Numerical Methods: Jeffrey R. Chasnov

Numerical Methods
Jeffrey R. Chasnov
Adapted for :
Numerical Methods for Engineers
Click to view a promotional video

The Hong Kong University of Science and Technology
Department of Mathematics
Clear Water Bay, Kowloon
Hong Kong
Copyright ○
c 2012 by Jeffrey Robert Chasnov
This work is licensed under the Creative Commons Attribution 3.0 Hong Kong License. To
view a copy of this license, visit http://creativecommons.org/licenses/by/3.0/hk/ or send
a letter to Creative Commons, 171 Second Street, Suite 300, San Francisco, California, 94105,
USA.
Preface
What follows were my lecture notes for Math 3311: Introduction to Numerical Meth-
ods, taught at the Hong Kong University of Science and Technology. Math 3311,
with two lecture hours per week, was primarily for non-mathematics majors and
was required by several engineering departments.
I also have some free online courses on Coursera. A lot of time and effort has gone
into their production, and the video lectures for these courses are of high quality.
You can click on the links below to explore these courses.
If you want to learn differential equations, have a look at
Differential Equations for Engineers
If your interests are matrices and elementary linear algebra, try
Matrix Algebra for Engineers
If you want to learn vector calculus (also known as multivariable calculus, or calcu-
lus three), you can sign up for
Vector Calculus for Engineers
And if your interest is numerical methods, have a go at
Numerical Methods for Engineers
Jeffrey R. Chasnov
Hong Kong
February 2021
iii
Contents
1 IEEE Arithmetic 1
1.1 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Numbers with a decimal or binary point . . . . . . . . . . . . . . . . . 1
1.3 Examples of binary numbers . . . . . . . . . . . . . . . . . . . . . . . . 1
1.4 Hex numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.5 4-bit unsigned integers as hex numbers . . . . . . . . . . . . . . . . . . 1
1.6 IEEE single precision format: . . . . . . . . . . . . . . . . . . . . . . . . 2
1.7 Special numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.8 Examples of computer numbers . . . . . . . . . . . . . . . . . . . . . . 3
1.9 Inexact numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.9.1 Find smallest positive integer that is not exact in single precision 4
1.10 Machine epsilon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.11 IEEE double precision format . . . . . . . . . . . . . . . . . . . . . . . . 5
1.12 Roundoff error example . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2 Root Finding 7
2.1 Bisection Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2 Newton’s Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.3 Secant Method . √
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.3.1 Estimate 2 = 1.41421356 using Newton’s Method . . . . . . . 8
2.3.2 Example of fractals using Newton’s Method . . . . . . . . . . . 8
2.4 Order of convergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.4.1 Newton’s Method . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.4.2 Secant Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3 Systems of equations 13
3.1 Gaussian Elimination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.2 LU decomposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.3 Partial pivoting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.4 Operation counts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.5 System of nonlinear equations . . . . . . . . . . . . . . . . . . . . . . . 20
4 Least-squares approximation 23
4.1 Fitting a straight line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.2 Fitting to a linear combination of functions . . . . . . . . . . . . . . . . 24
5 Interpolation 27
5.1 Polynomial interpolation . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
5.1.1 Vandermonde polynomial . . . . . . . . . . . . . . . . . . . . . . 27
5.1.2 Lagrange polynomial . . . . . . . . . . . . . . . . . . . . . . . . 28
5.1.3 Newton polynomial . . . . . . . . . . . . . . . . . . . . . . . . . 28
5.2 Piecewise linear interpolation . . . . . . . . . . . . . . . . . . . . . . . . 29
5.3 Cubic spline interpolation . . . . . . . . . . . . . . . . . . . . . . . . . . 30
5.4 Multidimensional interpolation . . . . . . . . . . . . . . . . . . . . . . . 33
v
CONTENTS
6 Integration 35
6.1 Elementary formulas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
6.1.1 Midpoint rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
6.1.2 Trapezoidal rule . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
6.1.3 Simpson’s rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
6.2 Composite rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
6.2.1 Trapezoidal rule . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
6.2.2 Simpson’s rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
6.3 Local versus global error . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
6.4 Adaptive integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
7 Ordinary differential equations 41

7.1 Examples of analytical solutions . . . . . . . . . . . . . . . . . . . . . . 41
7.1.1 Initial value problem . . . . . . . . . . . . . . . . . . . . . . . . . 41
7.1.2 Boundary value problems . . . . . . . . . . . . . . . . . . . . . . 42
7.1.3 Eigenvalue problem . . . . . . . . . . . . . . . . . . . . . . . . . 43
7.2 Numerical methods: initial value problem . . . . . . . . . . . . . . . . 43
7.2.1 Euler method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
7.2.2 Modified Euler method . . . . . . . . . . . . . . . . . . . . . . . 44
7.2.3 Second-order Runge-Kutta methods . . . . . . . . . . . . . . . . 45
7.2.4 Higher-order Runge-Kutta methods . . . . . . . . . . . . . . . . 46
7.2.5 Adaptive Runge-Kutta Methods . . . . . . . . . . . . . . . . . . 47
7.2.6 System of differential equations . . . . . . . . . . . . . . . . . . 47
7.3 Numerical methods: boundary value problem . . . . . . . . . . . . . . 48
7.3.1 Finite difference method . . . . . . . . . . . . . . . . . . . . . . . 48
7.3.2 Shooting method . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
7.4 Numerical methods: eigenvalue problem . . . . . . . . . . . . . . . . . 51
7.4.1 Finite difference method . . . . . . . . . . . . . . . . . . . . . . . 51
7.4.2 Shooting method . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
vi CONTENTS
Chapter 1
IEEE Arithmetic
1.1 Definitions
Bit = 0 or 1
Byte = 8 bits
Word = Reals: 4 bytes (single precision)
8 bytes (double precision)
= Integers: 1, 2, 4, or 8 byte signed
1, 2, 4, or 8 byte unsigned
1.2 Numbers with a decimal or binary point
·
Decimal: 103 102 101 100 10−1 10−2 10−3 10−4
Binary: 23 22 21 20 2−1 2−2 2−3 2−4
1.3 Examples of binary numbers

Decimal Binary
1 1
2 10
3 11
4 100
0.5 0.1
1.5 1.1
1.4 Hex numbers

{0, 1, 2, 3, . . . , 9, 10, 11, 12, 13, 14, 15} = {0, 1, 2, 3.......9, a,b,c,d,e,f}
1.5 4-bit unsigned integers as hex numbers

Decimal Binary Hex
1 0001 1
2 0010 2
3 0011 3
.. .. ..
. . .
10 1010 a
.. .. ..
. . .
15 1111 f
1
1.6. IEEE SINGLE PRECISION FORMAT:
1.6 IEEE single precision format:
s e f
z}|{ z }| {z }| {
· · · · · · · ·
0 1 2 3 4 5 6 7 8 9 31
# = (−1)s × 2e−127 × 1.f
where s = sign
e = biased exponent
p=e-127 = exponent
1.f = significand (use binary point)
1.7 Special numbers

Smallest exponent: e = 0000 0000, represents denormal numbers (1.f → 0.f)
Largest exponent: e = 1111 1111, represents ±∞, if f = 0
e = 1111 1111, represents NaN, if f 6= 0
Number Range: e = 1111 1111 = 28 - 1 = 255 reserved

e = 0000 0000 = 0 reserved
so, p = e - 127 is
1 - 127 ≤ p ≤ 254-127
-126 ≤ p ≤ 127
Smallest positive normal number

= 1.0000 0000 · · · · ·· 0000× 2−126
' 1.2 × 10−38
bin: 0000 0000 1000 0000 0000 0000 0000 0000
hex: 00800000
MATLAB: realmin(’single’)
Largest positive number
= 1.1111 1111 · · · · ·· 1111× 2127
= (1 + (1 − 2−23 )) × 2127
' 2128 ' 3.4 × 1038
bin: 0111 1111 0111 1111 1111 1111 1111 1111
hex: 7f7fffff
MATLAB: realmax(’single’)
Zero
bin: 0000 0000 0000 0000 0000 0000 0000 0000
hex: 00000000
Subnormal numbers
Allow 1.f → 0.f (in software)
Smallest positive number = 0.0000 0000 · · · · · 0001 × 2−126
= 2−23 × 2−126 ' 1.4 × 10−45
2 CHAPTER 1. IEEE ARITHMETIC

1.8. EXAMPLES OF COMPUTER NUMBERS
1.8 Examples of computer numbers

What is 1.0, 2.0 & 1/2 in hex ?
1.0 = (−1)0 × 2(127−127) × 1.0

Therefore, s = 0, e = 0111 1111, f = 0000 0000 0000 0000 0000 000
bin: 0011 1111 1000 0000 0000 0000 0000 0000
hex: 3f80 0000
2.0 = (−1)0 × 2(128−127) × 1.0

Therefore, s = 0, e = 1000 0000, f = 0000 0000 0000 0000 0000 000
bin: 0100 00000 1000 0000 0000 0000 0000 0000
hex: 4000 0000
1/2 = (−1)0 × 2(126−127) × 1.0

Therefore, s = 0, e = 0111 1110, f = 0000 0000 0000 0000 0000 000
bin: 0011 1111 0000 0000 0000 0000 0000 0000
hex: 3f00 0000
1.9 Inexact numbers

Example:
1 1 1
= (−1)0 × × (1 + ),
3 4 3
so that p = e − 127 = −2 and e = 125 = 128 − 3, or in binary, e = 0111 1101. How is
f = 1/3 represented in binary? To compute binary number, multiply successively
by 2 as follows:
0.333 . . . 0.
0.666 . . . 0.0
1.333 . . . 0.01
0.666 . . . 0.010
1.333 . . . 0.0101
etc.
so that 1/3 exactly in binary is 0.010101 . . . . With only 23 bits to represent f , the
number is inexact and we have
f = 01010101010101010101011,
where we have rounded to the nearest binary number (here, rounded up). The
machine number 1/3 is then represented as
00111110101010101010101010101011
or in hex
3eaaaaab.
CHAPTER 1. IEEE ARITHMETIC 3

1.10. MACHINE EPSILON
1.9.1 Find smallest positive integer that is not exact in single pre-
cision
Let N be the smallest positive integer that is not exact. Now, I claim that
N − 2 = 223 × 1.11 . . . 1,
and
N − 1 = 224 × 1.00 . . . 0.
The integer N would then require a one-bit in the 2−24 position, which is not avail-
able. Therefore, the smallest positive integer that is not exact is 224 + 1 = 16 777 217.
In MATLAB, single(224 ) has the same value as single(224 + 1). Since single(224 + 1)
is exactly halfway between the two consecutive machine numbers 224 and 224 + 2,
MATLAB rounds to the number with a final zero-bit in f, which is 224 .
1.10 Machine epsilon

Machine epsilon (emach ) is the distance between 1 and the next largest number. If
0 ≤ δ < emach /2, then 1 + δ = 1 in computer math. Also since
x + y = x (1 + y/x ),
if 0 ≤ y/x < emach /2, then x + y = x in computer math.
Find emach
The number 1 in the IEEE format is written as
1 = 20 × 1.000 . . . 0,
with 23 0’s following the binary point. The number just larger than 1 has a 1 in the
23rd position after the decimal point. Therefore,
emach = 2−23 ≈ 1.192 × 10−7 .
What is the distance between 1 and the number just smaller than 1? Here, the
number just smaller than one can be written as
2−1 × 1.111 . . . 1 = 2−1 (1 + (1 − 2−23 )) = 1 − 2−24
Therefore, this distance is 2−24 = emach /2.
The spacing between numbers is uniform between powers of 2, with logarithmic
spacing of the powers of 2. That is, the spacing of numbers between 1 and 2 is 2−23 ,
between 2 and 4 is 2−22 , between 4 and 8 is 2−21 , etc. This spacing changes for
denormal numbers, where the spacing is uniform all the way down to zero.
Find the machine number just greater than 5

A rough estimate would be 5(1 + emach ) = 5 + 5emach , but this is not exact. The
exact answer can be found by writing
1
5 = 22 (1 + ),
4
so that the next largest number is
1
22 (1 + + 2−23 ) = 5 + 2−21 = 5 + 4emach .
4

1.11. IEEE DOUBLE PRECISION FORMAT
1.11 IEEE double precision format

Most computations take place in double precision, where round-off error is re-
duced, and all of the above calculations in single precision can be repeated for
double precision. The format is
s e f
z}|{ z }| {z }| {
· · · · · · · ·
0 1 2 3 4 5 6 7 8 9 10 11 12 63
# = (−1)s × 2e−1023 × 1.f
where s = sign
e = biased exponent
p=e-1023 = exponent
1.f = significand (use binary point)
1.12 Roundoff error example

Consider solving the quadratic equation
x2 + 2bx − 1 = 0,
where b is a parameter. The quadratic formula yields the two solutions

p
x± = −b ± b2 + 1.
Consider the solution with b > 0 and x > 0 (the x+ solution) given by
p
x = −b + b2 + 1. (1.1)
As b → ∞,
p
x = −b + b2 + 1
p
= −b + b 1 + 1/b2
p
= b( 1 + 1/b2 − 1)

1
≈ b 1+ 2 −1
2b
1
= .
2b
Now in double precision, realmin ≈ 2.2 × 10−308 and we would like x to be accurate
to this value before it goes to 0 via denormal numbers. Therefore, x should be
computed accurately to b ≈ 1/(2 × realmin) ≈ 2 × 10307 . What happens if we
compute (1.1) directly? √Then x = 0 when b2 + 1 = b2 , or 1 + 1/b2 = 1. That is
√
1/b2 = emach /2, or b = 2/ emach ≈ 108 .
CHAPTER 1. IEEE ARITHMETIC 5

1.12. ROUNDOFF ERROR EXAMPLE
For a subroutine written to compute the solution of a quadratic for a general

user, this is not good enough. The way for a software designer to solve this problem
is to compute the solution for x as
1
x= √ .
b (1 + 1 + 1/b2 )
In this form, if 1 + 1/b2 = 1, then x = 1/2b which is the correct asymptotic form.

