Chapter 6

Linear Transformations
6.1 Introduction to Linear Transformations
6.2 The Kernel and Range of a Linear Transformation
6.3 Matrices for Linear Transformations
6.4 Transition Matrices and Similarity

MAT1041 - Chapter 6 1
 6.1 Introduction to Linear Transformation
 Learn about functions that map a vector space V into a
vector space W --- T: V  W
V: domain of T
range
v
w image of v
T: V  W
W: codomain of T

MAT1041 - Chapter 6 6-2

Section 6-1

Map
 If v is in V and w is in W s.t. T(v) = w, then w is called the
image of v under T.

 The set of all images of vectors in V is called the range of T.

 The set of all v in V s.t. T(v) = w is called the preimage of w.

MAT1041 - Chapter 6 6-3

Section 6-1

Ex. 1: A function from R2 into R 2

 For any vector v = (v1, v2) in R2, and let T: R2 R2 be defined
by T(v1, v2) = (v1  v2, v1 + 2v2).

Find the image of v = (1, 2)

T(1, 2) =(1 2, 1+2·2) = (3, 3)

Find the preimage of w = (1, 11)

T(v1, v2) = (v1  v2, v1 + 2v2) = (1, 11)
 v1  v2 = 1 and v1 + 2v2 = 11
 v1 = 3; v2 = 4
MAT1041 - Chapter 6 6-4
Section 6-1

Linear Transformation
 Let V and W be vector spaces. The function T: V  W is
called a linear transformation of V into W if the
following two properties are true for all u and v in V and
for any scalar c.
1. T(u + v) = T(u) + T(v)
2. T(cu) = cT(u)

 A linear transformation is said to be operation reserving

(the operations of addition and scalar multiplication).
MAT1041 - Chapter 6 6-5
Section 6-1

Ex.2: Verifying a linear transformation from R 2 into R2

Show that the function T(v1, v2) = (v1  v2, v1 + 2v2) is a linear
transformation from R2 into R2.
Let v = (v1, v2) and u = (u1, u2)
 Vector addition: v + u = (v1 + u1, v2 + u2)

T(v + u) = T(v1 + u1, v2 + u2)

= ( (v1 + u1)  (v2 + u2), (v1 + u1) + 2(v2 + u2) )
= (v1  v2 , v1 + 2v2) + (u1  u2, u1 +2u2)
= T(v) + T(u)
 Scalar multiplication: cv = c(v1, v2) = (cv1, cv2)

T(cv) = (cv1  cv2, cv1+ 2cv2) = c(v1  v2, v1+ 2v2) = cT(v)
Therefore T is a linear transformation.
MAT1041 - Chapter 6 6-6
Section 6-1

Ex. 3: Not linear transformation

 f(x) = sin(x)
In general, sin(x1 + x2)  sin(x1) + sin(x2)
 f(x) = x2
In general, ( x1  x2 )2  x12  x22
 f(x) = x + 1
f(x1 + x2) = x1 + x2 + 1
f(x1) + f(x2) = (x1 + 1) + (x2 + 1) = x1 + x2 + 2
Thus, f(x1 + x2)  f(x1) + f(x2)
MAT1041 - Chapter 6 6-7
Section 6-1

Linear Operation & Zero / Identity Transformation

 A linear transformation T: V  V from a vector space

into itself is called a linear operator.

 Zero transformation (T: V  W): T(v) = 0, for all v

 Identity transformation (T: V  V): T(v) = v, for all v

MAT1041 - Chapter 6 6-8

Section 6-1

Thm 6.1: Linear transformations

Let T be a linear transformation from V into W, where u and
v are in V. Then the following properties are true.
1. T(0) = 0
2. T(v) = T(v)
3. T(u  v) = T(u)  T(v)
4. If v = c1v1 + c2v2 + … + cnvn, then
T(v) = c1T(v1) + c2T(v2) + … + cnT(vn)

MAT1041 - Chapter 6 6-9

Section 6-1

Proof of Theorem 6.1

1. Note that 0v = 0. Then it follows that
T(0) = T(0v) = 0T(v) = 0
2. Follow from v = (1)v, which implies that
T(v) = T[(1)v] = (1)T(v) = T(v)
3. Follow from u  v = u + (v), which implies that
T(u  v) = T[u + (1)v] = T(u) + (1)T(v)
= T(u)  T(v)
4. Left to you
MAT1041 - Chapter 6 6-10
Section 6-1

Remark of Theorem 6.1

 A linear transformation T: V  W is determined
completely by its action on a basis of V.

