevC, int newC)
{
// Base cases
if (x < 0 || x >= M || y < 0 || y >= N)
return;
if (screen[x][y] != prevC)
return;
// Replace the color at (x, y)
screen[x][y] = newC;
// Recur for north, east, south and west
floodFillUtil(screen, x+1, y, prevC, newC);
floodFillUtil(screen, x-1, y, prevC, newC);
floodFillUtil(screen, x, y+1, prevC, newC);

floodFillUtil(screen, x, y-1, prevC, newC);

}
// It mainly finds the previous color on (x, y) and
// calls floodFillUtil()
void floodFill(int screen[][N], int x, int y, int newC)
{
int prevC = screen[x][y];
floodFillUtil(screen, x, y, prevC, newC);
}
// Driver program to test above function
int main()
{
int screen[M][N] = {{1, 1, 1, 1, 1, 1, 1, 1},
{1, 1, 1, 1, 1, 1, 0, 0},
{1, 0, 0, 1, 1, 0, 1, 1},
{1, 2, 2, 2, 2, 0, 1, 0},
{1, 1, 1, 2, 2, 0, 1, 0},
{1, 1, 1, 2, 2, 2, 2, 0},
{1, 1, 1, 1, 1, 2, 1, 1},
{1, 1, 1, 1, 1, 2, 2, 1},
};
int x = 4, y = 4, newC = 3;
floodFill(screen, x, y, newC);
cout << "Updated screen after call to floodFill: \n";
for (int i=0; i<M; i++)
{
for (int j=0; j<N; j++)
cout << screen[i][j] << " ";
cout << endl;
}
}

Output:
Updated
1 1 1 1
1 1 1 1
1 0 0 1
1 3 3 3
1 1 1 3
1 1 1 3
1 1 1 1
1 1 1 1

screen after call to floodFill:

1 1 1 1
1 1 0 0
1 0 1 1
3 0 1 0
3 0 1 0
3 3 3 0
1 3 1 1
1 3 3 1

References:
http://en.wikipedia.org/wiki/Flood_fill

Given n appointments, find all conflicting appointments

Given n appointments, find all conflicting appointments.
Examples:
Input: appointments[] = { {1, 5} {3, 7}, {2, 6}, {10, 15}, {5, 6}, {4, 100}}
Output: Following are conflicting intervals
[3,7] Conflicts with [1,5]
[2,6] Conflicts with [1,5]
[5,6] Conflicts with [3,7]
[4,100] Conflicts with [1,5]

An appointment is conflicting, if it conflicts with any of the previous appointments in array.

A Simple Solution is to one by one process all appointments from second appointment to last. For every appointment i, check if it conflicts with i1, i-2, 0. The time complexity of this method is O(n2).
We can use Interval Tree to solve this problem in O(nLogn) time. Following is detailed algorithm.
1) Create an Interval Tree, initially with the first appointment.
2) Do following for all other appointments starting from the second one.
a) Check if the current appointment conflicts with any of the existing
appointments in Interval Tree. If conflicts, then print the current
appointment. This step can be done O(Logn) time.
b) Insert the current appointment in Interval Tree. This step also can
be done O(Logn) time.

Following is C++ implementation of above idea.

// C++ program to print all conflicting appointments in a
// given set of appointments
#include <iostream>
using namespace std;
// Structure to represent an interval
struct Interval
{
int low, high;
};
// Structure to represent a node in Interval Search Tree
struct ITNode
{
Interval *i; // 'i' could also be a normal variable
int max;
ITNode *left, *right;
};
// A utility function to create a new Interval Search Tree Node
ITNode * newNode(Interval i)
{
ITNode *temp = new ITNode;
temp->i = new Interval(i);
temp->max = i.high;
temp->left = temp->right = NULL;
};
// A utility function to insert a new Interval Search Tree
// Node. This is similar to BST Insert. Here the low value
// of interval is used tomaintain BST property
ITNode *insert(ITNode *root, Interval i)
{
// Base case: Tree is empty, new node becomes root
if (root == NULL)
return newNode(i);
// Get low value of interval at root
int l = root->i->low;
// If root's low value is smaller, then new interval
// goes to left subtree
if (i.low < l)
root->left = insert(root->left, i);
// Else, new node goes to right subtree.
else
root->right = insert(root->right, i);

// Update the max value of this ancestor if needed

if (root->max < i.high)
root->max = i.high;
return root;
}
// A utility function to check if given two intervals overlap
bool doOVerlap(Interval i1, Interval i2)
{
if (i1.low < i2.high && i2.low < i1.high)
return true;
return false;
}
// The main function that searches a given interval i
// in a given Interval Tree.
Interval *overlapSearch(ITNode *root, Interval i)
{
// Base Case, tree is empty
if (root == NULL) return NULL;
// If given interval overlaps with root
if (doOVerlap(*(root->i), i))
return root->i;
//
//
//
if

If left child of root is present and max of left child

is greater than or equal to given interval, then i may
overlap with an interval is left subtree
(root->left != NULL && root->left->max >= i.low)
return overlapSearch(root->left, i);

// Else interval can only overlap with right subtree

return overlapSearch(root->right, i);
}
// This function prints all conflicting appointments in a given
// array of apointments.
void printConflicting(Interval appt[], int n)
{
// Create an empty Interval Search Tree, add first
// appointment
ITNode *root = NULL;
root = insert(root, appt[0]);
// Process rest of the intervals
for (int i=1; i<n; i++)
{
// If current appointment conflicts with any of the
// existing intervals, print it
Interval *res = overlapSearch(root, appt[i]);
if (res != NULL)
cout << "[" << appt[i].low << "," << appt[i].high
<< "] Conflicts with [" << res->low << ","
<< res->high << "]\n";
// Insert this appointment
root = insert(root, appt[i]);
}
}
// Driver program to test above functions
int main()
{
// Let us create interval tree shown in above figure
Interval appt[] = { {1, 5}, {3, 7}, {2, 6}, {10, 15},
{5, 6}, {4, 100}};
int n = sizeof(appt)/sizeof(appt[0]);
cout << "Following are conflicting intervals\n";
printConflicting(appt, n);
return 0;
}

Output:
Following are conflicting intervals
[3,7] Conflicts with [1,5]
[2,6] Conflicts with [1,5]
[5,6] Conflicts with [3,7]

[4,100] Conflicts with [1,5]

Note that the above implementation uses simple Binary Search Tree insert operations. Therefore, time complexity of the above implementation is
more than O(nLogn). We can use Red-Black Tree or AVL Tree balancing techniques to make the above implementation O(nLogn).

Check a given sentence for a given set of simple grammer rules

A simple sentence if syntactically correct if it fulfills given rules. The following are given rules.
1. Sentence must start with a Uppercase character (e.g. Noun/ I/ We/ He etc.)
2. Then lowercase character follows.
3. There must be spaces between words.
4. Then the sentence must end with a full stop(.) after a word.
5. Two continuous spaces are not allowed.
6. Two continuous upper case characters are not allowed.
7. However the sentence can end after an upper case character.
Examples:
Correct sentences "My name is Ram."
"The vertex is S."
"I am single."
"I love Geeksquiz and Geeksforgeeks."
Incorrect sentence "My name is KG."
"I lovE cinema."
"GeeksQuiz. is a quiz site."
" You are my friend."
"I love cinema"

Question: Given a sentence, validate the given sentence for above given rules.
Algorithm :
1. Check for the corner cases
..1.a) Check if the first character is uppercase or not in the sentence.
..1.b) Check if the last character is a full stop or not.
2. For rest of the string, this problem could be solved by following a state diagram. Please refer to the below state diagram for that.