Chapter 2
Root Finding
Solve f ( x ) = 0 for x, when an explicit analytical solution is impossible.
2.1 Bisection Method

The bisection method is the easiest to numerically implement and almost always
works. The main disadvantage is that convergence is slow. If the bisection method
results in a computer program that runs too slow, then other faster methods may
be chosen; otherwise it is a good choice of method.
We want to construct a sequence x0 , x1 , x2 , . . . that converges to the root x = r
that solves f ( x ) = 0. We choose x0 and x1 such that x0 < r < x1 . We say that x0
and x1 bracket the root. With f (r ) = 0, we want f ( x0 ) and f ( x1 ) to be of opposite
sign, so that f ( x0 ) f ( x1 ) < 0. We then assign x2 to be the midpoint of x0 and x1 ,
that is x2 = ( x0 + x1 )/2, or
x − x0
x2 = x0 + 1 .
2
The sign of f ( x2 ) can then be determined. The value of x3 is then chosen as either
the midpoint of x0 and x2 or as the midpoint of x2 and x1 , depending on whether
x0 and x2 bracket the root, or x2 and x1 bracket the root. The root, therefore,
stays bracketed at all times. The algorithm proceeds in this fashion and is typically
stopped when the increment to the left side of the bracket (above, given by ( x1 −
x0 )/2) is smaller than some required precision.
2.2 Newton’s Method

This is the fastest method, but requires analytical computation of the derivative of
f ( x ). Also, the method may not always converge to the desired root.
We can derive Newton’s Method graphically, or by a Taylor series. We again
want to construct a sequence x0 , x1 , x2 , . . . that converges to the root x = r. Consider
the xn+1 member of this sequence, and Taylor series expand f ( xn+1 ) about the point
xn . We have
f ( x n +1 ) = f ( x n ) + ( x n +1 − x n ) f 0 ( x n ) + . . . .
To determine xn+1 , we drop the higher-order terms in the Taylor series, and assume
f ( xn+1 ) = 0. Solving for xn+1 , we have
f ( xn )
x n +1 = x n − .
f 0 ( xn )
Starting Newton’s Method requires a guess for x0 , hopefully close to the root x = r.
2.3 Secant Method

The Secant Method is second best to Newton’s Method, and is used when a faster
convergence than Bisection is desired, but it is too difficult or impossible to take an
7
2.3. SECANT METHOD
analytical derivative of the function f ( x ). We write in place of f 0 ( xn ),
f ( x n ) − f ( x n −1 )
f 0 ( xn ) ≈ .
x n − x n −1
Starting the Secant Method requires a guess for both x0 and x1 .

√
2.3.1 Estimate 2 = 1.41421356 using Newton’s Method
√
The 2 is the zero of the function f ( x ) = x2 − 2. To implement Newton’s Method,
we use f 0 ( x ) = 2x. Therefore, Newton’s Method is the iteration
xn2 − 2
x n +1 = x n − .
2xn
We take as our initial guess x0 = 1. Then
−1 3
x1 = 1 − = = 1.5,
2 2
9
3 − 2 17
x2 = − 4 = = 1.416667,
2 3 12
172
17 2 −2 577
x3 = − 12 17 = = 1.41426.
12 6
408
2.3.2 Example of fractals using Newton’s Method

Consider the complex roots of the equation f (z) = 0, where
f (z) = z3 − 1.
These roots are the three cubic roots of unity. With
ei2πn = 1, n = 0, 1, 2, . . .
the three unique cubic roots of unity are given by
1, ei2π/3 , ei4π/3 .
With
eiθ = cos θ + i sin θ,
√
and cos (2π/3) = −1/2, sin (2π/3) = 3/2, the three cubic roots of unity are
√ √
1 3 1 3
r1 = 1, r2 = − + i, r3 = − − i.
2 2 2 2
The interesting idea here is to determine which initial values of z0 in the complex
plane converge to which of the three cubic roots of unity.
Newton’s method is
z3 − 1
z n +1 = z n − n 2 .
3zn
If the iteration converges to r1 , we color z0 red; r2 , blue; r3 , green. The result will
be shown in lecture.
8 CHAPTER 2. ROOT FINDING

2.4. ORDER OF CONVERGENCE
2.4 Order of convergence

Let r be the root and xn be the nth approximation to the root. Define the error as
en = r − x n .
If for large n we have the approximate relationship
| e n +1 | = k | e n | p ,
with k a positive constant, then we say the root-finding numerical method is of

order p. Larger values of p correspond to faster convergence to the root. The order
of convergence of bisection is one: the error is reduced by approximately a factor of
2 with each iteration so that
1
| e n +1 | = | e n | .
2
We now find the order of convergence for Newton’s Method and for the Secant
Method.
2.4.1 Newton’s Method

We start with Newton’s Method
f ( xn )
x n +1 = x n − .
f 0 ( xn )
Subtracting both sides from r, we have
f ( xn )
r − x n +1 = r − x n + ,
f 0 ( xn )
or
f ( xn )
e n +1 = e n + . (2.1)
f 0 ( xn )
We use Taylor series to expand the functions f ( xn ) and f 0 ( xn ) about the root r,
using f (r ) = 0. We have
1
f ( xn ) = f (r ) + ( xn − r ) f 0 (r ) + ( xn − r )2 f 00 (r ) + . . . ,
2
0 1 2 00
= − en f (r ) + en f (r ) + . . . ;
2 (2.2)
1
f 0 ( xn ) = f 0 (r ) + ( xn − r ) f 00 (r ) + ( xn − r )2 f 000 (r ) + . . . ,
2
1
= f 0 (r ) − en f 00 (r ) + en2 f 000 (r ) + . . . .
2
To make further progress, we will make use of the following standard Taylor series:
1
= 1 + e + e2 + . . . , (2.3)
1−e
CHAPTER 2. ROOT FINDING 9

which converges for |e| < 1. Substituting (2.2) into (2.1), and using (2.3) yields
f ( xn )
e n +1 = e n +
f 0 ( xn )
−en f 0 (r ) + 12 en2 f 00 (r ) + . . .
= en +
f 0 (r ) − en f 00 (r ) + 12 en2 f 000 (r ) + . . .
00
−en + 12 en2 ff 0 ((rr)) + . . .
= en + f 00 (r )
1 − en f 0 (r )
+...
1 f 00 (r ) f 00 (r )

= en + −en + en2 0 +... 1 + en 0 +...
2 f (r ) f (r )
00 00 (r )
1 f ( r ) f
= en + −en + en2 − + . . .
2 f 0 (r ) f 0 (r )
1 f 00 (r ) 2
=− e +...
2 f 0 (r ) n
Therefore, we have shown that
| e n +1 | = k | e n | 2
as n → ∞, with
1 f 00 (r )

k= ,
2 f 0 (r )
provided f 0 (r ) 6= 0. Newton’s method is thus of order 2 at simple roots.
2.4.2 Secant Method

Determining the order of the Secant Method proceeds in a similar fashion. We start
with
( x n − x n −1 ) f ( x n )
x n +1 = x n − .
f ( x n ) − f ( x n −1 )
We subtract both sides from r and make use of
x n − x n −1 = ( r − x n −1 ) − ( r − x n )
= e n −1 − e n ,
and the Taylor series
1
f ( xn ) = −en f 0 (r ) + en2 f 00 (r ) + . . . ,
2
1
0
f ( xn−1 ) = −en−1 f (r ) + en2 −1 f 00 (r ) + . . . ,
2
so that
1
f ( xn ) − f ( xn−1 ) = (en−1 − en ) f 0 (r ) + (en2 − en2 −1 ) f 00 (r ) + . . .
2
1
= (en−1 − en ) f (r ) − (en−1 + en ) f 00 (r ) + . . . .
0
2

We therefore have
−en f 0 (r ) + 21 en2 f 00 (r ) + . . .
e n +1 = e n +
f 0 (r ) − 12 (en−1 + en ) f 00 (r ) + . . .
f 00 (r )
1 − 21 en f 0 (r )
+...
= en − en f 00 (r )
1 − 12 (en−1 + en ) f 0 (r) + . . .
1 f 00 (r ) f 00 (r )

1
= en − en 1 − en 0 +... 1 + ( e n −1 + e n ) 0 +...
2 f (r ) 2 f (r )
00
1 f (r )
=− e en + . . . ,
2 f 0 ( r ) n −1
or to leading order
1 f 00 (r )

| e n +1 | = 0 |e ||en |. (2.4)
2 f ( r ) n −1
The order of convergence is not yet obvious from this equation, and to determine
the scaling law we look for a solution of the form
| e n +1 | = k | e n | p .
From this ansatz, we also have
| e n | = k | e n −1 | p ,
and therefore
2
| e n +1 | = k p +1 | e n −1 | p .
Substitution into (2.4) results in
k f 00 (r )

2
k p +1 | e n −1 | p = | e | p +1 .
2 f 0 ( r ) n −1
Equating the coefficient and the power of en−1 results in
1 f 00 (r )

k p = 0 ,
2 f (r )
and
p2 = p + 1.
The order of convergence of the Secant Method, given by p, therefore is determined
to be the positive root of the quadratic equation p2 − p − 1 = 0, or
√
1+ 5
p= ≈ 1.618,
2
which coincidentally is a famous irrational number that is called The Golden Ra-
tio, and goes by the symbol Φ. We see that the Secant Method has an order of
convergence lying between the Bisection Method and Newton’s Method.
CHAPTER 2. ROOT FINDING 11


Chapter 3
Systems of equations
Consider the system of n linear equations and n unknowns, given by
a11 x1 + a12 x2 + · · · + a1n xn = b1 ,

a21 x1 + a22 x2 + · · · + a2n xn = b2 ,
.. ..
. .
an1 x1 + an2 x2 + · · · + ann xn = bn .
We can write this system as the matrix equation
Ax = b,
with      
a11 a12 ··· a1n x1 b1
 a21 a22 ··· a2n   x2   b2 
A= . ..  , x =  . , b =  . .
     