 If {v1, v2, …, vn} is a basis for the vector space V and if

T(v1), T(v2), …, T(vn) are given, then T(v) is determined
for any v in V.

MAT1041 - Chapter 6 6-11

Section 6-1

Ex 4: Linear transformations and bases

Let T: R3  R3 be a linear transformation s.t.
T(1, 0, 0) = (2, 1, 4)
T(0, 1, 0) = (1, 5, 2)
T(0, 0, 1) = (0, 3, 1).
Find T(2, 3, 2).

(2, 3, 2) = 2(1, 0, 0) + 3(0, 1, 0) 2(0, 0, 1)

T(2, 3, 2) = 2T(1, 0, 0) + 3T(0, 1, 0) 2T(0, 0, 1)
= 2(2, 1, 4) + 3(1, 5, 2) 2(0, 3, 1)
= (7, 7, 0)
MAT1041 - Chapter 6 6-12
Section 6-1

Ex 5: Linear transformation defined by a matrix

The function T: R2  R3 is defined as follows
 3 0
   v1 
T ( v)  Av   2 1  
 1  2  v2 
 Find T(v), where v = (2, 1)

 3 0 6 
   2  
T ( v)  Av   2 1    3
  1
 1  2 0
Therefore, T(2, 1) = (6, 3, 0)

MAT1041 - Chapter 6 6-13

Section 6-1

Example 5 (cont.)
 Show that T is a linear transformation from R2 to R3.
1. For any u and v in R2, we have
T(u + v) = A(u + v) = Au +Av = T(u) +T(v)
2. For any u in R2 and any scalar c, we have
T(cu) = A(cu) = c(Au) = cT(u)

Therefore, T is a linear transformation from R2 to R3.

MAT1041 - Chapter 6 6-14

Section 6-1

Thm 6.2: Linear transformation given by a matrix

Let A be an m  n matrix. The function T defined by T(v) = Av is
a linear transformation from Rn into Rm.

Note: In order to conform to matrix multiplication with an m  n

matrix, the vectors in Rn are represented by n  1 matrices and the
vectors in Rm are represented by m  1 matrices.

MAT1041 - Chapter 6 6-15

Section 6-1

Remark of Theorem 6.2

 The m  n matrix zero matrix corresponds to the zero transformation
from Rn into Rm.
 The n  n matrix identity matrix In corresponds to the identity
transformation from Rn into Rn.
 An m  n matrix A defines a linear transformation from Rn into Rm.

 a11 a12  a1n   v1   a11v1  a12v2   a1n vn 

a a22  a2 n  v2   a21v1  a22v2   a2 n vn 
Av   21

         
    
am1 am 2  amn  vn  am1v1  am 2v2   amn vn 
Rn Rm
MAT1041 - Chapter 6 6-16
Section 6-1

Ex 8: A projection in R 3
1 0 0
The linear transformation T: R3  R3 given by A  0 1 0
is called a projection in R3. 0 0 0
If v = (x, y, z) is a vector in R3, then T(v) = (x, y, 0). In other words, T
maps every vector in R3 to its orthogonal projection in the xy - plane.
z
( x, y, z )

y
x T ( x, y , z )
 ( x, y,0)

MAT1041 - Chapter 6 6-17

Section 6-1

Ex 9: Linear transformation from Mm,n to Mn,m

Let T: Mm,n  Mn,m be the function that maps m  n matrix A to its
transpose. That is, T ( A)  AT. Show that T is a linear transformation.

Proof: Let A and B be m  n matrix.

 T ( A  B)  ( A  B)T  AT  BT  T ( A)  T ( B)
and T (cA)  (cA)T  c( AT )  cT ( A)
T is a linear transformation form M m, n into M n, m

MAT1041 - Chapter 6 6-18

 6.2 The Kernel and Range of a Linear Transformation
 Definition of Kernel of a Linear Transformation
Let T:V  W be a linear transformation. Then the set of all
vectors v in V that satisfy T(v) = 0 is called the kernel of T and is
denoted by ker(T).

 The kernel of the zero transformation T: V  W consists of all of

V because T(v) = 0 for every v in V. That is, ker(T) = V.

 The kernel of the identity transformation T: V  V consists of

the single element 0. That is, ker(T) = {0}.

MAT1041 - Chapter 6 6-19

Section 6-2

Ex 3: Finding the kernel

Find the kernel of the projection T: R3  R3 given by
T(x, y, z) = (x, y, 0).
Sol: This linear transformation projects the vector (x, y, z)
in R3 to the vector (x, y, 0) in xy-plane. z
Therefore, ker(T) = { (0, 0, z) : zR} (0,0, z )

(0,0,0)

x y
MAT1041 - Chapter 6 6-20
Section 6-2

Ex 4: Finding the kernel

Find the kernel of T: R2  R3 given by
T(x1, x2) = (x1  2x2, 0, x1).