3. We need to maintain previous and current state of different characters in the string. Based on that we can always validate the sentence of every
character traversed.
A C based implementation is below. (By the way this sentence is also correct according to the rule and code)

C++
// C program to validate a given sentence for a set of rules
#include<stdio.h>
#include<string.h>
#include<stdbool.h>
// Method to check a given sentence for given rules
bool checkSentence(char str[])
{
// Calculate the length of the string.
int len = strlen(str);
// Check that the first character lies in [A-Z].
// Otherwise return false.
if (str[0] < 'A' || str[0] > 'Z')

return false;
//If the last character is not a full stop(.) no
//need to check further.
if (str[len - 1] != '.')
return false;
// Maintain 2 states. Previous and current state based
// on which vertex state you are. Initialise both with
// 0 = start state.
int prev_state = 0, curr_state = 0;
//Keep the index to the next character in the string.
int index = 1;
//Loop to go over the string.
while (str[index])
{
// Set states according to the input characters in the
// string and the rule defined in the description.
// If current character is [A-Z]. Set current state as 0.
if (str[index] >= 'A' && str[index] <= 'Z')
curr_state = 0;
// If current character is a space. Set current state as 1.
else if (str[index] == ' ')
curr_state = 1;
// If current character is [a-z]. Set current state as 2.
else if (str[index] >= 'a' && str[index] <= 'z')
curr_state = 2;
// If current state is a dot(.). Set current state as 3.
else if (str[index] == '.')
curr_state = 3;
// Validates all current state with previous state for the
// rules in the description of the problem.
if (prev_state == curr_state && curr_state != 2)
return false;
if (prev_state == 2 && curr_state == 0)
return false;
//
//
//
if

If we have reached last state and previous state is not 1,

then check next character. If next character is '\0', then
return true, else false
(curr_state == 3 && prev_state != 1)
return (str[index + 1] == '\0');

index++;
// Set previous state as current state before going over
// to the next character.
prev_state = curr_state;
}
return false;
}
// Driver program
int main()
{
char *str[] = { "I love cinema.", "The vertex is S.",
"I am single.", "My name is KG.",
"I lovE cinema.", "GeeksQuiz. is a quiz site.",
"I love Geeksquiz and Geeksforgeeks.",
" You are my friend.", "I love cinema" };
int str_size = sizeof(str) / sizeof(str[0]);
int i = 0;
for (i = 0; i < str_size; i++)
checkSentence(str[i])? printf("\"%s\" is correct \n", str[i]):
printf("\"%s\" is incorrect \n", str[i]);
return 0;
}

Python
# Python program to validate a given sentence for a set of rules

# Method to check a given sentence for given rules

def checkSentence(string):
# Calculate the length of the string.
length = len(string)
# Check that the first character lies in [A-Z].
# Otherwise return false.
if string[0] < 'A' or string[0] > 'Z':
return False
# If the last character is not a full stop(.) no
# need to check further.
if string[length-1] != '.':
return False
# Maintain 2 states. Previous and current state based
# on which vertex state you are. Initialise both with
# 0 = start state.
prev_state = 0
curr_state = 0
# Keep the index to the next character in the string.
index = 1
# Loop to go over the string.
while (string[index]):
# Set states according to the input characters in the
# string and the rule defined in the description.
# If current character is [A-Z]. Set current state as 0.
if string[index] >= 'A' and string[index] <= 'Z':
curr_state = 0
# If current character is a space. Set current state as 1.
elif string[index] == ' ':
curr_state = 1
# If current character is a space. Set current state as 2.
elif string[index] >= 'a' and string[index] <= 'z':
curr_state = 2
# If current character is a space. Set current state as 3.
elif string[index] == '.':
curr_state = 3
# Validates all current state with previous state for the
# rules in the description of the problem.
if prev_state == curr_state and curr_state != 2:
return False
# If we have reached last state and previous state is not 1,
# then check next character. If next character is '\0', then
# return true, else false
if prev_state == 2 and curr_state == 0:
return False
# Set previous state as current state before going over
# to the next character.
if curr_state == 3 and prev_state != 1:
return True
index += 1
prev_state = curr_state
return False
# Driver program
string = ["I love cinema.", "The vertex is S.",
"I am single.", "My name is KG.",
"I lovE cinema.", "GeeksQuiz. is a quiz site.",
"I love Geeksquiz and Geeksforgeeks.",
" You are my friend.", "I love cinema"]
string_size = len(string)
for i in xrange(string_size):
if checkSentence(string[i]):
print "\"" + string[i] + "\" is correct"
else:
print "\"" + string[i] + "\" is incorrect"
# This code is contributed by BHAVYA JAIN

"I love cinema." is correct

"The vertex is S." is correct
"I am single." is correct
"My name is KG." is incorrect
"I lovE cinema." is incorrect
"GeeksQuiz. is a quiz site." is incorrect
"I love Geeksquiz and Geeksforgeeks." is correct
" You are my friend." is incorrect
"I love cinema" is incorrect

Time complexity O(n), worst case as we have to traverse the full sentence where n is the length of the sentence.
Auxiliary space O(1)

Find Index of 0 to be replaced with 1 to get longest continuous sequence of 1s in a

binary array
Given an array of 0s and 1s, find the position of 0 to be replaced with 1 to get longest continuous sequence of 1s. Expected time complexity is
O(n) and auxiliary space is O(1).
Example:
Input:
arr[] = {1, 1, 0, 0, 1, 0, 1, 1, 1, 0, 1, 1, 1}
Output:
Index 9
Assuming array index starts from 0, replacing 0 with 1 at index 9 causes
the maximum continuous sequence of 1s.
Input:
arr[] = {1, 1, 1, 1, 0}
Output:
Index 4

A Simple Solution is to traverse the array, for every 0, count the number of 1s on both sides of it. Keep track of maximum count for any 0.
Finally return index of the 0 with maximum number of 1s around it. The time complexity of this solution is O(n2).
Using an Efficient Solution, the problem can solved in O(n) time. The idea is to keep track of three indexes, current index (curr), previous zero
index (prev_zero) and previous to previous zero index (prev_prev_zero). Traverse the array, if current element is 0, calculate the difference
between curr and prev_prev_zero (This difference minus one is the number of 1s around the prev_zero). If the difference between curr and
prev_prev_zero is more than maximum so far, then update the maximum. Finally return index of the prev_zero with maximum difference.
Following are C++ and Java implementations of the above algorithm.

C++
// C++ program to find Index of 0 to be replaced with 1 to get
// longest continuous sequence of 1s in a binary array
#include<iostream>
using namespace std;
// Returns index of 0 to be replaced with 1 to get longest
// continuous sequence of 1s. If there is no 0 in array, then
// it returns -1.
int maxOnesIndex(bool arr[], int n)
{
int max_count = 0; // for maximum number of 1 around a zero
int max_index; // for storing result
int prev_zero = -1; // index of previous zero
int prev_prev_zero = -1; // index of previous to previous zero
// Traverse the input array
for (int curr=0; curr<n; ++curr)
{
// If current element is 0, then calculate the difference
// between curr and prev_prev_zero
if (arr[curr] == 0)
{
// Update result if count of 1s around prev_zero is more
if (curr - prev_prev_zero > max_count)
{
max_count = curr - prev_prev_zero;
max_index = prev_zero;
}
// Update for next iteration
prev_prev_zero = prev_zero;
prev_zero = curr;
}
}
// Check for the last encountered zero
if (n-prev_prev_zero > max_count)
max_index = prev_zero;
return max_index;
}
// Driver program
int main()
{

bool arr[] = {1, 1, 0, 0, 1, 0, 1, 1, 1, 0, 1, 1, 1};

int n = sizeof(arr)/sizeof(arr[0]);
cout << "Index of 0 to be replaced is "
<< maxOnesIndex(arr, n);
return 0;
}

Java
// Java program to find Index of 0 to be replaced with 1 to get
// longest continuous sequence of 1s in a binary array
import java.io.*;
class Binary
{
// Returns index of 0 to be replaced with 1 to get longest
// continuous sequence of 1s. If there is no 0 in array, then
// it returns -1.
static int maxOnesIndex(int arr[], int n)
{
int max_count = 0; // for maximum number of 1 around a zero
int max_index=0; // for storing result
int prev_zero = -1; // index of previous zero
int prev_prev_zero = -1; // index of previous to previous zero
// Traverse the input array
for (int curr=0; curr<n; ++curr)
{
// If current element is 0, then calculate the difference
// between curr and prev_prev_zero
if (arr[curr] == 0)
{
// Update result if count of 1s around prev_zero is more
if (curr - prev_prev_zero > max_count)
{
max_count = curr - prev_prev_zero;
max_index = prev_zero;
}
// Update for next iteration
prev_prev_zero = prev_zero;
prev_zero = curr;
}
}
// Check for the last encountered zero
if (n-prev_prev_zero > max_count)
max_index = prev_zero;
return max_index;
}
// Driver program to test above function
public static void main(String[] args)
{
int arr[] = {1, 1, 0, 0, 1, 0, 1, 1, 1, 0, 1, 1, 1};
int n = arr.length;
System.out.println("Index of 0 to be replaced is "+
maxOnesIndex(arr, n));
}
}
/* This code is contributed by Devesh Agrawal */

Index of 0 to be replaced is 9

Time Complexity: O(n)

Auxiliary Space: O(1)

How to check if two given sets are disjoint?