.. ..
 .. . . .   ..   .. 
an1 an2 ··· ann xn bn
3.1 Gaussian Elimination

The standard numerical algorithm to solve a system of linear equations is called
Gaussian Elimination. We can illustrate this algorithm by example.
Consider the system of equations
−3x1 + 2x2 − x3 = −1,

6x1 − 6x2 + 7x3 = −7,
3x1 − 4x2 + 4x3 = −6.
To perform Gaussian elimination, we form an Augmented Matrix by combining the

matrix A with the column vector b:
 
−3 2 −1 −1
 6 −6 7 −7 .
3 −4 4 −6
Row reduction is then performed on this matrix. Allowed operations are (1) mul-
tiply any row by a constant, (2) add multiple of one row to another row, (3) inter-
change the order of any rows. The goal is to convert the original matrix into an
upper-triangular matrix.
We start with the first row of the matrix and work our way down as follows.
First we multiply the first row by 2 and add it to the second row, and add the first
row to the third row:  
−3 2 −1 −1
 0 −2 5 −9 .
0 −2 3 −7
13
3.2. LU DECOMPOSITION
We then go to the second row. We multiply this row by −1 and add it to the third
row:  
−3 2 −1 −1
 0 −2 5 −9 .
0 0 −2 2
The resulting equations can be determined from the matrix and are given by
−3x1 + 2x2 − x3 = −1
−2x2 + 5x3 = −9
−2x3 = 2.
These equations can be solved by backward substitution, starting from the last equa-
tion and working backwards. We have
−2x3 = 2 → x3 = −1
−2x2 = −9 − 5x3 = −4 → x2 = 2,
−3x1 = −1 − 2x2 + x3 = −6 → x1 = 2.
Therefore,    
x1 2
 x2  =  2  .
x3 −1
3.2 LU decomposition
The process of Gaussian Elimination also results in the factoring of the matrix A to
A = LU,
where L is a lower triangular matrix and U is an upper triangular matrix. Using the
same matrix A as in the last section, we show how this factorization is realized. We
have    
−3 2 −1 −3 2 −1
 6 −6 7  →  0 −2 5  = M1 A,
3 −4 4 0 −2 3
where     
1 0 0 −3 2 −1 −3 2 −1
M1 A =  2 1 0  6 −6 7  =  0 −2 5 .
1 0 1 3 −4 4 0 −2 3
Note that the matrix M1 performs row elimination on the first column. Two times
the first row is added to the second row and one times the first row is added to
the third row. The entries of the column of M1 come from 2 = −(6/ − 3) and
1 = −(3/ − 3) as required for row elimination. The number −3 is called the pivot.
The next step is
   
−3 2 −1 −3 2 −1
 0 −2 5 →  0 −2 5 = M2 (M1 A),
0 −2 3 0 0 −2
where
    
1 0 0 −3 2 −1 −3 2 −1
M2 (M1 A) =  0 1 0  0 −2 5 =  0 −2 5 .
0 −1 1 0 −2 3 0 0 −2
14 CHAPTER 3. SYSTEMS OF EQUATIONS

3.2. LU DECOMPOSITION
Here, M2 multiplies the second row by −1 = −(−2/ − 2) and adds it to the third
row. The pivot is −2.
We now have
M2 M1 A = U
or
A = M1−1 M2−1 U.
The inverse matrices are easy to find. The matrix M1 multiples the first row by 2
and adds it to the second row, and multiplies the first row by 1 and adds it to the
third row. To invert these operations, we need to multiply the first row by −2 and
add it to the second row, and multiply the first row by −1 and add it to the third
row. To check, with
M1 M1−1 = I,
we have     
1 0 0 1 0 0 1 0 0
 2 1 0  −2 1 0 = 0 1 0 .
1 0 1 −1 0 1 0 0 1
Similarly,  
1 0 0
M2−1 = 0 1 0
0 1 1
Therefore,
L = M1−1 M2−1
is given by
    
1 0 0 1 0 0 1 0 0
L =  −2 1 0 0 1 0 =  −2 1 0 ,
−1 0 1 0 1 1 −1 1 1
which is lower triangular. The off-diagonal elements of M1−1 and M2−1 are simply
combined to form L. Our LU decomposition is therefore
    
−3 2 −1 1 0 0 −3 2 −1
 6 −6 7  =  −2 1 0  0 −2 5 .
3 −4 4 −1 1 1 0 0 −2
Another nice feature of the LU decomposition is that it can be done by overwriting

A, therefore saving memory if the matrix A is very large.
The LU decomposition is useful when one needs to solve Ax = b for x when
A is fixed and there are many different b’s. First one determines L and U using
Gaussian elimination. Then one writes
(LU)x = L(Ux) = b.
We let
y = Ux,
and first solve
Ly = b
for y by forward substitution. We then solve
Ux = y
CHAPTER 3. SYSTEMS OF EQUATIONS 15

3.3. PARTIAL PIVOTING
for x by backward substitution. When we count operations, we will see that solving
(LU)x = b is significantly faster once L and U are in hand than solving Ax = b
directly by Gaussian elimination.
We now illustrate the solution of LUx = b using our previous example, where
     
1 0 0 −3 2 −1 −1
L =  −2 1 0 , U =  0 −2 5 , b =  −7 .
−1 1 1 0 0 −2 −6
With y = Ux, we first solve Ly = b, that is

    
1 0 0 y1 −1
 −2 1 0  y2  =  −7 .
−1 1 1 y3 −6
Using forward substitution
y1 = −1,
y2 = −7 + 2y1 = −9,
y3 = −6 + y1 − y2 = 2.
We now solve Ux = y, that is

    
−3 2 −1 x1 −1
 0 −2 5  x2  =  −9 .
0 0 −2 x3 2
Using backward substitution,
−2x3 = 2 → x3 = −1,
−2x2 = −9 − 5x3 = −4 → x2 = 2,
−3x1 = −1 − 2x2 + x3 = −6 → x1 = 2,
and we have once again determined

   
x1 2
 x2  =  2 .
x3 −1
3.3 Partial pivoting

When performing Gaussian elimination, the diagonal element that one uses during
the elimination procedure is called the pivot. To obtain the correct multiple, one
uses the pivot as the divisor to the elements below the pivot. Gaussian elimination
in this form will fail if the pivot is zero. In this situation, a row interchange must
be performed.
Even if the pivot is not identically zero, a small value can result in big round-
off errors. For very large matrices, one can easily lose all accuracy in the solution.
To avoid these round-off errors arising from small pivots, row interchanges are
made, and this technique is called partial pivoting (partial pivoting is in contrast
to complete pivoting, where both rows and columns are interchanged). We will
illustrate by example the LU decomposition using partial pivoting.

3.3. PARTIAL PIVOTING
Consider  
−2 2 −1
A =  6 −6 7 .
3 −8 4
We interchange rows to place the largest element (in absolute value) in the pivot, or
a11 , position. That is,  
6 −6 7
A →  −2 2 −1 = P12 A,
3 −8 4
where  
0 1 0
P12 = 1 0 0
0 0 1
is a permutation matrix that when multiplied on the left interchanges the first and
−1
second rows of a matrix. Note that P12 = P12 . The elimination step is then
 
6 −6 7
P12 A → 0 0 4/3 = M1 P12 A,
0 −5 1/2
where  
1 0 0
M1 =  1/3 1 0 .
−1/2 0 1
The final step requires one more row interchange:
 
6 −6 7
M1 P12 A → 0 −5 1/2 = P23 M1 P12 A = U.
0 0 4/3
Since the permutation matrices given by P are their own inverses, we can write our
result as
(P23 M1 P23 )P23 P12 A = U.
Multiplication of M on the left by P interchanges rows while multiplication on the
right by P interchanges columns. That is,
     
1 0 0 1 0 0 1 0 0
P23  1/3 1 0 P23 = −1/2 0 1 P23 = −1/2 1 0 .
−1/2 0 1 1/3 1 0 1/3 0 1
The net result on M1 is an interchange of the nondiagonal elements 1/3 and −1/2.
We can then multiply by the inverse of (P23 M1 P23 ) to obtain
P23 P12 A = (P23 M1 P23 )−1 U,
which we write as
PA = LU.
Instead of L, MATLAB will write this as
A = (P−1 L)U.

3.4. OPERATION COUNTS
For convenience, we will just denote (P−1 L) by L, but by L here we mean a permu-
tated lower triangular matrix.
For example, in MATLAB, to solve Ax = b for x using Gaussian elimination,
one types
x = A \ b;
where \ solves for x using the most efficient algorithm available, depending on the
form of A. If A is a general n × n matrix, then first the LU decomposition of A is
found using partial pivoting, and then x is determined from permuted forward and
backward substitution. If A is upper or lower triangular, then forward or backward
substitution (or their permuted version) is used directly.
If there are many different right-hand-sides, one would first directly find the
LU decomposition of A using a function call, and then solve using \. That is, one
would iterate for different b’s the following expressions:
[LU] = lu(A);
y = L \ b;
x = U \ y;
where the second and third lines can be shortened to
x = U \ (L \ b);
where the parenthesis are required. In lecture, I will demonstrate these solutions in
MATLAB using the matrix A = [−2, 2, −1; 6, −6, 7; 3, −8, 4]; which is the example
in the notes.
3.4 Operation counts

To estimate how much computational time is required for an algorithm, one can
count the number of operations required (multiplications, divisions, additions and
subtractions). Usually, what is of interest is how the algorithm scales with the size
of the problem. For example, suppose one wants to multiply two full n × n matrices.
The calculation of each element requires n multiplications and n − 1 additions, or
say 2n − 1 operations. There are n2 elements to compute so that the total operation
count is n2 (2n − 1). If n is large, we might want to know what will happen to
the computational time if n is doubled. What matters most is the fastest-growing,
leading-order term in the operation count. In this matrix multiplication example,
the operation count is n2 (2n − 1) = 2n3 − n2 , and the leading-order term is 2n3 .
The factor of 2 is unimportant for the scaling, and we say that the algorithm scales
like O(n3 ), which is read as “big Oh of n cubed.” When using the big-Oh notation,
we will drop both lower-order terms and constant multipliers.
The big-Oh notation tells us how the computational time of an algorithm scales.
For example, suppose that the multiplication of two large n × n matrices took a
computational time of Tn seconds. With the known operation count going like
O(n3 ), we can write
Tn = kn3
for some unknown constant k. To determine how much longer the multiplication

3.4. OPERATION COUNTS
of two 2n × 2n matrices will take, we write
T2n = k(2n)3
= 8kn3
= 8Tn ,
so that doubling the size of the matrix is expected to increase the computational
time by a factor of 23 = 8.
Running MATLAB on my computer, the multiplication of two 2048 × 2048 ma-
trices took about 0.75 sec. The multiplication of two 4096 × 4096 matrices took about
6 sec, which is 8 times longer. Timing of code in MATLAB can be found using the
built-in stopwatch functions tic and toc.
What is the operation count and therefore the scaling of Gaussian elimination?
Consider an elimination step with the pivot in the ith row and ith column. There
are both n − i rows below the pivot and n − i columns to the right of the pivot. To
perform elimination of one row, each matrix element to the right of the pivot must
be multiplied by a factor and added to the row underneath. This must be done for
all the rows. There are therefore (n − i )(n − i ) multiplication-additions to be done
for this pivot. Since we are interested in only the scaling of the algorithm, I will just
count a multiplication-addition as one operation.
To find the total operation count, we need to perform elimination using n − 1
pivots, so that
n −1
op. counts = ∑ ( n − i )2
i =1
= ( n − 1)2 + ( n − 2)2 + . . . (1)2
n −1
= ∑ i2 .
i =1
Three summation formulas will come in handy. They are

n
∑ 1 = n,
i =1
n
1
∑ i = 2 n ( n + 1),
i =1
n
1
∑ i2 = 6 n(2n + 1)(n + 1),
i =1
which can be proved by mathematical induction, or derived by some tricks.

The operation count for Gaussian elimination is therefore
n −1
op. counts = ∑ i2
i =1
1
= (n − 1)(2n − 1)(n).
6
The leading-order term is therefore n3 /3, and we say that Gaussian elimination
scales like O(n3 ). Since LU decomposition with partial pivoting is essentially Gaus-
sian elimination, that algorithm also scales like O(n3 ).

3.5. SYSTEM OF NONLINEAR EQUATIONS
However, once the LU decomposition of a matrix A is known, the solution of

Ax = b can proceed by a forward and backward substitution. How does a back-
ward substitution, say, scale? For backward substitution, the matrix equation to be
solved is of the form
    
a1,1 a1,2 · · · a1,n−1 a1,n x1 b1
 0
 a2,2 · · · a2,n−1 a2,n   x2   b2 
   
 .. .. .. .
.. . .
..   ..  =  ..  .
.
   
 .
 . .    
 0 0 · · · an−1,n−1 an−1,n   xn−1  bn−1 
0 0 ··· 0 an,n xn bn
The solution for xi is found after solving for x j with j > i. The explicit solution for
xi is given by !
n
1
xi = bi − ∑ ai,j x j .
ai,i j = i +1
The solution for xi requires n − i + 1 multiplication-additions, and since this must

be done for n such i0 s, we have
n
op. counts = ∑n−i+1
i =1
= n + ( n − 1) + · · · + 1
n
= ∑i
i =1
1
= n ( n + 1).
2
The leading-order term is n2 /2 and the scaling of backward substitution is O(n2 ).
After the LU decomposition of a matrix A is found, only a single forward and back-
ward substitution is required to solve Ax = b, and the scaling of the algorithm to
solve this matrix equation is therefore still O(n2 ). For large n, one should expect
that solving Ax = b by a forward and backward substitution should be substan-
tially faster than a direct solution using Gaussian elimination.
3.5 System of nonlinear equations

A system of nonlinear equations can be solved using a version of Newton’s Method.
We illustrate this method for a system of two equations and two unknowns. Sup-
pose that we want to solve
f ( x, y) = 0, g( x, y) = 0,
for the unknowns x and y. We want to construct two simultaneous sequences
x0 , x1 , x2 , . . . and y0 , y1 , y2 , . . . that converge to the roots. To construct these se-
quences, we Taylor series expand f ( xn+1 , yn+1 ) and g( xn+1 , yn+1 ) about the point
( xn , yn ). Using the partial derivatives f x = ∂ f /∂x, f y = ∂ f /∂y, etc., the two-
dimensional Taylor series expansions, displaying only the linear terms, are given
by
f ( x n +1 , y n +1 ) = f ( x n , y n ) + ( x n +1 − x n ) f x ( x n , y n )
+ ( y n +1 − y n ) f y ( x n , y n ) + . . .

g ( x n +1 , y n +1 ) = g ( x n , y n ) + ( x n +1 − x n ) g x ( x n , y n )
+ ( y n +1 − y n ) g y ( x n , y n ) + . . . .
To obtain Newton’s method, we take f ( xn+1 , yn+1 ) = 0, g( xn+1 , yn+1 ) = 0 and drop
higher-order terms above linear. Although one can then find a system of linear
equations for xn+1 and yn+1 , it is more convenient to define the variables
∆xn = xn+1 − xn , ∆yn = yn+1 − yn .
The iteration equations will then be given by
xn+1 = xn + ∆xn , yn+1 = yn + ∆yn ;
and the linear equations to be solved for ∆xn and ∆yn are given by
∆xn

fx fy −f
= ,
g x gy ∆yn −g
where f , g, f x , f y , gx , and gy are all evaluated at the point ( xn , yn ). The two-
dimensional case is easily generalized to n dimensions. The matrix of partial deriva-
tives is called the Jacobian Matrix.
We illustrate Newton’s Method by finding the steady state solution of the Lorenz
equations, given by
σ (y − x ) = 0,
rx − y − xz = 0,
xy − bz = 0,
where x, y, and z are the unknown variables and σ, r, and b are the known param-
eters. Here, we have a three-dimensional homogeneous system f = 0, g = 0, and
h = 0, with
f ( x, y, z) = σ(y − x ),
g( x, y, z) = rx − y − xz,
h( x, y, z) = xy − bz.
The partial derivatives can be computed to be
f x = −σ, f y = σ, f z = 0,
gx = r − z, gy = −1, gz = − x,
h x = y, hy = x, hz = −b.
The iteration equation is therefore

∆xn
    
−σ σ 0 σ (yn − xn )
r − zn −1 − xn   ∆yn  = − rxn − yn − xn zn  ,
yn xn −b ∆zn xn yn − bzn
with
xn+1 = xn + ∆xn ,
yn+1 = yn + ∆yn ,
zn+1 = zn + ∆zn .
The MATLAB program that solves this system is contained in newton_system.m.