Sol: The kernel of T is the set of all x = (x1, x2) in R2 s.t.

T(x1, x2) = (x1  2x2, 0, x1) = (0, 0, 0).
Therefore, (x1, x2) = (0, 0).
 ker(T) = { (0, 0) } = { 0 }

MAT1041 - Chapter 6 6-21

Section 6-2

Ex 5: Finding the kernel

 1  1  2
Find the kernel of T: R3  R2 defined by T(x) = Ax, where A   1 2 
 3 

Sol: The kernel of T is the set of all x = (x1, x2, x3) in R3 s.t.
T(x1, x2, x3) = (0, 0). That is,
 x1   x1   t 
 1  1  2    0     
 1 2   x2       x2    t , t  R
 3
 x3   0
 x3   t 
Therefore, ker(T) = { t (1, 1,1): tR } = span{ (1, 1,1) }

MAT1041 - Chapter 6 6-22

Section 6-2

Thm 6.3: Kernel is a subspace

The kernel of a linear transformation T: V  W is a subspace of the domain V.

Proof: 1. ker(T) is a nonempty subset of V.

2. Let u and v be vectors in ker(T). Then
T(u + v) = T(u) + T(v) = 0 + 0 = 0 (vector addition)
Thus, u + v is in the kernel
3. If c is any scalar, then T(cu) = cT(u) = c0 = 0
(scalar multiplication), Thus, cu is in the kernel.

 The kernel of T sometimes called the nullspace of T.

MAT1041 - Chapter 6 6-23

Section 6-2

Ex 6: Finding a basis for kernel

Let T: R5  R4 be defined by T(x) = Ax, where x is in R5
 1 2 0 1  1
 2 1 3 1 0 .
and A
 1 0 2 0 1
 
 0 0 0 2 8
Find a basis for ker(T) as a subspace of R5.

MAT1041 - Chapter 6 6-24

Section 6-2

Example 6 (cont.)
 x1  0  x1   2  1
 1 2 0 1  1        1  2
 2 1 3 1 0      
x2 0 x2    
  x3   0   x3   s  1  t  0
 1 0 2 0 1          
  4     4 
x 0 x  0   4
 0 0 0 2 8
 x5  0  x5   0  1
Thus one basis for the kernel T is given by
B = { (2, 1, 1, 0, 0), (1, 2, 0, 4, 1) }

MAT1041 - Chapter 6 6-25

Section 6-2

Solution Space
 A basis for the kernel of a linear transformation T(x) = Ax was
found by solving the homogeneous system given by Ax = 0.

 It is the same produce used to find the solution space of Ax = 0.

MAT1041 - Chapter 6 6-26

Section 6-2 The Range of a Linear Transform

Thm 6.4: Range is a subspace

The range of a linear transformation T: V  W is a subspace of the domain W.
 range(T) = { T(v): v is in V }

 ker(T) is a subspace of V.

Proof: 1. range(T) is a nonempty because T(0) = 0.

2. Let T(u) and T(v) be vectors in range(T). Because u and v are in V,
it follows that u + v is also in V. Hence the sum
T(u) + T(v) = T(u + v) is in the range of T. (vector addition)
3. Let T(u) be a vector in the range of T and let c be a scalar. Because
u is in V, it follows that cu is also in V. Hence, cT(u) = T(cu) is in the
range of T. (scalar multiplication)

MAT1041 - Chapter 6 6-27

Section 6-2

Figure 6.6
Kernel ker(T) is a subspace of V

V Codomain

0
Domain T: V  W
W

Range
range(T) is a subspace of W

MAT1041 - Chapter 6 6-28

Section 6-2

Column Space
 To find a basis for the range of a linear transformation defined by
T(x) = Ax, observe that the range consists of all vectors b such that
the system Ax = b is consistent.
 a11 a12  a1n   x1   a11   a1n   b1 
a a22  a2 n   x2  a  a  b 
Ax   21
 x1  21     xn  2 n    2   b
            
        
am1 am 2  amn   xn   m1 
a amn  bm 

 b is in the range of T if and only if b is a linear combination of the

column vectors of A.

MAT1041 - Chapter 6 6-29

Section 6-2

Corollary of Theorems 6.3 & 6.4

Let T: Rn  Rm be the linear transformation given by T(x) = Ax.