Given two sets represented by two arrays, how to check if the given two sets are disjoint or not? It may be assumed that the given arrays have no
duplicates.
Difficulty Level: Rookie
Input: set1[] = {12, 34, 11, 9, 3}
set2[] = {2, 1, 3, 5}
Output: Not Disjoint
3 is common in two sets.
Input: set1[] = {12, 34, 11, 9, 3}
set2[] = {7, 2, 1, 5}
Output: Yes, Disjoint
There is no common element in two sets.

Method 1 (Simple)
Iterate through every element of first set and search it in other set, if any element is found, return false. If no element is found, return tree. Time
complexity of this method is O(mn).
Following is C++ implementation of above idea.
// A Simple C++ program to check if two sets are disjoint
#include<iostream>
using namespace std;
// Returns true if set1[] and set2[] are disjoint, else false
bool areDisjoint(int set1[], int set2[], int m, int n)
{
// Take every element of set1[] and search it in set2
for (int i=0; i<m; i++)
for (int j=0; j<n; j++)
if (set1[i] == set2[j])
return false;
// If no element of set1 is present in set2
return true;
}
// Driver program to test above function
int main()
{
int set1[] = {12, 34, 11, 9, 3};
int set2[] = {7, 2, 1, 5};
int m = sizeof(set1)/sizeof(set1[0]);
int n = sizeof(set2)/sizeof(set2[0]);
areDisjoint(set1, set2, m, n)? cout << "Yes" : cout << " No";
return 0;
}

Output:
Yes

Method 2 (Use Sorting and Merging)

1) Sort first and second sets.
2) Use merge like process to compare elements.
Following is C++ implementation of above idea.
// A Simple C++ program to check if two sets are disjoint
#include<iostream>
#include<algorithm>
using namespace std;
// Returns true if set1[] and set2[] are disjoint, else false
bool areDisjoint(int set1[], int set2[], int m, int n)
{
// Sort the given two sets
sort(set1, set1+m);
sort(set2, set2+n);
// Check for same elements using merge like process
int i = 0, j = 0;
while (i < m && j < n)
{

if (set1[i] < set2[j])

i++;
else if (set2[j] < set1[i])
j++;
else /* if set1[i] == set2[j] */
return false;
}
return true;
}
// Driver program to test above function
int main()
{
int set1[] = {12, 34, 11, 9, 3};
int set2[] = {7, 2, 1, 5};
int m = sizeof(set1)/sizeof(set1[0]);
int n = sizeof(set2)/sizeof(set2[0]);
areDisjoint(set1, set2, m, n)? cout << "Yes" : cout << " No";
return 0;
}

Output:
Yes

Time complexity of above solution is O(mLogm + nLogn).

The above solution first sorts both sets, then takes O(m+n) time to find intersection. If we are given that the input sets are sorted, then this method
is best among all.
Method 3 (Use Sorting and Binary Search)
This is similar to method 1. Instead of linear search, we use Binary Search.
1) Sort first set.
2) Iterate through every element of second set, and use binary search to search every element in first set. If element is found return it.
Time complexity of this method is O(mLogm + nLogm)
Method 4 (Use Binary Search Tree)
1) Create a self balancing binary search tree (Red Black, AVL, Splay, etc) of all elements in first set.
2) Iterate through all elements of second set and search every element in the above constructed Binary Search Tree. If element is found, return
false.
3) If all elements are absent, return true.
Time complexity of this method is O(mLogm + nLogm).
Method 5 (Use Hashing)
1) Create an empty hash table.
2) Iterate through the first set and store every element in hash table.
3) Iterate through second set and check if any element is present in hash table. If present, then return false, else ignore the element.
4) If all elements of second set are not present in hash table, return true.
Following is Java implementation of this method.
/* Java program to check if two sets are distinct or not */
import java.util.*;
class Main
{
// This function prints all distinct elements
static boolean areDisjoint(int set1[], int set2[])
{
// Creates an empty hashset
HashSet<Integer> set = new HashSet<>();
// Traverse the first set and store its elements in hash
for (int i=0; i<set1.length; i++)
set.add(set1[i]);
// Traverse the second set and check if any element of it
// is already in hash or not.
for (int i=0; i<set2.length; i++)
if (set.contains(set2[i]))
return false;
return true;

}
// Driver method to test above method
public static void main (String[] args)
{
int set1[] = {10, 5, 3, 4, 6};
int set2[] = {8, 7, 9, 3};
if (areDisjoint(set1, set2)
System.out.println("Yes");
else
System.out.println("No");
}
}

Output:
Yes

Time complexity of the above implementation is O(m+n) under the assumption that hash set operations like add() and contains() work in O(1)
time.

Minimum Number of Platforms Required for a Railway/Bus Station

Given arrival and departure times of all trains that reach a railway station, find the minimum number of platforms required for the railway station so
that no train waits.
We are given two arrays which represent arrival and departure times of trains that stop
Examples:
Input: arr[] = {9:00, 9:40, 9:50, 11:00, 15:00, 18:00}
dep[] = {9:10, 12:00, 11:20, 11:30, 19:00, 20:00}
Output: 3
There are at-most three trains at a time (time between 11:00 to 11:20)

We can solve the above problem in O(nLogn) time. The idea is to consider all evens in sorted order. Once we have all events in sorted order,
we can trace the number of trains at any time keeping track of trains that have arrived, but not departed.
For example consider the above example.
arr[] = {9:00, 9:40, 9:50, 11:00, 15:00, 18:00}
dep[] = {9:10, 12:00, 11:20, 11:30, 19:00, 20:00}
All events sorted by time.
Total platforms at any time can be obtained by subtracting total
departures from total arrivals by that time.
Time
Event Type
Total Platforms Needed at this Time
9:00
Arrival
1
9:10
Departure
0
9:40
Arrival
1
9:50
Arrival
2
11:00
Arrival
3
11:20
Departure
2
11:30
Departure
1
12:00
Departure
0
15:00
Arrival
1
18:00
Arrival
2
19:00
Departure
1
20:00
Departure
0
Minimum Platforms needed on railway station = Maximum platforms
needed at any time
= 3

Following is C++ implementation of above approach. Note that the implementation doesnt create a single sorted list of all events, rather it
individually sorts arr[] and dep[] arrays, and then uses merge process of merge sort to process them together as a single sorted array.
// Program to find minimum number of platforms required on a railway station
#include<iostream>
#include<algorithm>
using namespace std;
// Returns minimum number of platforms reqquired
int findPlatform(int arr[], int dep[], int n)
{
// Sort arrival and departure arrays
sort(arr, arr+n);
sort(dep, dep+n);
// plat_needed indicates number of platforms needed at a time
int plat_needed = 1, result = 1;
int i = 1, j = 0;
// Similar to merge in merge sort to process all events in sorted order
while (i < n && j < n)
{
// If next event in sorted order is arrival, increment count of
// platforms needed
if (arr[i] < dep[j])
{
plat_needed++;
i++;
if (plat_needed > result) // Update result if needed
result = plat_needed;
}
else // Else decrement count of platforms needed
{
plat_needed--;
j++;
}

}
return result;
}
// Driver program to test methods of graph class
int main()
{
int arr[] = {900, 940, 950, 1100, 1500, 1800};
int dep[] = {910, 1200, 1120, 1130, 1900, 2000};
int n = sizeof(arr)/sizeof(arr[0]);
cout << "Minimum Number of Platforms Required = "
<< findPlatform(arr, dep, n);
return 0;
}

Output:
Minimum Number of Platforms Required = 3

Algorithmic Paradigm: Dynamic Programming

Time Complexity: O(nLogn), assuming that a O(nLogn) sorting algorithm for sorting arr[] and dep[].