Chapter 4
Least-squares approximation
The method of least-squares is commonly used to fit a parameterized curve to
experimental data. In general, the fitting curve is not expected to pass through the
data points, making this problem substantially different from the one of interpola-
tion.
We consider here only the simplest case of the same experimental error for all
the data points. Let the data to be fitted be given by ( xi , yi ), with i = 1 to n.
4.1 Fitting a straight line

Suppose the fitting curve is a line. We write for the fitting curve
y( x ) = αx + β.
The distance ri from the data point ( xi , yi ) and the fitting curve is given by
ri = yi − y ( xi )
= yi − (αxi + β).
A least-squares fit minimizes the sum of the squares of the ri ’s. This minimum can
be shown to result in the most probable values of α and β.
We define
n
ρ= ∑ ri2
i =1
n 2
= ∑ yi − (αxi + β) .
i =1
To minimize ρ with respect to α and β, we solve

∂ρ ∂ρ
= 0, = 0.
∂α ∂β
Taking the partial derivatives, we have
n
∂ρ
= ∑ 2(− xi ) yi − (αxi + β) = 0,

∂α i =1
n
∂ρ
= ∑ 2(−1) yi − (αxi + β) = 0.

∂β i =1
These equations form a system of two linear equations in the two unknowns α and
β, which is evident when rewritten in the form
n n n
α ∑ xi2 + β ∑ xi = ∑ xi yi ,
i =1 i =1 i =1
n n
α ∑ xi + βn = ∑ yi .
i =1 i =1
23
4.2. FITTING TO A LINEAR COMBINATION OF FUNCTIONS
These equations can be solved either analytically, or numerically in MATLAB, where

the matrix form is
n n
∑i=1 xi2 ∑in=1 xi ∑ i =1 x i y i

α
= .
∑in=1 xi n β ∑in=1 yi
A proper statistical treatment of this problem should also consider an estimate of
the errors in α and β as well as an estimate of the goodness-of-fit of the data to the
model. We leave these further considerations to a statistics class.
4.2 Fitting to a linear combination of functions

Consider the general fitting function
m
y( x ) = ∑ c j f j ( x ),
j =1
where we assume m functions f j ( x ). For example, if we want to fit a cubic poly-

nomial to the data, then we would have m = 4 and take f 1 = 1, f 2 = x, f 3 = x2
and f 4 = x3 . Typically, the number of functions f j is less than the number of data
points; that is, m < n, so that a direct attempt to solve for the c j ’s such that the fit-
ting function exactly passes through the n data points would result in n equations
and m unknowns. This would be an over-determined linear system that in general
has no solution.
We now define the vectors
   
y1 c1
 y2   c2 
y =  . , c =  . ,
   
..  .. 
yn cm
and the n-by-m matrix

 
f 1 ( x1 ) f 2 ( x1 ) ··· f m ( x1 )
 f 1 ( x2 ) f 2 ( x2 ) ··· f m ( x2 ) 
A= . ..  . (4.1)
 
.. ..
 .. . . . 
f 1 ( xn ) f 2 ( xn ) ··· f m ( xn )
With m < n, the equation Ac = y is over-determined. We let
r = y − Ac
be the residual vector, and let

n
ρ= ∑ ri2 .
i =1
The method of least squares minimizes ρ with respect to the components of c. Now,
using T to signify the transpose of a matrix, we have
ρ = rT r
= (y − Ac)T (y − Ac)
= y T y − c T AT y − y T Ac + c T AT Ac.
24 CHAPTER 4. LEAST-SQUARES APPROXIMATION

Since ρ is a scalar, each term in the above expression must be a scalar, and since the
transpose of a scalar is equal to the scalar, we have
T
c T AT y = c T AT y = y T Ac.
Therefore,
ρ = y T y − 2y T Ac + c T AT Ac.
To find the minimum of ρ, we will need to solve ∂ρ/∂c j = 0 for j = 1, . . . , m.
To take the derivative of ρ, we switch to a tensor notation, using the Einstein sum-
mation convention, where repeated indices are summed over their allowable range.
We can write
T
ρ = yi yi − 2yi Aik ck + ci Aik Akl cl .
Taking the partial derivative, we have
∂ρ ∂c ∂c T ∂c
= −2yi Aik k + i Aik T
Akl cl + ci Aik Akl l .
∂c j ∂c j ∂c j ∂c j
Now, (
∂ci 1, if i = j;
=
∂c j 0, otherwise.
Therefore,
∂ρ
= −2yi Aij + ATjk Akl cl + ci Aik
T
Akj .
∂c j
Now,
T
ci Aik Akj = ci Aki Akj
= Akj Aki ci
= ATjk Aki ci
= ATjk Akl cl .
Therefore,
∂ρ
= −2yi Aij + 2ATjk Akl cl .
∂c j
With the partials set equal to zero, we have
ATjk Akl cl = yi Aij ,
or
ATjk Akl cl = ATji yi ,
In vector notation, we have
AT Ac = AT y. (4.2)
Equation (4.2) is the so-called normal equation, and can be solved for c by Gaus-
sian elimination using the MATLAB backslash operator. After constructing the
matrix A given by (4.1), and the vector y from the data, one can code in MATLAB
c = ( A0 A)\( A0 y);
But in fact the MATLAB back slash operator will automatically solve the normal
equations when the matrix A is not square, so that the MATLAB code
c = A\y;
yields the same result.
CHAPTER 4. LEAST-SQUARES APPROXIMATION 25

26 CHAPTER 4. LEAST-SQUARES APPROXIMATION

Chapter 5
Interpolation
Consider the following problem: Given the values of a known function y = f ( x )
at a sequence of ordered points x0 , x1 , . . . , xn , find f ( x ) for arbitrary x. When x0 ≤
x ≤ xn , the problem is called interpolation. When x < x0 or x > xn the problem is
called extrapolation.
With yi = f ( xi ), the problem of interpolation is basically one of drawing a
smooth curve through the known points ( x0 , y0 ), ( x1 , y1 ), . . . , ( xn , yn ). This is not the
same problem as drawing a smooth curve that approximates a set of data points that
have experimental error. This latter problem is called least-squares approximation.
Here, we will consider three interpolation algorithms: (1) polynomial interpola-
tion; (2) piecewise linear interpolation, and; (3) cubic spline interpolation.
5.1 Polynomial interpolation

The n + 1 points ( x0 , y0 ), ( x1 , y1 ), . . . , ( xn , yn ) can be interpolated by a unique poly-
nomial of degree n. When n = 1, the polynomial is a linear function; when n = 2,
the polynomial is a quadratic function. There are three standard algorithms that
can be used to construct this unique interpolating polynomial, and we will present
all three here, not so much because they are all useful, but because it is interesting
to learn how these three algorithms are constructed.
When discussing each algorithm, we define Pn ( x ) to be the unique nth degree
polynomial that passes through the given n + 1 data points.
5.1.1 Vandermonde polynomial

This Vandermonde polynomial is the most straightforward construction of the in-
terpolating polynomial Pn ( x ). We simply write
Pn ( x ) = c0 x n + c1 x n−1 + · · · + cn .
Then we can immediately form n + 1 linear equations for the n + 1 unknown coef-
ficients c0 , c1 , . . . , cn using the n + 1 known points:
y0 = c0 x0n + c1 x0n−1 + · · · + cn−1 x0 + cn

y2 = c0 x1n + c1 x1n−1 + · · · + cn−1 x1 + cn
.. .. ..
. . .
yn = c0 xnn + c1 xnn−1 + · · · + cn−1 xn + cn .
The system of equations in matrix form is
x0n x0n−1 ···

    
x0 1 c0 y0
 xn
 1 x1n−1 ··· x1 1  c1   y1 
  ..  =  ..  .
   
 . .. .. ..
 .. . . .  .   . 
xnn xnn−1 ··· xn 1 cn yn
27
5.1. POLYNOMIAL INTERPOLATION
The matrix is called the Vandermonde matrix, and can be constructed using the
MATLAB function vander.m. The system of linear equations can be solved in MAT-
LAB using the \ operator, and the MATLAB function polyval.m can used to inter-
polate using the c coefficients. I will illustrate this in class and place the code on
the website.
5.1.2 Lagrange polynomial
The Lagrange polynomial is the most clever construction of the interpolating poly-
nomial Pn ( x ), and leads directly to an analytical formula. The Lagrange polynomial
is the sum of n + 1 terms and each term is itself a polynomial of degree n. The full
polynomial is therefore of degree n. Counting from 0, the ith term of the Lagrange
polynomial is constructed by requiring it to be zero at x j with j 6= i, and equal to yi
when j = i. The polynomial can be written as
( x − x1 )( x − x2 ) · · · ( x − xn )y0 ( x − x0 )( x − x2 ) · · · ( x − xn )y1
Pn ( x ) = +
( x0 − x1 )( x0 − x2 ) · · · ( x0 − xn ) ( x1 − x0 )( x1 − x2 ) · · · ( x1 − xn )
( x − x0 )( x − x1 ) · · · ( x − xn−1 )yn
+···+ .
( xn − x0 )( xn − x1 ) · · · ( xn − xn−1 )
It can be clearly seen that the first term is equal to zero when x = x1 , x2 , . . . , xn and
equal to y0 when x = x0 ; the second term is equal to zero when x = x0 , x2 , . . . xn and
equal to y1 when x = x1 ; and the last term is equal to zero when x = x0 , x1 , . . . xn−1
and equal to yn when x = xn . The uniqueness of the interpolating polynomial
implies that the Lagrange polynomial must be the interpolating polynomial.
5.1.3 Newton polynomial
The Newton polynomial is somewhat more clever than the Vandermonde polyno-
mial because it results in a system of linear equations that is lower triangular, and
therefore can be solved by forward substitution. The interpolating polynomial is
written in the form
Pn ( x ) = c0 + c1 ( x − x0 ) + c2 ( x − x0 )( x − x1 ) + · · · + cn ( x − x0 ) · · · ( x − xn−1 ),
which is clearly a polynomial of degree n. The n + 1 unknown coefficients given by

the c’s can be found by substituting the points ( xi , yi ) for i = 0, . . . , n:
y0 = c0 ,
y1 = c0 + c1 ( x1 − x0 ),
y2 = c0 + c1 ( x2 − x0 ) + c2 ( x2 − x0 )( x2 − x1 ),
.. .. ..
. . .
yn = c0 + c1 ( xn − x0 ) + c2 ( xn − x0 )( xn − x1 ) + · · · + cn ( xn − x0 ) · · · ( xn − xn−1 ).
28 CHAPTER 5. INTERPOLATION
5.2. PIECEWISE LINEAR INTERPOLATION
This system of linear equations is lower triangular as can be seen from the matrix
form
  
1 0 0 ··· 0 c0
1
 ( x1 − x0 ) 0 ··· 0   c1 
 
 .. .. .. .. ..   .. 
. . . . .  . 
1 ( xn − x0 ) ( xn − x0 )( xn − x1 ) · · · ( x n − x 0 ) · · · ( x n − x n −1 ) cn
 
y0
 y1 
=  . ,
 
 .. 
yn
and so theoretically can be solved faster than the Vandermonde polynomial. In

practice, however, there is little difference because polynomial interpolation is only
useful when the number of points to be interpolated is small.
5.2 Piecewise linear interpolation