[Theorem 6.3] The kernel of T is equal to the solution space of Ax = 0.

[Theorem 6.4] The column space of A is equal to the range of T.

MAT1041 - Chapter 6 6-30

Section 6-2

Ex 7: Finding a basis for range

Let T: R5  R4 be the linear transform given in Example 6. Find a
basis for the range of T.
Sol: The row echelon of A:

 1 2 0 1  1  1 0 2 0  1
 2 1 3 1 0  0 1  1 0  2 
A  
 1 0  2 0 1 0 0 0 1 4
   
 0 0 0 2 8 0 0 0 0 0
One basis for the range of T is
B = { (1, 2, 1, 0), (2, 1, 0, 0), (1, 1, 0, 2) }

MAT1041 - Chapter 6 6-31

Section 6-2

Rank and Nullity

Let T: V  W be a linear transformation.

 The dimension of the kernel of T is called the nullity of T and

is denoted by nullity(T).

 The dimension of the range of T is called the rank of T and is

denoted by rank(T).

MAT1041 - Chapter 6 6-32

Section 6-2

Thm 6.5: Sum of rank and nullity

 Let T: V  W be a linear transformation from an n-dimension vector
space V into a vector space W. Then the sum of the dimensions of the
range and the kernel is equal to the dimension of the domain. That is,

rank(T) + nullity(T) = n
or
dim(range) + dim(kernel) = dim(domain)

MAT1041 - Chapter 6 6-33

Section 6-2

Proof of Theorem 6.5

The linear transformation from an n-dimension vector space into an m-
dimension vector space can be represented by a matrix, i.e., T(x) = Ax
where A is an m  n matrix. Assume that the matrix A has a rank of r.
Then,
rank(T) = dim(range of T) = dim(column space) = rank(A) = r
From Theorem 4.7, we have
nullity(T) = dim(kernel of T) = dim(solution space) = n  r
Thus,
rank(T) + nullity(T) = n + (n  r) = n

MAT1041 - Chapter 6 6-34

Section 6-2

Ex 8: Finding the rank & nullity

Find the rank and nullity of T: R3  R3 defined by the matrix
1 0  2
A  0 1 1 
0 0 0 

Sol: Because rank(A) = 2, the rank of T is 2.

The nullity is dim(domain) – rank = 3 – 2 = 1.

MAT1041 - Chapter 6 6-35

Section 6-2

Ex 8: Finding the rank & nullity

Let T: R5  R7 be a linear transformation
 Find the dimension of the kernel of T if the dimension of the range is 2.

dim(kernel) = n – dim(range) = 5 – 2 = 3

 Find the rank of T if the nullity of T is 4.

rank(T) = n – nullity(T) = 5 – 4 = 1

 Find the rank of T if ker(T) = {0}.

rank(T) = n – nullity(T) = 5 – 0 = 5

MAT1041 - Chapter 6 6-36

Section 6-2 One-to-One & Onto Linear Transformation

One-to-One Mapping
 A linear transformation T :VW is said to be one-to-one if and
only if for all u and v in V, T(u) = T(v) implies that u = v.

V V

W W

T T

One-to-one Not one-to-one

MAT1041 - Chapter 6 6-37
Section 6-2

Thm 6.6: One-to-one Linear

transformation
Let T :VW be a linear transformation. Then T is one-to-one if and
only if ker(T) = {0}.

Proof:」Suppose T is one-to-one. Then T(v) = 0 can have

only one solution: v = 0. In this case, ker(T) = {0}.
」Suppose ker(T) = {0} and T(u) = T(v). Because T is a
linear transformation, it follows that
T(u – v) = T(u) – T(v) = 0
This implies that u – v lies in the kernel of T and
must therefore equal 0. Hence u – v = 0 and u = v,
and we can conclude that T is one-to-one.
MAT1041 - Chapter 6 6-38
Section 6-2

Example 10
 The linear transformation T: Mm,n  Mn,m given by T ( A)  AT is
one-to-one because its kernel consists of only the m  n zero matrix.

 The zero transformation T: R3  R3 is not one-to-one because its

kernel is all of R3.

MAT1041 - Chapter 6 6-39

Section 6-2

Onto Linear Transformation

 A linear transformation T :VW is said to be onto if every
element in W has a preiamge in V.

 T is onto W when W is equal to the range of T.

 [Thm 6.7] Let T :VW be a linear transformation, where W is

finite dimensional. Then T is onto if and only if the rank of T is
equal to the dimension of W, i.e., rank(T) = dim(W).