Length of the largest subarray with contiguous elements | Set 1

Given an array of distinct integers, find length of the longest subarray which contains numbers that can be arranged in a continuous sequence.
Examples:
Input: arr[] = {10, 12, 11};
Output: Length of the longest contiguous subarray is 3
Input: arr[] = {14, 12, 11, 20};
Output: Length of the longest contiguous subarray is 2
Input: arr[] = {1, 56, 58, 57, 90, 92, 94, 93, 91, 45};
Output: Length of the longest contiguous subarray is 5

The important thing to note in question is, it is given that all elements are distinct. If all elements are distinct, then a subarray has contiguous
elements if and only if the difference between maximum and minimum elements in subarray is equal to the difference between last and first indexes
of subarray. So the idea is to keep track of minimum and maximum element in every subarray.
The following is C++ implementation of above idea.
#include<iostream>
using namespace std;
// Utility functions to find minimum and maximum of
// two elements
int min(int x, int y) { return (x < y)? x : y; }
int max(int x, int y) { return (x > y)? x : y; }
// Returns length of the longest contiguous subarray
int findLength(int arr[], int n)
{
int max_len = 1; // Initialize result
for (int i=0; i<n-1; i++)
{
// Initialize min and max for all subarrays starting with i
int mn = arr[i], mx = arr[i];
// Consider all subarrays starting with i and ending with j
for (int j=i+1; j<n; j++)
{
// Update min and max in this subarray if needed
mn = min(mn, arr[j]);
mx = max(mx, arr[j]);
// If current subarray has all contiguous elements
if ((mx - mn) == j-i)
max_len = max(max_len, mx-mn+1);
}
}
return max_len; // Return result
}
// Driver program to test above function
int main()
{
int arr[] = {1, 56, 58, 57, 90, 92, 94, 93, 91, 45};
int n = sizeof(arr)/sizeof(arr[0]);
cout << "Length of the longest contiguous subarray is "
<< findLength(arr, n);
return 0;
}

Output:
Length of the longest contiguous subarray is 5

Time Complexity of the above solution is O(n2).

We will soon be covering solution for the problem where duplicate elements are allowed in subarray.

Length of the largest subarray with contiguous elements | Set 2

Given an array of integers, find length of the longest subarray which contains numbers that can be arranged in a continuous sequence.
In the previous post, we have discussed a solution that assumes that elements in given array are distinct. Here we discuss a solution that works
even if the input array has duplicates.
Examples:
Input: arr[] = cc
Output: Length of the longest contiguous subarray is 4
Input: arr[] = {10, 12, 12, 10, 10, 11, 10};
Output: Length of the longest contiguous subarray is 2

The idea is similar to previous post. In the previous post, we checked whether maximum value minus minimum value is equal to ending index minus
starting index or not. Since duplicate elements are allowed, we also need to check if the subarray contains duplicate elements or not. For example,
the array {12, 14, 12} follows the first property, but numbers in it are not contiguous elements.
To check duplicate elements in a subarray, we create a hash set for every subarray and if we find an element already in hash, we dont consider the
current subarray.
Following is Java implementation of the above idea.
/* Java program to find length of the largest subarray which has
all contiguous elements */
import java.util.*;
class Main
{
// This function prints all distinct elements
static int findLength(int arr[])
{
int n = arr.length;
int max_len = 1; // Inialize result
// One by one fix the starting points
for (int i=0; i<n-1; i++)
{
// Create an empty hash set and add i'th element
// to it.
HashSet<Integer> set = new HashSet<>();
set.add(arr[i]);
// Initialize max and min in current subarray
int mn = arr[i], mx = arr[i];
// One by one fix ending points
for (int j=i+1; j<n; j++)
{
// If current element is already in hash set, then
// this subarray cannot contain contiguous elements
if (set.contains(arr[j]))
break;
// Else add curremt element to hash set and update
// min, max if required.
set.add(arr[j]);
mn = Math.min(mn, arr[j]);
mx = Math.max(mx, arr[j]);
// We have already cheched for duplicates, now check
// for other property and update max_len if needed
if (mx-mn == j-i)
max_len = Math.max(max_len, mx-mn+1);
}
}
return max_len; // Return result
}
// Driver method to test above method
public static void main (String[] args)
{
int arr[] = {10, 12, 12, 10, 10, 11, 10};
System.out.println("Length of the longest contiguous subarray is " +
findLength(arr));
}
}

Output:
Length of the longest contiguous subarray is 2

Time complexity of the above solution is O(n2) under the assumption that hash set operations like add() and contains() work in O(1) time.

Print all increasing sequences of length k from first n natural numbers

Given two positive integers n and k, print all increasing sequences of length k such that the elements in every sequence are from first n natural
numbers.
Examples:
Input: k = 2, n = 3
Output: 1 2
1 3
2 3
Input: k = 5, n = 5
Output: 1 2 3 4 5
Input: k = 3, n = 5
Output: 1 2 3
1 2 4
1 2 5
1 3 4
1 3 5
1 4 5
2 3 4
2 3 5
2 4 5
3 4 5

Its a good recursion question. The idea is to create an array of length k. The array stores current sequence. For every position in array, we check
the previous element and one by one put all elements greater than the previous element. If there is no previous element (first position), we put all
numbers from 1 to n.
Following is C++ implementation of above idea.
// C++ program to print all increasing sequences of
// length 'k' such that the elements in every sequence
// are from first 'n' natural numbers.
#include<iostream>
using namespace std;
// A utility function to print contents of arr[0..k-1]
void printArr(int arr[], int k)
{
for (int i=0; i<k; i++)
cout << arr[i] << " ";
cout << endl;
}
// A recursive function to print all increasing sequences
// of first n natural numbers. Every sequence should be
// length k. The array arr[] is used to store current
// sequence.
void printSeqUtil(int n, int k, int &len, int arr[])
{
// If length of current increasing sequence becomes k,
// print it
if (len == k)
{
printArr(arr, k);
return;
}
// Decide the starting number to put at current position:
// If length is 0, then there are no previous elements
// in arr[]. So start putting new numbers with 1.
// If length is not 0, then start from value of
// previous element plus 1.
int i = (len == 0)? 1 : arr[len-1] + 1;
// Increase length
len++;
// Put all numbers (which are greater than the previous
// element) at new position.
while (i<=n)
{
arr[len-1] = i;
printSeqUtil(n, k, len, arr);
i++;
}

// This is important. The variable 'len' is shared among

// all function calls in recursion tree. Its value must be
// brought back before next iteration of while loop
len--;
}
// This function prints all increasing sequences of
// first n natural numbers. The length of every sequence
// must be k. This function mainly uses printSeqUtil()
void printSeq(int n, int k)
{
int arr[k]; // An array to store individual sequences
int len = 0; // Initial length of current sequence
printSeqUtil(n, k, len, arr);
}
// Driver program to test above functions
int main()
{
int k = 3, n = 7;
printSeq(n, k);
return 0;
}

Output:
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
4
4
4
5

2
2
2
2
2
3
3
3
3
4
4
4
5
5
6
3
3
3
3
4
4
4
5
5
6
4
4
4
5
5
6
5
5
6
6

3
4
5
6
7
4
5
6
7
5
6
7
6
7
7
4
5
6
7
5
6
7
6
7
7
5
6
7
6
7
7
6
7
7
7

Given two strings, find if first string is a subsequence of second

Given two strings str1 and str2, find if str1 is a subsequence of str2. A subsequence is a sequence that can be derived from another sequence by
deleting some elements without changing the order of the remaining elements (source: wiki). Expected time complexity is linear.
Examples:
Input: str1 = "AXY", str2 = "ADXCPY"
Output: True (str1 is a subsequence of str2)
Input: str1 = "AXY", str2 = "YADXCP"
Output: False (str1 is not a subsequence of str2)
Input: str1 = "gksrek", str2 = "geeksforgeeks"
Output: True (str1 is a subsequence of str2)

The idea is simple, we traverse both strings from one side to other side (say from rightmost character to leftmost). If we find a matching character,
we move ahead in both strings. Otherwise we move ahead only in str2.
Following is Recursive Implementation in C++ and Python of the above idea.