Instead of constructing a single global polynomial that goes through all the points,
one can construct local polynomials that are then connected together. In the the
section following this one, we will discuss how this may be done using cubic poly-
nomials. Here, we discuss the simpler case of linear polynomials. This is the default
interpolation typically used when plotting data.
Suppose the interpolating function is y = g( x ), and as previously, there are
n + 1 points to interpolate. We construct the function g( x ) out of n local linear
polynomials. We write
g ( x ) = gi ( x ) , for xi ≤ x ≤ xi+1 ,
where
g i ( x ) = a i ( x − x i ) + bi ,
and i = 0, 1, . . . , n − 1.
We now require y = gi ( x ) to pass through the endpoints ( xi , yi ) and ( xi+1 , yi+1 ).
We have
y i = bi ,
y i + 1 = a i ( x i + 1 − x i ) + bi .
Therefore, the coefficients of gi ( x ) are determined to be
y i +1 − y i
ai = , bi = y i .
x i +1 − x i
Although piecewise linear interpolation is widely used, particularly in plotting rou-

tines, it suffers from a discontinuity in the derivative at each point. This results in a
function which may not look smooth if the points are too widely spaced. We next
consider a more challenging algorithm that uses cubic polynomials.
CHAPTER 5. INTERPOLATION 29
5.3. CUBIC SPLINE INTERPOLATION
5.3 Cubic spline interpolation

The n + 1 points to be interpolated are again
( x0 , y0 ), ( x1 , y1 ), . . . ( x n , y n ).
Here, we use n piecewise cubic polynomials for interpolation,
g i ( x ) = a i ( x − x i ) 3 + bi ( x − x i ) 2 + c i ( x − x i ) + d i , i = 0, 1, . . . , n − 1,
with the global interpolation function written as
g ( x ) = gi ( x ) , for xi ≤ x ≤ xi+1 .
To achieve a smooth interpolation we impose that g( x ) and its first and second
derivatives are continuous. The requirement that g( x ) is continuous (and goes
through all n + 1 points) results in the two constraints
gi ( x i ) = y i , i = 0 to n − 1, (5.1)
gi ( x i + 1 ) = y i + 1 , i = 0 to n − 1. (5.2)
The requirement that g0 ( x ) is continuous results in
gi0 ( xi+1 ) = gi0+1 ( xi+1 ), i = 0 to n − 2. (5.3)
And the requirement that g00 ( x ) is continuous results in
gi00 ( xi+1 ) = gi00+1 ( xi+1 ), i = 0 to n − 2. (5.4)
There are n cubic polynomials gi ( x ) and each cubic polynomial has four free co-
efficients; there are therefore a total of 4n unknown coefficients. The number of
constraining equations from (5.1)-(5.4) is 2n + 2(n − 1) = 4n − 2. With 4n − 2 con-
straints and 4n unknowns, two more conditions are required for a unique solution.
These are usually chosen to be extra conditions on the first g0 ( x ) and last gn−1 ( x )
polynomials. We will discuss these extra conditions later.
We now proceed to determine equations for the unknown coefficients of the
cubic polynomials. The polynomials and their first two derivatives are given by
g i ( x ) = a i ( x − x i ) 3 + bi ( x − x i ) 2 + c i ( x − x i ) + d i , (5.5)
gi0 ( x ) = 3ai ( x − xi )2 + 2bi ( x − xi ) + ci , (5.6)
gi00 ( x ) = 6ai ( x − xi ) + 2bi . (5.7)
We will consider the four conditions (5.1)-(5.4) in turn. From (5.1) and (5.5), we
have
di = yi , i = 0 to n − 1, (5.8)
which directly solves for all of the d-coefficients.
To satisfy (5.2), we first define
h i = x i +1 − x i ,
and
f i = y i +1 − y i .
Now, from (5.2) and (5.5), using (5.8), we obtain the n equations
ai h3i + bi h2i + ci hi = f i , i = 0 to n − 1. (5.9)
From (5.3) and (5.6) we obtain the n − 1 equations
3ai h2i + 2bi hi + ci = ci+1 , i = 0 to n − 2. (5.10)
From (5.4) and (5.7) we obtain the n − 1 equations
3ai hi + bi = bi+1 i = 0 to n − 2. (5.11)
It is will be useful to include a definition of the coefficient bn , which is now missing.

(The index of the cubic polynomial coefficients only go up to n − 1.) We simply
extend (5.11) up to i = n − 1 and so write
3an−1 hn−1 + bn−1 = bn , (5.12)
which can be viewed as a definition of bn .

We now proceed to eliminate the sets of a- and c-coefficients (with the d-coefficients
already eliminated in (5.8)) to find a system of linear equations for the b-coefficients.
From (5.11) and (5.12), we can solve for the n a-coefficients to find
1
ai = ( b − bi ) , i = 0 to n − 1. (5.13)
3hi i+1
From (5.9), we can solve for the n c-coefficients as follows:
1
ci = f i − ai h3i − bi h2i
hi

1 1
= fi − (bi+1 − bi ) h3i − bi h2i
hi 3hi
f 1
= i − hi (bi+1 + 2bi ) , i = 0 to n − 1. (5.14)
hi 3
We can now find an equation for the b-coefficients by substituting (5.8), (5.13)
and (5.14) into (5.10):

1 fi 1
3 (bi+1 − bi ) h2i + 2bi hi + − hi (bi+1 + 2bi )
3hi hi 3

f i +1 1
= − hi+1 (bi+2 + 2bi+1 ) ,
h i +1 3
which simplifies to
1 2 1 f f
h i bi + ( h i + h i + 1 ) bi + 1 + h i + 1 bi + 2 = i + 1 − i , (5.15)
3 3 3 h i +1 hi
an equation that is valid for i = 0 to n − 2. Therefore, (5.15) represent n − 1 equa-

tions for the n + 1 unknown b-coefficients. Accordingly, we write the matrix equa-
tion for the b-coefficients, leaving the first and last row absent, as
  
... ... ... ... missing ... ... b0
 1 h0 2 1
3 ( h0 + h1 ) 3 h1 ... 0 0 0 b1
 
3  
 .. .. .. .. .. .. ..  .. 
 .
 . . . . . . 
 . 

1 2 1   bn − 1 
3 h n −2 3 ( h n −2
+ h n −1 ) h
 0 0 0 ... 3 n −1
... ... ... ... missing ... ... bn
missing
 
 f1 − f0 
 h1 h0 
=
 .. 
.
 . 
 f n −1 f n −2 

h n −1 − h n −2

missing
Once the missing first and last equations are specified, the matrix equation for the
b-coefficients can be solved by Gaussian elimination. And once the b-coefficients are
determined, the a- and c-coefficients can also be determined from (5.13) and (5.14),
with the d-coefficients already known from (5.8). The piecewise cubic polynomials,
then, are known and g( x ) can be used for interpolation to any value x satisfying
x0 ≤ x ≤ x n .
The missing first and last equations can be specified in several ways, and here
we show the two ways that are allowed by the MATLAB function spline.m. The
first way should be used when the derivative g0 ( x ) is known at the endpoints x0
and xn ; that is, suppose we know the values of α and β such that
g00 ( x0 ) = α, gn0 −1 ( xn ) = β.
From the known value of α, and using (5.6) and (5.14), we have
α = c0
f 1
= 0 − h0 (b1 + 2b0 ).
h0 3
Therefore, the missing first equation is determined to be
2 1 f0
h0 b0 + h0 b1 = − α. (5.16)
3 3 h0
From the known value of β, and using (5.6), (5.13), and (5.14), we have
β = 3an−1 h2n−1 + 2bn−1 hn−1 + cn−1

1 2 f n −1 1
=3 (bn − bn−1 ) hn−1 + 2bn−1 hn−1 + − hn−1 (bn + 2bn−1 ) ,
3hn−1 h n −1 3
which simplifies to
1 2 f
h b + h bn = β − n − 1 , (5.17)
3 n −1 n −1 3 n −1 h n −1
to be used as the missing last equation.
The second way of specifying the missing first and last equations is called the
not-a-knot condition, which assumes that
g0 ( x ) = g1 ( x ) , g n −2 ( x ) = g n −1 ( x ).
5.4. MULTIDIMENSIONAL INTERPOLATION
Considering the first of these equations, from (5.5) we have
a0 ( x − x0 )3 + b0 ( x − x0 )2 + c0 ( x − x0 ) + d0
= a1 ( x − x1 )3 + b1 ( x − x1 )2 + c1 ( x − x1 ) + d1 .
Now two cubic polynomials can be proven to be identical if at some value of x,

the polynomials and their first three derivatives are identical. Our conditions of
continuity at x = x1 already require that at this value of x these two polynomials
and their first two derivatives are identical. The polynomials themselves will be
identical, then, if their third derivatives are also identical at x = x1 , or if
a0 = a1 .
From (5.13), we have

1 1
(b − b0 ) = (b2 − b1 ),
3h0 1 3h1
or after simplification
h1 b0 − (h0 + h1 )b1 + h0 b2 = 0, (5.18)
which provides us our missing first equation. A similar argument at x = xn − 1
also provides us with our last equation,
hn−1 bn−2 − (hn−2 + hn−1 )bn−1 + hn−2 bn = 0. (5.19)
The MATLAB subroutines spline.m and ppval.m can be used for cubic spline
interpolation (see also interp1.m). I will illustrate these routines in class and post
sample code on the course web site.
5.4 Multidimensional interpolation

Suppose we are interpolating the value of a function of two variables,
z = f ( x, y).
The known values are given by
zij = f ( xi , y j ),
with i = 0, 1, . . . , n and j = 0, 1, . . . , n. Note that the ( x, y) points at which f ( x, y)

are known lie on a grid in the x − y plane.
Let z = g( x, y) be the interpolating function, satisfying zij = g( xi , y j ). A two-
dimensional interpolation to find the value of g at the point ( x, y) may be done by
first performing n + 1 one-dimensional interpolations in y to find the value of g at
the n + 1 points ( x0 , y), ( x1 , y), . . . , ( xn , y), followed by a single one-dimensional
interpolation in x to find the value of g at ( x, y).
In other words, two-dimensional interpolation on a grid of dimension (n + 1) ×
(n + 1) is done by first performing n + 1 one-dimensional interpolations to the
value y followed by a single one-dimensional interpolation to the value x. Two-
dimensional interpolation can be generalized to higher dimensions. The MATLAB
functions to perform two- and three-dimensional interpolation are interp2.m and
interp3.m.
5.4. MULTIDIMENSIONAL INTERPOLATION
Chapter 6
Integration
We want to construct numerical algorithms that can perform definite integrals
of the form Z b
I= f ( x )dx. (6.1)
a
Calculating these definite integrals numerically is called numerical integration, nu-
merical quadrature, or more simply quadrature.
6.1 Elementary formulas

We first consider integration from 0 to h, with h small, to serve as the building blocks
for integration over larger domains. We here define Ih as the following integral:
Z h
Ih = f ( x )dx. (6.2)
0
To perform this integral, we consider a Taylor series expansion of f ( x ) about the

value x = h/2:
( x − h/2)2 00
f ( x ) = f (h/2) + ( x − h/2) f 0 (h/2) + f (h/2)
2
( x − h/2)3 000 ( x − h/2)4 0000
+ f (h/2) + f (h/2) + . . .
6 24
6.1.1 Midpoint rule

The midpoint rule makes use of only the first term in the Taylor series expansion.
Here, we will determine the error in this approximation. Integrating,
( x − h/2)2 00
Z h
Ih = h f (h/2) + ( x − h/2) f 0 (h/2) + f (h/2)
0 2
( x − h/2)3 000 ( x − h/2)4 0000

+ f (h/2) + f (h/2) + . . . dx.
6 24
Changing variables by letting y = x − h/2 and dy = dx, and simplifying the integral
depending on whether the integrand is even or odd, we have
Ih = h f (h/2)
y2 00 y3 000 y4 0000
Z h/2
+ y f 0 (h/2) + f (h/2) + f (h/2) + f (h/2) + . . . dy
−h/2 2 6 24
Z h/2 4
y 0000

= h f (h/2) + y2 f 00 (h/2) + f (h/2) + . . . dy.
0 12
The integrals that we need here are
h h
h3 h5
Z Z
2 2
y2 dy = , y4 dy = .
0 24 0 160
35
6.2. COMPOSITE RULES
Therefore,
h3 00 h5 0000
Ih = h f (h/2) + f (h/2) + f (h/2) + . . . . (6.3)
24 1920
6.1.2 Trapezoidal rule