 One-to-one: ker(T) = {0} or nullity(T) = 0

MAT1041 - Chapter 6 6-40
Section 6-2

Thm 6.8: One-to-one and onto

linear transformation
 Let T :VW be a linear transformation with vector spaces V and W
both of dimension n. Then T is one-to-one if and only if it is onto.

Proof:
」If T is one-to-one, then ker(T) = {0} and dim(ker(T)) = 0. In
this case, dim(range of T) = n – dim(ker(T)) = n = dim(W).
By Theorem 6.7, T is onto.

」If T is onto, then dim(range of T) = dim(W) = n. Which by

Theorem 6.5 implies that dim(ker(T)) = 0.
By Theorem 6.6, T is onto-to-one.
MAT1041 - Chapter 6 6-41
Section 6-2

Example 11
The linear transformation T:RnRm is given by T(x) = Ax. Find the nullity
and rank of T and determine whether T is one-to-one, onto, or either.
1 2 0 1 2 1 2 0
    1 2 0  0 1 1 
(a) A  0 1 1, (b) A  0 1, (c) A    , ( d ) A   
 0 1  1
0 0 1 0 0 0 0 0

T:RnRm dim(domain) rank(T) nullity(T) one-to-one onto

(a) T:R3 R3 3 3 0 Yes Yes
(b) T:R2 R3 2 2 0 Yes No
(c) T:R3 R2 3 2 1 No Yes
(d) T:R3 R3 3 2 1 No No

MAT1041 - Chapter 6 6-42

 6.3 Matrices for
Linear Transformation
 Which one is better?
T ( x1, x2 , x3 )  (2 x1  x2  x3 ,  x1  3x2  2 x3 , 3x2  4 x3 )
 2 1  1  x1  Simpler to write. Simpler
T (x)  Ax   1 3  2  x2  to read, and more adapted
 0 3 4   x3  for computer use.

 The key to representing a linear transformation T:VW by a matrix is

to determine how it acts on a basis of V.

 Once you know the image of every vector in the basis, you can use the
properties of linear transformations to determine T(v) for any v in V.
MAT1041 - Chapter 6 6-43
Section 6-3

Thm 6.10: Standard matrix for

a linear transformation
Let T: RnRm be a linear transformation such that
1  a11  0  a12  0  a1n 
0   a  1  a  0   a 
T (e1 )  T (   )   21 , T (e 2 )  T (   )   22 ,  , T (e n )  T (   )   2 n 
           
           
   m1 
0 a    m2 
0 a 1 amn 
Then the m  n matrix whose n columns corresponds to T (ei ), is such
that T(v) = Av for every v in Rn. A is called the standard matrix for T.
 a11 a12  an1 
a a22  an 2 
A   21
    
 
am1 am 2  amn 
MAT1041 - Chapter 6 6-44
Section 6-3

Proof of Theorem 6.10

Let v  v1 v2  vn T  v1e1  v2e2    vnen  R n .
Because T is a linear transformation, we have
T ( v)  T (v1e1  v2e 2    vne n )
 T (v1e1 )  T (v2e 2 )    T (vnen )
 v1T (e1 )  v2T (e 2 )    vnT (en )
On the other hand,
 a11 a12  an1   v1   a11   a12   a1n 
a a  a  v  a  a  a 
Av   21 22 n 2     v    v      v  2n 
2 21 22

      
1
  2   n
 
        
 m1 m 2
a a  a mn   n 
v  m1 
a  m2 
a amn 
 v1T (e1 )  v2T (e 2 )    vnT (e n )  T ( v)
MAT1041 - Chapter 6 6-45
Section 6-3

Example 1
Find the standard matrix for the linear transformation T: R3R2
defined by T(x, y, z) = ( x – 2y, 2x + y)
Sol:  1 
    1 
T (e1 )  T (1, 0, 0)  (1, 2)  T (e1 )  T  0    
 0    2 
  
 0  
      2
T (e 2 )  T (0, 1, 0)  (2, 1)  T (e 2 )  T  1    
 0   1
 
 0  
    0 
T (e3 )  T (0, 0, 1)  (0, 0)  T (e3 )  T  0    
 1  0
  
MAT1041 - Chapter 6 6-46
Section 6-3

Example 1 (cont.)
 1  2 0
A  T (e1 ) T (e3 )  
T (e2 ) 
 2 1 0 
Note that
 x  x
   1  2 0    x  2 y 
A y      y   

 z    z  
2 1 0 2 x y
 
which is equivalent to T(x, y, z) = ( x – 2y, 2x + y).

MAT1041 - Chapter 6 6-47