C/C++
// Recursive C++ program to check if a string is subsequence of another string
#include<iostream>
#include<cstring>
using namespace std;
// Returns true if str1[] is a subsequence of str2[]. m is
// length of str1 and n is length of str2
bool isSubSequence(char str1[], char str2[], int m, int n)
{
// Base Cases
if (m == 0) return true;
if (n == 0) return false;
// If last characters of two strings are matching
if (str1[m-1] == str2[n-1])
return isSubSequence(str1, str2, m-1, n-1);
// If last characters are not matching
return isSubSequence(str1, str2, m, n-1);
}
// Driver program to test methods of graph class
int main()
{
char str1[] = "gksrek";
char str2[] = "geeksforgeeks";
int m = strlen(str1);
int n = strlen(str2);
isSubSequence(str1, str2, m, n)? cout << "Yes ":
cout << "No";
return 0;
}

Python
# Recursive Python program to check if a string is subsequence
# of another string
# Returns true if str1[] is a subsequence of str2[]. m is
# length of str1 and n is length of str2
def isSubSequence(string1, string2, m, n):
# Base Cases
if m == 0:
return True
if n == 0:
return False
# If last characters of two strings are matching
if string1[m-1] == string2[n-1]:
return isSubSequence(string1, string2, m-1, n-1)
# If last characters are not matching
return isSubSequence(string1, string2, m, n-1)
# Driver program to test the above function
string1 = "gksrek"

string2 = "geeksforgeeks"
m = len(string1)
n = len(string2)
if isSubSequence(string1, string2, m, n):
print "Yes"
else:
print "No"
# This code is contributed by BHAVYA JAIN

Yes

Following is Iterative Implementation in C++ for the same.

// Iterative C++ program to check if a string is subsequence of another string
#include<iostream>
#include<cstring>
using namespace std;
// Returns true if str1[] is a subsequence of str2[]. m is
// length of str1 and n is length of str2
bool isSubSequence(char str1[], char str2[], int m, int n)
{
int j = 0; // For index of str1 (or subsequence
// Traverse str2 and str1, and compare current character
// of str2 with first unmatched char of str1, if matched
// then move ahead in str1
for (int i=0; i<n&&j<m; i++)
if (str1[j] == str2[i])
j++;
// If all characters of str1 were found in str2
return (j==m);
}
// Driver program to test methods of graph class
int main()
{
char str1[] = "gksrek";
char str2[] = "geeksforgeeks";
int m = strlen(str1);
int n = strlen(str2);
isSubSequence(str1, str2, m, n)? cout << "Yes ":
cout << "No";
return 0;
}

Output:
Yes

Time Complexity of both implementations above is O(n) where n is the length of str2.

Snake and Ladder Problem

Given a snake and ladder board, find the minimum number of dice throws required to reach the destination or last cell from source or 1st cell.
Basically, the player has total control over outcome of dice throw and wants to find out minimum number of throws required to reach last cell.
If the player reaches a cell which is base of a ladder, the player has to climb up that ladder and if reaches a cell is mouth of the snake, has to go
down to the tail of snake without a dice throw.

For example consider the board shown on right side (taken from here), the minimum number of dice throws required to reach cell 30 from cell 1 is
3. Following are steps.
a) First throw two on dice to reach cell number 3 and then ladder to reach 22
b) Then throw 6 to reach 28.
c) Finally through 2 to reach 30.
There can be other solutions as well like (2, 2, 6), (2, 4, 4), (2, 3, 5).. etc.
Following is C++ implementation of the above idea. The input is represented by two things, first is N which is number of cells in the given board,
second is an array move[0N-1] of size N. An entry move[i] is -1 if there is no snake and no ladder from i, otherwise move[i] contains index of
destination cell for the snake or the ladder at i.
// C++ program to find minimum number of dice throws required to
// reach last cell from first cell of a given snake and ladder
// board
#include<iostream>
#include <queue>
using namespace std;
// An entry in queue used in BFS
struct queueEntry
{
int v;
// Vertex number
int dist; // Distance of this vertex from source
};
// This function returns minimum number of dice throws required to
// Reach last cell from 0'th cell in a snake and ladder game.
// move[] is an array of size N where N is no. of cells on board
// If there is no snake or ladder from cell i, then move[i] is -1
// Otherwise move[i] contains cell to which snake or ladder at i
// takes to.
int getMinDiceThrows(int move[], int N)
{
// The graph has N vertices. Mark all the vertices as
// not visited
bool *visited = new bool[N];
for (int i = 0; i < N; i++)
visited[i] = false;
// Create a queue for BFS
queue<queueEntry> q;
// Mark the node 0 as visited and enqueue it.
visited[0] = true;
queueEntry s = {0, 0}; // distance of 0't vertex is also 0
q.push(s); // Enqueue 0'th vertex
// Do a BFS starting from vertex at index 0
queueEntry qe; // A queue entry (qe)
while (!q.empty())
{
qe = q.front();
int v = qe.v; // vertex no. of queue entry

// If front vertex is the destination vertex,

// we are done
if (v == N-1)
break;
// Otherwise dequeue the front vertex and enqueue
// its adjacent vertices (or cell numbers reachable
// through a dice throw)
q.pop();
for (int j=v+1; j<=(v+6) && j<N; ++j)
{
// If this cell is already visited, then ignore
if (!visited[j])
{
// Otherwise calculate its distance and mark it
// as visited
queueEntry a;
a.dist = (qe.dist + 1);
visited[j] = true;
//
//
//
if

Check if there a snake or ladder at 'j'

then tail of snake or top of ladder
become the adjacent of 'i'
(move[j] != -1)
a.v = move[j];
else
a.v = j;
q.push(a);
}
}
}
// We reach here when 'qe' has last vertex
// return the distance of vertex in 'qe'
return qe.dist;
}
// Driver program to test methods of graph class
int main()
{
// Let us construct the board given in above diagram
int N = 30;
int moves[N];
for (int i = 0; i<N; i++)
moves[i] = -1;
// Ladders
moves[2] = 21;
moves[4] = 7;
moves[10] = 25;
moves[19] = 28;
// Snakes
moves[26]
moves[20]
moves[16]
moves[18]

=
=
=
=

0;
8;
3;
6;

cout << "Min Dice throws required is " << getMinDiceThrows(moves, N);
return 0;
}

Output:
Min Dice throws required is 3

Time complexity of the above solution is O(N) as every cell is added and removed only once from queue. And a typical enqueue or dequeue
operation takes O(1) time.