From the Taylor series expansion of f ( x ) about x = h/2, we have
h 0 h2 00 h3 000 h4 0000
f (0) = f (h/2) − f (h/2) + f (h/2) − f (h/2) + f (h/2) + . . . ,
2 8 48 384
and
h 0 h2 00 h3 000 h4 0000
f (h) = f (h/2) + f (h/2) + f (h/2) + f (h/2) + f (h/2) + . . . .
2 8 48 384
Adding and multiplying by h/2 we obtain
h h3 00 h5 0000
f (0) + f (h) = h f (h/2) + f (h/2) + f (h/2) + . . . .
2 8 384
We now substitute for the first term on the right-hand-side using the midpoint rule
formula:
h3 00 h5 0000

h
f ( 0 ) + f ( h ) = Ih − f (h/2) − f (h/2)
2 24 1920
h3 00 h5 0000
+ f (h/2) + f (h/2) + . . . ,
8 384
and solving for Ih , we find
h h3 00 h5 0000
Ih = f (0) + f ( h ) − f (h/2) − f (h/2) + . . . . (6.4)
2 12 480
6.1.3 Simpson’s rule

To obtain Simpson’s rule, we combine the midpoint and trapezoidal rule to elimi-
nate the error term proportional to h3 . Multiplying (6.3) by two and adding to (6.4),
we obtain

1 2 1
3Ih = h 2 f (h/2) + ( f (0) + f (h)) + h5 − f 0000 (h/2) + . . . ,
2 1920 480
or
h h5 0000
Ih = f (0) + 4 f (h/2) + f (h) − f (h/2) + . . . .
6 2880
Usually, Simpson’s rule is written by considering the three consecutive points 0, h
and 2h. Substituting h → 2h, we obtain the standard result
h h5 0000
I2h = f (0) + 4 f (h) + f (2h) − f (h) + . . . . (6.5)
3 90
6.2 Composite rules

We now use our elementary formulas obtained for (6.2) to perform the integral
given by (6.1).
36 CHAPTER 6. INTEGRATION
6.2. COMPOSITE RULES
6.2.1 Trapezoidal rule

We suppose that the function f ( x ) is known at the n + 1 points labeled as x0 , x1 , . . . , xn ,
with the endpoints given by x0 = a and xn = b. Define
f i = f ( x i ), h i = x i +1 − x i .
Then the integral of (6.1) may be decomposed as

Z b n −1 Z x i +1
a
f ( x )dx = ∑ xi
f ( x )dx
i =0
n −1 Z h i
= ∑ 0
f ( xi + s)ds,
i =0
where the last equality arises from the change-of-variables s = x − xi . Applying the
trapezoidal rule to the integral, we have
Z b n −1
hi
a
f ( x )dx = ∑ 2 i
( f + f i +1 ) . (6.6)
i =0
If the points are not evenly spaced, say because the data are experimental values,
then the hi may differ for each value of i and (6.6) is to be used directly.
However, if the points are evenly spaced, say because f ( x ) can be computed, we
have hi = h, independent of i. We can then define
xi = a + ih, i = 0, 1, . . . , n;
and since the end point b satisfies b = a + nh, we have
b−a
h= .
n
The composite trapezoidal rule for evenly space points then becomes
Z b n −1
h
a
f ( x )dx =
2 ∑ ( f i + f i +1 )
i =0
h
= ( f 0 + 2 f 1 + · · · + 2 f n −1 + f n ) . (6.7)
2
The first and last terms have a multiple of one; all other terms have a multiple of
two; and the entire sum is multiplied by h/2.
6.2.2 Simpson’s rule

We here consider the composite Simpson’s rule for evenly space points. We apply
Simpson’s rule over intervals of 2h, starting from a and ending at b:
Z b
h h
f ( x )dx = ( f0 + 4 f1 + f2 ) + ( f2 + 4 f3 + f4 ) + . . .
a 3 3
h
+ ( f n −2 + 4 f n −1 + f n ) .
3
CHAPTER 6. INTEGRATION 37
6.3. LOCAL VERSUS GLOBAL ERROR
Note that n must be even for this scheme to work. Combining terms, we have
Z b
h
f ( x )dx = ( f 0 + 4 f 1 + 2 f 2 + 4 f 3 + 2 f 4 + · · · + 4 f n −1 + f n ) .
a 3
The first and last terms have a multiple of one; the even indexed terms have a
multiple of 2; the odd indexed terms have a multiple of 4; and the entire sum is
multiplied by h/3.
6.3 Local versus global error

Consider the elementary formula (6.4) for the trapezoidal rule, written in the form
Z h
h h3 00
f ( x )dx = f (0) + f ( h ) − f ( ξ ),
0 2 12
where ξ is some value satisfying 0 ≤ ξ ≤ h, and we have used Taylor’s theorem

with the mean-value form of the remainder. We can also represent the remainder
as
h3
− f 00 (ξ ) = O(h3 ),
12
where O(h3 ) signifies that when h is small, halving of the grid spacing h decreases
the error in the elementary trapezoidal rule by a factor of eight. We call the terms
represented by O(h3 ) the Local Error.
More important is the Global Error which is obtained from the composite formula
(6.7) for the trapezoidal rule. Putting in the remainder terms, we have
h3 n−1 00
Z b
h
12 i∑
f ( x )dx = ( f 0 + 2 f 1 + · · · + 2 f n −1 + f n ) − f ( ξ i ),
a 2 =0
where ξ i are values satisfying xi ≤ ξ i ≤ xi+1 . The remainder can be rewritten as
h3 n−1 00 nh3
00
∑

− f (ξ i ) = − f (ξ i ) ,
12 i=0 12
where f 00 (ξ i ) is the average value of all the f 00 (ξ i )’s. Now,

b−a
n= ,
h
so that the error term becomes
nh3
00 (b − a)h2
00
− f (ξ i ) = − f (ξ i )
12 12
= O( h2 ).
Therefore, the global error is O(h2 ). That is, a halving of the grid spacing only
decreases the global error by a factor of four.
Similarly, Simpson’s rule has a local error of O(h5 ) and a global error of O(h4 ).
6.4. ADAPTIVE INTEGRATION
a d c e b
Figure 6.1: Adaptive Simpson quadrature: Level 1.
6.4 Adaptive integration

The useful MATLAB function quad.m performs numerical integration using adap-
tive Simpson quadrature. The idea is to let the computation itself decide on the grid
size required to achieve a certain level of accuracy. Moreover, the grid size need not
be the same over the entire region of integration.
We begin the adaptive integration at what is called Level 1. The uniformly
spaced points at which the function f ( x ) is to be evaluated are shown in Fig. 6.1.
The distance between the points a and b is taken to be 2h, so that
b−a
h= .
2
Integration using Simpson’s rule (6.5) with grid size h yields
h h5 0000
I= f ( a) + 4 f (c) + f (b) − f ( ξ ),
3 90
where ξ is some value satisfying a ≤ ξ ≤ b.

Integration using Simpson’s rule twice with grid size h/2 yields
h (h/2)5 0000 (h/2)5 0000

I= f ( a) + 4 f (d) + 2 f (c) + 4 f (e) + f (b) − f (ξ l ) − f ( ξ r ),
6 90 90
with ξ l and ξ r some values satisfying a ≤ ξ l ≤ c and c ≤ ξ r ≤ b.

We now define
h
S1 = f ( a) + 4 f (c) + f (b) ,
3
h
S2 = f ( a) + 4 f (d) + 2 f (c) + 4 f (e) + f (b) ,
6
h5
E1 = − f 0000 (ξ ),
90
h5
f 0000 (ξ l ) + f 0000 (ξ r ) .

E2 = − 5
2 · 90
Now we ask whether S2 is accurate enough, or must we further refine the calcula-
tion and go to Level 2? To answer this question, we make the simplifying approxi-
mation that all of the fourth-order derivatives of f ( x ) in the error terms are equal;
that is,
f 0000 (ξ ) = f 0000 (ξ l ) = f 0000 (ξ r ) = C.
CHAPTER 6. INTEGRATION 39
6.4. ADAPTIVE INTEGRATION
Then
h5
E1 = − C,
90
h5 1
E2 = − 4 C= E .
2 · 90 16 1
Then since
S1 + E1 = S2 + E2 ,
and
E1 = 16E2 ,
we have for our estimate for the error term E2 ,
1
E2 = ( S2 − S1 ) .
15
Therefore, given some specific value of the tolerance tol, if

1
(S2 − S1 ) < tol,
15
then we can accept S2 as I. If the tolerance is not achieved for I, then we proceed to
Level 2.
The computation at Level 2 further divides the integration interval from a to b
into the two integration intervals a to c and c to b, and proceeds with the above
procedure independently on both halves. Integration can be stopped on either half
provided the tolerance is less than tol/2 (since the sum of both errors must be less
than tol). Otherwise, either half can proceed to Level 3, and so on.
As a side note, the two values of I given above (for integration with step size h
and h/2) can be combined to give a more accurate value for I given by
16S2 − S1
I= + O( h7 ),
15
where the error terms of O(h5 ) approximately cancel. This free lunch, so to speak,
is called Richardson’s extrapolation.
Chapter 7
Ordinary differential equations

We now discuss the numerical solution of ordinary differential equations. These
include the initial value problem, the boundary value problem, and the eigenvalue
problem. Before proceeding to the development of numerical methods, we review
the analytical solution of some classic equations.
7.1 Examples of analytical solutions

7.1.1 Initial value problem
One classic initial value problem is the RC circuit. With R the resistor and C the
capacitor, the differential equation for the charge q on the capacitor is given by
dq q
R + = 0. (7.1)
dt C
If we consider the physical problem of a charged capacitor connected in a closed
circuit to a resistor, then the initial condition is q(0) = q0 , where q0 is the initial
charge on the capacitor.
The differential equation (7.1) is separable, and separating and integrating from
time t = 0 to t yields
Z q Z t
dq 1
=− dt,
q0 q RC 0
which can be integrated and solved for q = q(t):
q(t) = q0 e−t/RC .
The classic second-order initial value problem is the RLC circuit, with differen-
tial equation
d2 q dq q
L 2 + R + = 0.
dt dt C
Here, a charged capacitor is connected to a closed circuit, and the initial conditions
satisfy
dq
q (0) = q0 , (0) = 0.
dt
The solution is obtained for the second-order equation by the ansatz
q(t) = ert ,
which results in the following so-called characteristic equation for r:
1
Lr2 + Rr + = 0.
C
If the two solutions for r are distinct and real, then the two found exponential
solutions can be multiplied by constants and added to form a general solution. The
constants can then be determined by requiring the general solution to satisfy the
two initial conditions. If the roots of the characteristic equation are complex or
degenerate, a general solution to the differential equation can also be found.
41
7.1. EXAMPLES OF ANALYTICAL SOLUTIONS
7.1.2 Boundary value problems

The dimensionless equation for the temperature y = y( x ) along a linear heat-
conducting rod of length unity, and with an applied external heat source f ( x ),
is given by the differential equation
d2 y
−
= f ( x ), (7.2)
dx2
with 0 ≤ x ≤ 1. Boundary conditions are usually prescribed at the end points of
the rod, and here we assume that the temperature at both ends are maintained at
zero so that
y(0) = 0, y(1) = 0.
The assignment of boundary conditions at two separate points is called a two-
point boundary value problem, in contrast to the initial value problem where the
boundary conditions are prescribed at only a single point. Two-point boundary
value problems typically require a more sophisticated algorithm for a numerical
solution than initial value problems.
Here, the solution of (7.2) can proceed by integration once f ( x ) is specified. We
assume that
f ( x ) = x (1 − x ),
so that the maximum of the heat source occurs in the center of the rod, and goes to
zero at the ends.
The differential equation can then be written as
d2 y
= − x (1 − x ).
dx2
The first integration results in
dy
Z
= ( x2 − x )dx
dx
x3 x2
= − + c1 ,
3 2
where c1 is the first integration constant. Integrating again,
Z 3
x2

x
y( x ) = − + c1 dx
3 2
x4 x3
=− + c1 x + c2 ,
12 6
where c2 is the second integration constant. The two integration constants are de-
termined by the boundary conditions. At x = 0, we have
0 = c2 ,
and at x = 1, we have
1 1
0= − + c1 ,
12 6
so that c1 = 1/12. Our solution is therefore
x4 x3 x
y( x ) = − +
12 6 12
1
= x (1 − x )(1 + x − x2 ).
12
42 CHAPTER 7. ORDINARY DIFFERENTIAL EQUATIONS

7.2. NUMERICAL METHODS: INITIAL VALUE PROBLEM
The temperature of the rod is maximum at x = 1/2 and goes smoothly to zero at
the ends.
7.1.3 Eigenvalue problem