Write a function that returns 2 for input 1 and returns 1 for 2

Write a function which returns 1 that 2 is passed and return 2 when 1 is passed.
Source: Adobe Interview Experience | Set 19 (For MTS)
A simple solution is to compare the passed value with 1 and 2.
int invert(int x)
{
if (x == 1) return 2;
else return 1;
}

Another solution is to use subtraction

int invertSub(int x)
{
return (3-x);
}

We can also use bitwise xor operator.

int invertXOR(int x)
{
return (x ^ 1 ^ 2);
}

Connect n ropes with minimum cost

There are given n ropes of different lengths, we need to connect these ropes into one rope. The cost to connect two ropes is equal to sum of their
lengths. We need to connect the ropes with minimum cost.
For example if we are given 4 ropes of lengths 4, 3, 2 and 6. We can connect the ropes in following ways.
1) First connect ropes of lengths 2 and 3. Now we have three ropes of lengths 4, 6 and 5.
2) Now connect ropes of lengths 4 and 5. Now we have two ropes of lengths 6 and 9.
3) Finally connect the two ropes and all ropes have connected.
Total cost for connecting all ropes is 5 + 9 + 15 = 29. This is the optimized cost for connecting ropes. Other ways of connecting ropes would
always have same or more cost. For example, if we connect 4 and 6 first (we get three strings of 3, 2 and 10), then connect 10 and 3 (we get two
strings of 13 and 2). Finally we connect 13 and 2. Total cost in this way is 10 + 13 + 15 = 38.
Following is complete algorithm for finding the minimum cost for connecting n ropes.
Let there be n ropes of lengths stored in an array len[0..n-1]
1) Create a min heap and insert all lengths into the min heap.
2) Do following while number of elements in min heap is not one.
a) Extract the minimum and second minimum from min heap
b) Add the above two extracted values and insert the added value to the min-heap.
3) Return the value of only left item in min heap.
Following is C++ implementation of above algorithm.
// C++ program for connecting n ropes with minimum cost
#include <iostream>
using namespace std;
// A Min Heap: Collection of min heap nodes
struct MinHeap
{
unsigned size;
// Current size of min heap
unsigned capacity; // capacity of min heap
int *harr; // Attay of minheap nodes
};
// A utility function to create a min heap of given capacity
struct MinHeap* createMinHeap(unsigned capacity)
{
struct MinHeap* minHeap = new MinHeap;
minHeap->size = 0; // current size is 0
minHeap->capacity = capacity;
minHeap->harr = new int[capacity];
return minHeap;
}
// A utility function to swap two min heap nodes
void swapMinHeapNode(int* a, int* b)
{
int temp = *a;
*a = *b;
*b = temp;
}
// The standard minHeapify function.
void minHeapify(struct MinHeap* minHeap, int idx)
{
int smallest = idx;
int left = 2 * idx + 1;
int right = 2 * idx + 2;
if (left < minHeap->size &&
minHeap->harr[left] < minHeap->harr[smallest])
smallest = left;
if (right < minHeap->size &&
minHeap->harr[right] < minHeap->harr[smallest])
smallest = right;
if (smallest != idx)
{
swapMinHeapNode(&minHeap->harr[smallest], &minHeap->harr[idx]);
minHeapify(minHeap, smallest);
}
}

// A utility function to check if size of heap is 1 or not

int isSizeOne(struct MinHeap* minHeap)
{
return (minHeap->size == 1);
}
// A standard function to extract minimum value node from heap
int extractMin(struct MinHeap* minHeap)
{
int temp = minHeap->harr[0];
minHeap->harr[0] = minHeap->harr[minHeap->size - 1];
--minHeap->size;
minHeapify(minHeap, 0);
return temp;
}
// A utility function to insert a new node to Min Heap
void insertMinHeap(struct MinHeap* minHeap, int val)
{
++minHeap->size;
int i = minHeap->size - 1;
while (i && (val < minHeap->harr[(i - 1)/2]))
{
minHeap->harr[i] = minHeap->harr[(i - 1)/2];
i = (i - 1)/2;
}
minHeap->harr[i] = val;
}
// A standard funvtion to build min heap
void buildMinHeap(struct MinHeap* minHeap)
{
int n = minHeap->size - 1;
int i;
for (i = (n - 1) / 2; i >= 0; --i)
minHeapify(minHeap, i);
}
// Creates a min heap of capacity equal to size and inserts all values
// from len[] in it. Initially size of min heap is equal to capacity
struct MinHeap* createAndBuildMinHeap(int len[], int size)
{
struct MinHeap* minHeap = createMinHeap(size);
for (int i = 0; i < size; ++i)
minHeap->harr[i] = len[i];
minHeap->size = size;
buildMinHeap(minHeap);
return minHeap;
}
// The main function that returns the minimum cost to connect n ropes of
// lengths stored in len[0..n-1]
int minCost(int len[], int n)
{
int cost = 0; // Initialize result
// Create a min heap of capacity equal to n and put all ropes in it
struct MinHeap* minHeap = createAndBuildMinHeap(len, n);
// Iterate while size of heap doesn't become 1
while (!isSizeOne(minHeap))
{
// Extract two minimum length ropes from min heap
int min
= extractMin(minHeap);
int sec_min = extractMin(minHeap);
cost += (min + sec_min); // Update total cost
// Insert a new rope in min heap with length equal to sum
// of two extracted minimum lengths
insertMinHeap(minHeap, min+sec_min);
}
// Finally return total minimum cost for connecting all ropes
return cost;
}
// Driver program to test above functions
int main()
{
int len[] = {4, 3, 2, 6};

int size = sizeof(len)/sizeof(len[0]);

cout << "Total cost for connecting ropes is " << minCost(len, size);
return 0;
}

Output:
Total cost for connecting ropes is 29

Time Complexity: Time complexity of the algorithm is O(nLogn) assuming that we use a O(nLogn) sorting algorithm. Note that heap operations
like insert and extract take O(Logn) time.
Algorithmic Paradigm: Greedy Algorithm

A simple implementation with STL in C++

Following is a simple implementation that uses priority_queue available in STL. Thanks to Pango89 for providing below code.
#include<iostream>
#include<queue>
using namespace std;
int minCost(int arr[], int n)
{
// Create a priority queue ( http://www.cplusplus.com/reference/queue/priority_queue/ )
// By default 'less' is used which is for decreasing order
// and 'greater' is used for increasing order
priority_queue< int, vector<int>, greater<int> > pq(arr, arr+n);
// Initialize result
int res = 0;
// While size of priority queue is more than 1
while (pq.size() > 1)
{
// Extract shortest two ropes from pq
int first = pq.top();
pq.pop();
int second = pq.top();
pq.pop();
// Connect the ropes: update result and
// insert the new rope to pq
res += first + second;
pq.push(first + second);
}
return res;
}
// Driver program to test above function
int main()
{
int len[] = {4, 3, 2, 6};
int size = sizeof(len)/sizeof(len[0]);
cout << "Total cost for connecting ropes is " << minCost(len, size);
return 0;
}

Output:
Total cost for connecting ropes is 29

Find the number of valid parentheses expressions of given length

Given a number n find the number of valid parentheses expressions of that length.
Input: 2
Output: 1
There is only possible valid expression of length 2, "()"
Input: 4
Output: 2
Possible valid expression of length 4 are "(())" and "()()"
Input: 6
Output: 5
Possible valid expressions are ((())), ()(()), ()()(), (())() and (()())

This is mainly an application of Catalan Numbers. Total possible valid expressions for input n is n/2th Catalan Number if n is even and 0 if n is odd.
Following is C++ implementation of the above idea.
// C++ program to find valid paranthesisations of length n
// The majority of code is taken from method 3 of
// http://www.geeksforgeeks.org/program-nth-catalan-number/
#include<iostream>
using namespace std;
// Returns value of Binomial Coefficient C(n, k)
unsigned long int binomialCoeff(unsigned int n, unsigned int k)
{
unsigned long int res = 1;
// Since C(n, k) = C(n, n-k)
if (k > n - k)
k = n - k;
// Calculate value of [n*(n-1)*---*(n-k+1)] / [k*(k-1)*---*1]
for (int i = 0; i < k; ++i)
{
res *= (n - i);
res /= (i + 1);
}
return res;
}
// A Binomial coefficient based function to find nth catalan
// number in O(n) time
unsigned long int catalan(unsigned int n)
{
// Calculate value of 2nCn
unsigned long int c = binomialCoeff(2*n, n);
// return 2nCn/(n+1)
return c/(n+1);
}
// Function to find possible ways to put balanced parenthesis
// in an expression of length n
unsigned long int findWays(unsigned n)
{
// If n is odd, not possible to create any valid parentheses
if (n & 1) return 0;
// Otherwise return n/2'th Catalan Numer
return catalan(n/2);
}
// Driver program to test above functions
int main()
{
int n = 6;
cout << "Total possible expressions of length "
<< n << " is " << findWays(6);
return 0;
}