The classic eigenvalue problem obtained by solving the wave equation by separation
of variables is given by
d2 y
+ λ2 y = 0,
dx2
with the two-point boundary conditions y(0) = 0 and y(1) = 0. Notice that
y( x ) = 0 satisfies both the differential equation and the boundary conditions. Other
nonzero solutions for y = y( x ) are possible only for certain discrete values of λ.
These values of λ are called the eigenvalues of the differential equation.
We proceed by first finding the general solution to the differential equation. It
is easy to see that this solution is
y( x ) = A cos λx + B sin λx.
Imposing the first boundary condition at x = 0, we obtain
A = 0.
The second boundary condition at x = 1 results in
B sin λ = 0.
Since we are searching for a solution where y = y( x ) is not identically zero, we

must have
λ = π, 2π, 3π, . . . .
The corresponding negative values of λ are also solutions, but their inclusion only
changes the corresponding values of the unknown B constant. A linear superposi-
tion of all the solutions results in the general solution
∞
y( x ) = ∑ Bn sin nπx.
n =1
For each eigenvalue nπ, we say there is a corresponding eigenfunction sin nπx.
When the differential equation can not be solved analytically, a numerical method
should be able to solve for both the eigenvalues and eigenfunctions.
7.2 Numerical methods: initial value problem

We begin with the simple Euler method, then discuss the more sophisticated Runge-
Kutta methods, and conclude with the Runge-Kutta-Fehlberg method, as imple-
mented in the MATLAB function ode45.m. Our differential equations are for x =
x (t), where the time t is the independent variable, and we will make use of the
notation ẋ = dx/dt. This notation is still widely used by physicists and descends
directly from the notation originally used by Newton.
CHAPTER 7. ORDINARY DIFFERENTIAL EQUATIONS 43

7.2.1 Euler method

The Euler method is the most straightforward method to integrate a differential
equation. Consider the first-order differential equation
ẋ = f (t, x ), (7.3)
with the initial condition x (0) = x0 . Define tn = n∆t and xn = x (tn ). A Taylor
series expansion of xn+1 results in
xn+1 = x (tn + ∆t)

= x (tn ) + ∆t ẋ (tn ) + O(∆t2 )
= x (tn ) + ∆t f (tn , xn ) + O(∆t2 ).
The Euler Method is therefore written as
xn+1 = x (tn ) + ∆t f (tn , xn ).
We say that the Euler method steps forward in time using a time-step ∆t, starting
from the initial value x0 = x (0). The local error of the Euler Method is O(∆t2 ).
The global error, however, incurred when integrating to a time T, is a factor of 1/∆t
larger and is given by O(∆t). It is therefore customary to call the Euler Method a
first-order method.
7.2.2 Modified Euler method

This method is of a type that is called a predictor-corrector method. It is also the
first of what are Runge-Kutta methods. As before, we want to solve (7.3). The idea
is to average the value of ẋ at the beginning and end of the time step. That is, we
would like to modify the Euler method and write
1
xn+1 = xn + ∆t f (tn , xn ) + f (tn + ∆t, xn+1 ) .

2
The obvious problem with this formula is that the unknown value xn+1 appears
on the right-hand-side. We can, however, estimate this value, in what is called the
predictor step. For the predictor step, we use the Euler method to find
p
xn+1 = xn + ∆t f (tn , xn ).
The corrector step then becomes
1 p
xn+1 = xn + ∆t f (tn , xn ) + f (tn + ∆t, xn+1 ) .

2
The Modified Euler Method can be rewritten in the following form that we will
later identify as a Runge-Kutta method:
k1 = ∆t f (tn , xn ),
k2 = ∆t f (tn + ∆t, xn + k1 ), (7.4)
1
x n +1 = x n + ( k 1 + k 2 ).
2

7.2.3 Second-order Runge-Kutta methods

We now derive all second-order Runge-Kutta methods. Higher-order methods can
be similarly derived, but require substantially more algebra.
We consider the differential equation given by (7.3). A general second-order
Runge-Kutta method may be written in the form
k1 = ∆t f (tn , xn ),
k2 = ∆t f (tn + α∆t, xn + βk1 ), (7.5)
xn+1 = xn + ak1 + bk2 ,
with α, β, a and b constants that define the particular second-order Runge-Kutta

method. These constants are to be constrained by setting the local error of the
second-order Runge-Kutta method to be O(∆t3 ). Intuitively, we might guess that
two of the constraints will be a + b = 1 and α = β.
We compute the Taylor series of xn+1 directly, and from the Runge-Kutta method,
and require them to be the same to order ∆t2 . First, we compute the Taylor series
of xn+1 . We have
xn+1 = x (tn + ∆t)

1
= x (tn ) + ∆t ẋ (tn ) + (∆t)2 ẍ (tn ) + O(∆t3 ).
2
Now,
ẋ (tn ) = f (tn , xn ).
The second derivative is more complicated and requires partial derivatives. We
have

d
ẍ (tn ) = f (t, x (t))
dt t=tn
= f t (tn , xn ) + ẋ (tn ) f x (tn , xn )
= f t ( t n , x n ) + f ( t n , x n ) f x ( t n , x n ).
Therefore,
1
xn+1 = xn + ∆t f (tn , xn ) + (∆t)2 f t (tn , xn ) + f (tn , xn ) f x (tn , xn ) .

(7.6)
2
Second, we compute xn+1 from the Runge-Kutta method given by (7.5). Substi-
tuting in k1 and k2 , we have

xn+1 = xn + a∆t f (tn , xn ) + b∆t f tn + α∆t, xn + β∆t f (tn , xn ) .
We Taylor series expand using

f tn + α∆t, xn + β∆t f (tn , xn )
= f (tn , xn ) + α∆t f t (tn , xn ) + β∆t f (tn , xn ) f x (tn , xn ) + O(∆t2 ).
The Runge-Kutta formula is therefore
xn+1 = xn + ( a + b)∆t f (tn , xn )

+ (∆t)2 αb f t (tn , xn ) + βb f (tn , xn ) f x (tn , xn ) + O(∆t3 ). (7.7)


Comparing (7.6) and (7.7), we find
a + b = 1,
αb = 1/2,
βb = 1/2.
There are three equations for four parameters, and there exists a family of second-
order Runge-Kutta methods.
The Modified Euler Method given by (7.4) corresponds to α = β = 1 and a =
b = 1/2. Another second-order Runge-Kutta method, called the Midpoint Method,
corresponds to α = β = 1/2, a = 0 and b = 1. This method is written as
k1 = ∆t f (tn , xn ),

1 1
k2 = ∆t f tn + ∆t, xn + k1 ,
2 2
x n +1 = x n + k 2 .
7.2.4 Higher-order Runge-Kutta methods

The general second-order Runge-Kutta method was given by (7.5). The general
form of the third-order method is given by
k1 = ∆t f (tn , xn ),
k2 = ∆t f (tn + α∆t, xn + βk1 ),
k3 = ∆t f (tn + γ∆t, xn + δk1 + ek2 ),
xn+1 = xn + ak1 + bk2 + ck3 .
The following constraints on the constants can be guessed: α = β, γ = δ + e, and

a + b + c = 1. Remaining constraints need to be derived.
The fourth-order method has a k1 , k2 , k3 and k4 . The fifth-order method requires
up to k6 . The table below gives the order of the method and the number of stages
required.
order 2 3 4 5 6 7 8
minimum # stages 2 3 4 6 7 9 11
Because of the jump in the number of stages required between the fourth-order
and fifth-order method, the fourth-order Runge-Kutta method has some appeal.
The general fourth-order method starts with 13 constants, and one then finds 11
constraints. A particularly simple fourth-order method that has been widely used
is given by
k1 = ∆t f (tn , xn ),

1 1
k2 = ∆t f tn + ∆t, xn + k1 ,
2 2

1 1
k3 = ∆t f tn + ∆t, xn + k2 ,
2 2
k4 = ∆t f (tn + ∆t, xn + k3 ) ,
1
xn+1 = xn + (k1 + 2k2 + 2k3 + k4 ) .
6

7.2.5 Adaptive Runge-Kutta Methods

As in adaptive integration, it is useful to devise an ode integrator that automatically
finds the appropriate ∆t. The Dormand-Prince Method, which is implemented
in MATLAB’s ode45.m, finds the appropriate step size by comparing the results
of a fifth-order and fourth-order method. It requires six function evaluations per
time step, and constructs both a fifth-order and a fourth-order method from these
function evaluations.
Suppose the fifth-order method finds xn+1 with local error O(∆t6 ), and the
fourth-order method finds xn0 +1 with local error O(∆t5 ). Let ε be the desired er-
ror tolerance of the method, and let e be the actual error. We can estimate e from
the difference between the fifth- and fourth-order methods; that is,
e = | xn+1 − xn0 +1 |.
Now e is of O(∆t5 ), where ∆t is the step size taken. Let ∆τ be the estimated step
size required to get the desired error ε. Then we have
e/ε = (∆t)5 /(∆τ )5 ,
or solving for ∆τ,

ε 1/5
∆τ = ∆t .
e
On the one hand, if e < ε, then we accept xn+1 and do the next time step using
the larger value of ∆τ. On the other hand, if e > ε, then we reject the integration
step and redo the time step using the smaller value of ∆τ. In practice, one usually
increases the time step slightly less and decreases the time step slightly more to
prevent the waste of too many failed time steps.
7.2.6 System of differential equations

Our numerical methods can be easily adapted to solve higher-order differential
equations, or equivalently, a system of differential equations. First, we show how
a second-order differential equation can be reduced to two first-order equations.
Consider
ẍ = f (t, x, ẋ ).
This second-order equation can be rewritten as two first-order equations by defining
u = ẋ. We then have the system
ẋ = u,
u̇ = f (t, x, u).
This trick also works for higher-order equation. For another example, the third-
order equation
...
x = f (t, x, ẋ, ẍ ),
can be written as
ẋ = u,
u̇ = v,
v̇ = f (t, x, u, v).

7.3. NUMERICAL METHODS: BOUNDARY VALUE PROBLEM
Now, we show how to generalize Runge-Kutta methods to a system of differen-

tial equations. As an example, consider the following system of two odes,
ẋ = f (t, x, y),
ẏ = g(t, x, y),
with the initial conditions x (0) = x0 and y(0) = y0 . The generalization of the
commonly used fourth-order Runge-Kutta method would be
k1 = ∆t f (tn , xn , yn ),
l1 = ∆tg(tn , xn , yn ),

1 1 1
k2 = ∆t f tn + ∆t, xn + k , yn + l1 ,
2 2 1 2

1 1 1
l2 = ∆tg tn + ∆t, xn + k , yn + l1 ,
2 2 1 2

1 1 1
k3 = ∆t f tn + ∆t, xn + k2 , yn + l2 ,
2 2 2

1 1 1
l3 = ∆tg tn + ∆t, xn + k2 , yn + l2 ,
2 2 2
k4 = ∆t f (tn + ∆t, xn + k3 , yn + l3 ) ,
l4 = ∆tg (tn + ∆t, xn + k3 , yn + l3 ) ,
1
x n +1 = x n + (k + 2k2 + 2k3 + k4 ) ,
6 1
1
y n +1 = yn + (l1 + 2l2 + 2l3 + l4 ) .
6
7.3 Numerical methods: boundary value problem

7.3.1 Finite difference method
We consider first the differential equation
d2 y
−
= f ( x ), 0 ≤ x ≤ 1, (7.8)
dx2
with two-point boundary conditions
y(0) = A, y(1) = B.
Equation (7.8) can be solved by quadrature, but here we will demonstrate a numer-
ical solution using a finite difference method.
We begin by discussing how to numerically approximate derivatives. Consider
the Taylor series approximation for y( x + h) and y( x − h), given by
1 1 3 000 1 4 0000
y( x + h) = y( x ) + hy0 ( x ) + h2 y00 ( x ) + h y (x) + h y (x) + . . . ,
2 6 24
1 1 3 000 1 4 0000
y( x − h) = y( x ) − hy0 ( x ) + h2 y00 ( x ) − h y (x) + h y (x) + . . . .
2 6 24

The standard definitions of the derivatives give the first-order approximations
y( x + h) − y( x )
y0 ( x ) = + O( h ),
h
y( x ) − y( x − h)
y0 ( x ) = + O( h ).
h
The more widely-used second-order approximation is called the central difference
approximation and is given by
y( x + h) − y( x − h)
y0 ( x ) = + O( h2 ).
2h
The finite difference approximation to the second derivative can be found from
considering
1 4 0000
y( x + h) + y( x − h) = 2y( x ) + h2 y00 ( x ) + h y (x) + . . . ,
12
from which we find
y( x + h) − 2y( x ) + y( x − h)
y00 ( x ) = + O( h2 ).
h2
Sometimes a second-order method is required for x on the boundaries of the do-
main. For a boundary point on the left, a second-order forward difference method
requires the additional Taylor series
4
y( x + 2h) = y( x ) + 2hy0 ( x ) + 2h2 y00 ( x ) + h3 y000 ( x ) + . . . .
3
We combine the Taylor series for y( x + h) and y( x + 2h) to eliminate the term pro-
portional to h2 :
y( x + 2h) − 4y( x + h) = −3y( x ) − 2hy0 ( x ) + O(h3 ).
Therefore,
−3y( x ) + 4y( x + h) − y( x + 2h)
y0 ( x ) = + O( h2 ).
2h
For a boundary point on the right, we send h → −h to find
3y( x ) − 4y( x − h) + y( x − 2h)

y0 ( x ) = + O( h2 ).
2h
We now write a finite difference scheme to solve (7.8). We discretize x by defin-
ing xi = ih, i = 0, 1, . . . , n + 1. Since xn+1 = 1, we have h = 1/(n + 1). The functions
y( x ) and f ( x ) are discretized as yi = y( xi ) and f i = f ( xi ). The second-order differ-
ential equation (7.8) then becomes for the interior points of the domain
−yi−1 + 2yi − yi+1 = h2 f i , i = 1, 2, . . . n,
with the boundary conditions y0 = A and yn+1 = B. We therefore have a linear

system of equations to solve. The first and nth equation contain the boundary
conditions and are given by
2y1 − y2 = h2 f 1 + A,
−yn−1 + 2yn = h2 f n + B.