Output:
Total possible expressions of length 6 is 5

Time Complexity: O(n)

Longest Monotonically Increasing Subsequence Size (N log N): Simple

implementation
Given an array of random numbers, find longest monotonically increasing subsequence (LIS) in the array.
If you want to understand the O(NlogN) approach, its explained very clearly here.
In this post, a simple and time saving implementation of O(NlogN) approach using stl is discussed. Below is the code for LIS O(NlogN) :
// A simple C++ implementation to find LIS
#include<iostream>
#include<iostream>
#include<algorithm>
#include<set>
using namespace std;
// Return length of LIS in arr[] of size N
int lis(int arr[], int N)
{
int i;
set<int> s;
set<int>::iterator k;
for (i=0; i<N; i++)
{
// Check if the element was actually inserted
// An element in set is not inserted if it is
// already present. Please see
// http://www.cplusplus.com/reference/set/set/insert/
if (s.insert(arr[i]).second)
{
// Find the position of inserted element in iterator k
k = s.find(arr[i]);
k++; // Find the next greater element in set
// If the new element is not inserted at the end, then
// remove the greater element next to it (This is tricky)
if (k!=s.end()) s.erase(k);
}
}
// Note that set s may not contain actual LIS, but its size gives
// us the length of LIS
return s.size();
}
int main()
{
int arr[] = {8, 9, 12, 10, 11};
int n = sizeof(arr)/sizeof(arr[0]);
cout << lis(arr, n)<< endl;
}

Output:
4

Generate all binary permutations such that there are more or equal 1s than 0s before
every point in all permutations
Generate all permutations of given length such that every permutation has more or equal 1s than 0s in all prefixes of the permutation.
Examples:
Input: len = 4
Output: 1111 1110 1101 1100 1011 1010
Note that a permutation like 0101 can not be in output because
there are more 0's from index 0 to 2 in this permutation.
Input: len = 3
Output: 111 110 101
Input: len = 2
Output: 11 10

Like permutation generation problems, recursion is the simplest approach to solve this. We start with an empty string, attach 1 to it and recur.
While recurring, if we find more 1s at any point, we append a 0 and make one more recursive call.
// C++ program to generate all permutations of 1's and 0's such that
// every permutation has more 1's than 0's at all indexes.
#include <iostream>
#include <cstring>
using namespace std;
// ones & zeroes --> counts of 1's and 0's in current string 'str'
// len ---> desired length of every permutation
void generate(int ones, int zeroes, string str, int len)
{
// If length of current string becomes same as desired length
if (len == str.length())
{
cout << str << " ";
return;
}
// Append a 1 and recur
generate(ones+1, zeroes, str+"1", len);
// If there are more 1's, append a 0 as well, and recur
if (ones > zeroes)
generate(ones, zeroes+1, str+"0", len);
}
// Driver program to test above function
int main()
{
string str = "";
generate(0, 0, str, 4);
return 0;
}

Output:
1111 1110 1101 1100 1011 1010

Lexicographically minimum string rotation

Write code to find lexicographic minimum in a circular array, e.g. for the array BCABDADAB, the lexicographic minimum is ABBCABDAD.
Source: Google Written Test
More Examples:
Input: GEEKSQUIZ
Output: EEKSQUIZG
Input: GFG
Output: FGG
Input: GEEKSFORGEEKS
Output: EEKSFORGEEKSG

Following is a simple solution. Let the given string be str

1) Concatenate str with itself and store in a temporary string say concat.
2) Create an array of strings to store all rotations of str. Let the array be arr.
3) Find all rotations of str by taking substrings of concat at index 0, 1, 2..n-1. Store these rotations in arr[]
4) Sort arr[] and return arr[0].
Following is C++ implementation of above solution.
// A simple C++ program to find lexicographically minimum rotation
// of a given string
#include <iostream>
#include <algorithm>
using namespace std;
// This functionr return lexicographically minimum
// rotation of str
string minLexRotation(string str)
{
// Find length of given string
int n = str.length();
// Create an array of strings to store all rotations
string arr[n];
// Create a concatenation of string with itself
string concat = str + str;
// One by one store all rotations of str in array.
// A rotation is obtained by getting a substring of concat
for (int i = 0; i < n; i++)
arr[i] = concat.substr(i, n);
// Sort all rotations
sort(arr, arr+n);
// Return the first rotation from the sorted array
return arr[0];
}
// Driver program to test above function
int main()
{
cout << minLexRotation("GEEKSFORGEEKS") << endl;
cout << minLexRotation("GEEKSQUIZ") << endl;
cout << minLexRotation("BCABDADAB") << endl;
}

Output:
EEKSFORGEEKSG
EEKSQUIZG
ABBCABDAD

Time complexity of the above solution is O(n2Logn) under the assumption that we have used a O(nLogn) sorting algorithm.
This problem can be solved using more efficient methods like Booths Algorithm which solves the problem in O(n) time. We will soon be covering
these methods as separate posts.

Construct an array from its pair-sum array

Given a pair-sum array and size of the original array (n), construct the original array.
A pair-sum array for an array is the array that contains sum of all pairs in ordered form. For example pair-sum array for arr[] = {6, 8, 3, 4} is {14,
9, 10, 11, 12, 7}.
In general, pair-sum array for arr[0..n-1] is {arr[0]+arr[1], arr[0]+arr[2], ., arr[1]+arr[2], arr[1]+arr[3], ., arr[2]+arr[3], arr[2]+arr[4], ., arr[n2]+arr[n-1}.
Given a pair-sum array, construct the original array.
Let the given array be pair[] and let there be n elements in original array. If we take a look at few examples, we can observe that arr[0] is half of
pair[0] + pair[1] pair[n-1]. Note that the value of pair[0] + pair[1] pair[n-1] is (arr[0] + arr[1]) + (arr[0] + arr[2]) (arr[1] + arr[2]).
Once we have evaluated arr[0], we can evaluate other elements by subtracting arr[0]. For example arr[1] can be evaluated by subtracting arr[0]
from pair[0], arr[2] can be evaluated by subtracting arr[0] from pair[1].
Following are C++ and Java implementations of the above idea.

C++
#include <iostream>
using namespace std;
// Fills element in arr[] from its pair sum array pair[].
// n is size of arr[]
void constructArr(int arr[], int pair[], int n)
{
arr[0] = (pair[0]+pair[1]-pair[n-1]) / 2;
for (int i=1; i<n; i++)
arr[i] = pair[i-1]-arr[0];
}
// Driver program to test above function
int main()
{
int pair[] = {15, 13, 11, 10, 12, 10, 9, 8, 7, 5};
int n = 5;
int arr[n];
constructArr(arr, pair, n);
for (int i = 0; i < n; i++)
cout << arr[i] << " ";
return 0;
}

Java
import java.io.*;
class PairSum {
// Fills element in arr[] from its pair sum array pair[].
// n is size of arr[]
static void constructArr(int arr[], int pair[], int n)
{
arr[0] = (pair[0]+pair[1]-pair[n-1]) / 2;
for (int i=1; i<n; i++)
arr[i] = pair[i-1]-arr[0];
}
// Driver program to test above function
public static void main(String[] args)
{
int pair[] = {15, 13, 11, 10, 12, 10, 9, 8, 7, 5};
int n = 5;
int[] arr = new int[n];
constructArr(arr, pair, n);
for (int i = 0; i < n; i++)
System.out.print(arr[i]+" ");
}
}
/* This code is contributed by Devesh Agrawal */

8 7 5 3 2

Time complexity of constructArr() is O(n) where n is number of elements in arr[].