The second and third equations, etc., are
−y1 + 2y2 − y3 = h2 f 2 ,
−y2 + 2y3 − y4 = h2 f 3 ,
...
In matrix form, we have

    2 
2 −1 0 0 ... 0 0 0 y1 h f1 + A
 −1 2
 2 − 1 0 ... 0 0 0   y2   h2 f 2 
   
 0 −1 2 − 1 ... 0 0 0  y3   h f 3 
..   ..  =  .
    
 .. .. .. .. .. .. .. ..
 .
 . . . . . . .  .  
    . 

 0 0 0 0 ... −1 2 −1   y n −1   2
h f n −1 
0 0 0 0 ... 0 −1 2 yn h2 f n + B
The matrix is tridiagonal, and a numerical solution by Guassian elimination can be

done quickly. The matrix itself is easily constructed using the MATLAB function
diag.m and ones.m. As excerpted from the MATLAB help page, the function call
ones(m,n) returns an m-by-n matrix of ones, and the function call diag(v,k), when
v is a vector with n components, is a square matrix of order n+abs(k) with the el-
ements of v on the k-th diagonal: k = 0 is the main diagonal, k > 0 is above the
main diagonal and k < 0 is below the main diagonal. The n × n matrix above can
be constructed by the MATLAB code
M=diag(-ones(n-1,1),-1)+diag(2*ones(n,1),0)+diag(-ones(n-1,1),1); .
The right-hand-side, provided f is a given n-by-1 vector, can be constructed by

the MATLAB code
b=hˆ2*f; b(1)=b(1)+A; b(n)=b(n)+B;
and the solution for u is given by the MATLAB code
y=M\b;
7.3.2 Shooting method

The finite difference method can solve linear odes. For a general ode of the form
d2 y
= f ( x, y, dy/dx ),
dx2
with y(0) = A and y(1) = B, we use a shooting method. First, we formulate the
ode as an initial value problem. We have
dy
= z,
dx
dz
= f ( x, y, z).
dx

7.4. NUMERICAL METHODS: EIGENVALUE PROBLEM
The initial condition y(0) = A is known, but the second initial condition z(0) = b
is unknown. Our goal is to determine b such that y(1) = B.
In fact, this is a root-finding problem for an appropriately defined function. We
define the function F = F (b) such that
F (b) = y(1) − B.
In other words, F (b) is the difference between the value of y(1) obtained from
integrating the differential equations using the initial condition z(0) = b, and B.
Our root-finding routine will want to solve F (b) = 0. (The method is called shooting
because the slope of the solution curve for y = y( x ) at x = 0 is given by b, and the
solution hits the value y(1) at x = 1. This looks like pointing a gun and trying to
shoot the target, which is B.)
To determine the value of b that solves F (b) = 0, we iterate using the Secant
method, given by
bn − bn − 1
bn + 1 = bn − F ( bn ) .
F ( bn ) − F ( bn − 1 )
We need to start with two initial guesses for b, solving the ode for the two
corresponding values of y(1). Then the Secant Method will give us the next value
of b to try, and we iterate until |y(1) − B| < tol, where tol is some specified tolerance
for the error.
7.4 Numerical methods: eigenvalue problem

For illustrative purposes, we develop our numerical methods for what is perhaps
the simplest eigenvalue ode. With y = y( x ) and 0 ≤ x ≤ 1, this simple ode is given
by
y00 + λ2 y = 0. (7.9)
To solve (7.9) numerically, we will develop both a finite difference method and a
shooting method. Furthermore, we will show how to solve (7.9) with homogeneous
boundary conditions on either the function y or its derivative y0 .
7.4.1 Finite difference method

We first consider solving (7.9) with the homogeneous boundary conditions y(0) =
y(1) = 0. In this case, we have already shown that the eigenvalues of (7.9) are given
by λ = π, 2π, 3π, . . . .
With n interior points, we have xi = ih for i = 0, . . . , n + 1, and h = 1/(n + 1).
Using the centered-finite-difference approximation for the second derivative, (7.9)
becomes
yi−1 − 2yi + yi+1 = −h2 λ2 yi . (7.10)
Applying the boundary conditions y0 = yn+1 = 0, the first equation with i = 1, and
the last equation with i = n, are given by
−2y1 + y2 = −h2 λ2 y1 ,
yn−1 − 2yn = −h2 λ2 yn .
The remaining n − 2 equations are given by (7.10) for i = 2, . . . , n − 1.

It is of interest to see how the solution develops with increasing n. The smallest
possible value is n = 1, corresponding to a single interior point, and since h = 1/2
we have
1
−2y1 = − λ2 y1 ,
4
2
√
so that λ = 8, or λ = 2 2 = 2.8284. This is to be compared to the first eigenvalue
which is λ = π. When n = 2, we have h = 1/3, and the resulting two equations
written in matrix form are given by

−2 1 y1 1 y
= − λ2 1 .
1 −2 y2 9 y2
This is a matrix eigenvalue problem with the eigenvalue given by µ = −λ2 /9. The
solution for µ is arrived at by solving

−2 − µ 1
det = 0,
1 −2 − µ
with resulting quadratic equation
(2 + µ)2 − 1 = 0.
√ √
The solutions are µ = −1, −3, and since λ = 3 −µ, we have λ = 3, 3 3 = 5.1962.
These two eigenvalues serve as rough approximations to the first two eigenvalues
π and 2π.
With A an n-by-n matrix, the MATLAB variable mu=eig(A) is a vector containing
the n eigenvalues of the matrix A. The built-in function eig.m can therefore be used
to find the eigenvalues. With n grid points, the smaller eigenvalues will converge
more rapidly than the larger ones.
We can also consider boundary conditions on the derivative, or mixed boundary
conditions. For example, consider the mixed boundary conditions given by y(0) = 0
and y0 (1) = 0. The eigenvalues of (7.9) can then be determined analytically to be
λi = (i − 1/2)π, with i a natural number.
The difficulty we now face is how to implement a boundary condition on the
derivative. Our computation of y00 uses a second-order method, and we would
like the computation of the first derivative to also be second order. The condition
y0 (1) = 0 occurs on the right-most boundary, and we can make use of the second-
order backward-difference approximation to the derivative that we have previously
derived. This finite-difference approximation for y0 (1) can be written as
3yn+1 − 4yn + yn−1
y0n+1 = . (7.11)
2h
Now, the nth finite-difference equation was given by
yn−1 − 2yn + yn+1 = −h2 yn ,
and we now replace the value yn+1 using (7.11); that is,
1
2hy0n+1 + 4yn − yn−1 .

y n +1 =
3
Implementing the boundary condition y0n+1 = 0, we have
4 1
y n +1 = y n − y n −1 .
3 3

Therefore, the nth finite-difference equation becomes
2 2
y − y n = − h2 λ2 y n .
3 n −1 3
For example, when n = 2, the finite difference equations become

−2 1 y1 1 2 y1
2 =− λ .
3 − 23 y2 9 y2
The eigenvalues of the matrix are now the solution of

2 2
( µ + 2) µ + − = 0,
3 3
or
3µ2 + 8µ + 2 = 0.
√
Therefore, µ = (−4 ± 10)/3, and we find λ = 1.5853, 4.6354, which are approxi-
mations to π/2 and 3π/2.
7.4.2 Shooting method

We apply the shooting method to solve (7.9) with boundary conditions y(0) =
y(1) = 0. The initial value problem to solve is
y0 = z,
z0 = −λ2 y,
with known boundary condition y(0) = 0 and an unknown boundary condition

on y0 (0). In fact, any nonzero boundary condition on y0 (0) can be chosen: the
differential equation is linear and the boundary conditions are homogeneous, so
that if y( x ) is an eigenfunction then so is Ay( x ). What we need to find here is the
value of λ such that y(1) = 0. In other words, choosing y0 (0) = 1, we solve
F (λ) = 0, (7.12)
where F (λ) = y(1), obtained by solving the initial value problem. Again, an itera-
tion for the roots of F (λ) can be done using the Secant Method. For the eigenvalue
problem, there are an infinite number of roots, and the choice of the two initial
guesses for λ will then determine to which root the iteration will converge.
For this simple problem, it is possible to write explicitly the equation F (λ) = 0.
The general solution to (7.9) is given by
y( x ) = A cos (λx ) + B sin (λx ).
The initial condition y(0) = 0 yields A = 0. The initial condition y0 (0) = 1 yields
B = 1/λ.
Therefore, the solution to the initial value problem is
sin (λx )
y( x ) = .
λ

The function F (λ) = y(1) is therefore given by
sin λ
F (λ) = ,
λ
and the roots occur when λ = π, 2π, . . . .
If the boundary conditions were y(0) = 0 and y0 (1) = 0, for example, then we
would simply redefine F (λ) = y0 (1). We would then have
cos λ
F (λ) = ,
λ
and the roots occur when λ = π/2, 3π/2, . . . .

Numerical Methods: Jeffrey R. Chasnov

Uploaded by

Numerical Methods: Jeffrey R. Chasnov

Uploaded by

Numerical Methods

Click to view a promotional video

If you want to learn differential equations, have a look at

Differential Equations for Engineers

If your interests are matrices and elementary linear algebra, try

Matrix Algebra for Engineers

Vector Calculus for Engineers

And if your interest is numerical methods, have a go at

Numerical Methods for Engineers

7 Ordinary differential equations 41

1.2 Numbers with a decimal or binary point

1.3 Examples of binary numbers

1.4 Hex numbers

1.5 4-bit unsigned integers as hex numbers

1.6 IEEE single precision format:

1.7 Special numbers

Number Range: e = 1111 1111 = 28 - 1 = 255 reserved

Smallest positive normal number

2 CHAPTER 1. IEEE ARITHMETIC

1.8 Examples of computer numbers

1.0 = (−1)0 × 2(127−127) × 1.0

2.0 = (−1)0 × 2(128−127) × 1.0

1/2 = (−1)0 × 2(126−127) × 1.0

1.9 Inexact numbers

CHAPTER 1. IEEE ARITHMETIC 3

1.10 Machine epsilon

Find the machine number just greater than 5

4 CHAPTER 1. IEEE ARITHMETIC

1.11 IEEE double precision format

1.12 Roundoff error example

where b is a parameter. The quadratic formula yields the two solutions

CHAPTER 1. IEEE ARITHMETIC 5

For a subroutine written to compute the solution of a quadratic for a general

6 CHAPTER 1. IEEE ARITHMETIC

2.1 Bisection Method

2.2 Newton’s Method

2.3 Secant Method

analytical derivative of the function f ( x ). We write in place of f 0 ( xn ),

Starting the Secant Method requires a guess for both x0 and x1 .

2.3.2 Example of fractals using Newton’s Method

These roots are the three cubic roots of unity. With

the three unique cubic roots of unity are given by

8 CHAPTER 2. ROOT FINDING

2.4 Order of convergence

If for large n we have the approximate relationship

with k a positive constant, then we say the root-finding numerical method is of

2.4.1 Newton’s Method

Subtracting both sides from r, we have

CHAPTER 2. ROOT FINDING 9

Therefore, we have shown that

2.4.2 Secant Method

and the Taylor series

10 CHAPTER 2. ROOT FINDING

From this ansatz, we also have

Equating the coefficient and the power of en−1 results in

CHAPTER 2. ROOT FINDING 11

12 CHAPTER 2. ROOT FINDING

a11 x1 + a12 x2 + · · · + a1n xn = b1 ,

We can write this system as the matrix equation

3.1 Gaussian Elimination

−3x1 + 2x2 − x3 = −1,

To perform Gaussian elimination, we form an Augmented Matrix by combining the

14 CHAPTER 3. SYSTEMS OF EQUATIONS

Another nice feature of the LU decomposition is that it can be done by overwriting

CHAPTER 3. SYSTEMS OF EQUATIONS 15

With y = Ux, we first solve Ly = b, that is

Using forward substitution

We now solve Ux = y, that is

Using backward substitution,

and we have once again determined

3.3 Partial pivoting

16 CHAPTER 3. SYSTEMS OF EQUATIONS

P23 P12 A = (P23 M1 P23 )−1 U,

CHAPTER 3. SYSTEMS OF EQUATIONS 17

where the second and third lines can be shortened to

3.4 Operation counts