Program to evaluate simple expressions

You are given a string that represent an expression of digits and operands. E.g. 1+2*3, 1-2+4. You need to evaluate the string or the expression.
NO BODMAS is followed. If the expression is of incorrect syntax return -1.
Test cases:
a) 1+2*3 will be evaluated to 9.
b) 4-2+6*3 will be evaluated to 24.
c) 1++2 will be evaluated to -1(INVALID).
Also, in the string spaces can occur. For that case we need to ignore the spaces. Like :- 1*2 -1 is equals to 1.
Source: Amazon Interview Question
It is strongly recommend to minimize the browser and try this yourself first.
The idea is simple start from the first character and traverse from left to right and check for errors like two consecutive operators and operands.
We also keep track of result and update the result while traversing the expression.
Following is C++ program to evaluate the given expression.
// C++ program to evaluate a given expression
#include <ioexpeam>
using namespace std;
// A utility function to check if a given character is operand
bool isOperand(char c) { return (c >= '0' && c <= '9'); }
// utility function to find value of and operand
int value(char c) { return (c - '0'); }
// This function evaluates simple expressions. It returns -1 if the
// given expression is invalid.
int evaluate(char *exp)
{
// Base Case: Given expression is empty
if (*exp == '\0') return -1;
// The first character must be an operand, find its value
int res = value(exp[0]);
// Traverse the remaining characters in pairs
for (int i = 1; exp[i]; i += 2)
{
// The next character must be an operator, and
// next to next an operand
char opr = exp[i], opd = exp[i+1];
// If next to next character is not an operand
if (!isOperand(opd)) return -1;
// Update result according to the operator
if (opr == '+')
res += value(opd);
else if (opr == '-') res -= value(opd);
else if (opr == '*') res *= value(opd);
else if (opr == '/') res /= value(opd);
// If not a valid operator
else
return -1;
}
return res;
}
// Driver program to test above function
int main()
{
char expr1[] = "1+2*5+3";
int res = evaluate(expr1);
(res == -1)? cout << expr1 << " is " << "Invalid\n":
cout << "Value of " << expr1 << " is " << res << endl;
char expr2[] = "1+2*3";
res = evaluate(expr2);
(res == -1)? cout << expr2 << " is " << "Invalid\n":
cout << "Value of " << expr2 << " is " << res << endl;
char expr3[] = "4-2+6*3";
res = evaluate(expr3);
(res == -1)? cout << expr3 << " is " << "Invalid\n":
cout << "Value of " << expr3 << " is " << res << endl;

char expr4[] = "1++2";

res = evaluate(expr4);
(res == -1)? cout << expr4 << " is " << "Invalid\n":
cout << "Value of " << expr4 << " is " << res << endl;
return 0;
}

Output:
Value of 1+2*5+3 is 18
Value of 1+2*3 is 9
Value of 4-2+6*3 is 24
1++2 is Invalid

The above code doesnt handle spaces. We can handle spaces by first removing all spaces from the given string. A better solution is to handle
spaces in single traversal. This is left as an exercise.
Time Complexity is O(n) where n is length of the given expression.

Check if characters of a given string can be rearranged to form a palindrome

Given a string, Check if characters of the given string can be rearranged to form a palindrome.
For example characters of geeksogeeks can be rearranged to form a palindrome geeksoskeeg, but characters of geeksforgeeks cannot be
rearranged to form a palindrome.
A set of characters can form a palindrome if at most one character occurs odd number of times and all characters occur even number of times.
A simple solution is to run two loops, the outer loop picks all characters one by one, the inner loop counts number of occurrences of the picked
character. We keep track of odd counts. Time complexity of this solution is O(n2).
We can do it in O(n) time using a count array. Following are detailed steps.
1) Create a count array of alphabet size which is typically 256. Initialize all values of count array as 0.
2) Traverse the given string and increment count of every character.
3) Traverse the count array and if the count array has more than one odd values, return false. Otherwise return true.
Following is C++ implementation of above approach.
#include <iostream>
using namespace std;
#define NO_OF_CHARS 256
/* function to check whether characters of a string can form
a palindrome */
bool canFormPalindrome(char *str)
{
// Create a count array and initialize all values as 0
int count[NO_OF_CHARS] = {0};
// For each character in input strings, increment count in
// the corresponding count array
for (int i = 0; str[i]; i++)
count[str[i]]++;
// Count odd occurring characters
int odd = 0;
for (int i = 0; i < NO_OF_CHARS; i++)
if (count[i] & 1)
odd++;
// Return true if odd count is 0 or 1, otherwise false
return (odd <= 1);
}
/* Driver program to test to pront printDups*/
int main()
{
canFormPalindrome("geeksforgeeks")? cout << "Yes\n":
cout << "No\n";
canFormPalindrome("geeksogeeks")? cout << "Yes\n":
cout << "No\n";
return 0;
}

Output:
No
Yes

Print all pairs of anagrams in a given array of strings

Given an array of strings, find all anagram pairs in the given array.
Example:
Input: arr[] = {"geeksquiz", "geeksforgeeks", "abcd",
"forgeeksgeeks", "zuiqkeegs"};
Output: (geeksforgeeks, forgeeksgeeks), (geeksquiz, zuiqkeegs)

We can find whether two strings are anagram or not in linear time using count array (see method 2 of this).
One simple idea to find whether all anagram pairs is to run two nested loops. The outer loop picks all strings one by one. The inner loop checks
whether remaining strings are anagram of the string picked by outer loop. Following is C++ implementation of this simple approach.
#include <iostream>
using namespace std;
#define NO_OF_CHARS 256
/* function to check whether two strings are anagram of each other */
bool areAnagram(string str1, string str2)
{
// Create two count arrays and initialize all values as 0
int count[NO_OF_CHARS] = {0};
int i;
// For each character in input strings, increment count in
// the corresponding count array
for (i = 0; str1[i] && str2[i]; i++)
{
count[str1[i]]++;
count[str2[i]]--;
}
// If both strings are of different length. Removing this condition
// will make the program fail for strings like "aaca" and "aca"
if (str1[i] || str2[i])
return false;
// See if there is any non-zero value in count array
for (i = 0; i < NO_OF_CHARS; i++)
if (count[i])
return false;
return true;
}
// This function prints all anagram pairs in a given array of strigns
void findAllAnagrams(string arr[], int n)
{
for (int i = 0; i < n; i++)
for (int j = i+1; j < n; j++)
if (areAnagram(arr[i], arr[j]))
cout << arr[i] << " is anagram of " << arr[j] << endl;
}
/* Driver program to test to pront printDups*/
int main()
{
string arr[] = {"geeksquiz", "geeksforgeeks", "abcd",
"forgeeksgeeks", "zuiqkeegs"};
int n = sizeof(arr)/sizeof(arr[0]);
findAllAnagrams(arr, n);
return 0;
}

Output:
geeksquiz is anagram of zuiqkeegs
geeksforgeeks is anagram of forgeeksgeeks

The time complexity of the above solution is O(n2*m) where n is number of strings and m is maximum length of a string.
Optimizations:
We can optimize the above solution using following approaches.
1) Using sorting: We can sort array of strings so that all anagrams come together. Then print all anagrams by linearly traversing the sorted array.
The time complexity of this solution is O(mnLogn) (We would be doing O(nLogn) comparisons in sorting and a comparison would take O(m)
time)

2) Using Hashing: We can build a hash function like XOR or sum of ASCII values of all characters for a string. Using such a hash function, we
can build a hash table. While building the hash table, we can check if a value is already hashed. If yes, we can call areAnagrams() to check if two
strings are actually anagrams (Note that xor or sum of ASCII values is not sufficient, see Kaushik Leles comment here)