VIETNAM NATIONAL UNIVERSITY-HCMC

International University
Lecture Notes/Slides for

APPLIED LINEAR ALGEBRA

Chapter 4
Linear Transformations

Chapter 4, Linear Transformations Page 1/58

Transformation by Matrix Multiplication

Example
 
1 2 0
Consider the matrix A = . By matrix multiplication, A
2 1 0
transforms vectors inR3 into 2
 vectors in R .
x
Consider the vector y . Transforming this vector by A looks like:

z
 
  x  
1 2 0  y = x + 2y
2 1 0 2x + y
z

For example:  
  1  
1 2 0  2 = 5
2 1 0 4
3
Chapter 4, Linear Transformations Transformations Page 2/58
 Transformations
Definition
A transformation is a function T : Rn → Rm , sometimes written
T
Rn → Rm ,
and is called a transformation from Rn to Rm . If m = n, then we say T is
a transformation of Rn .

What do we mean by a function?

Informally, a function T : Rn → Rm is a rule that assigns exactly one
vector of Rm to each vector of Rn .
We use the notation T (~x ) to mean the transformation T applied to the
vector ~x .

Definition
If T acts by matrix multiplication of a matrix A (such as the previous
example), we call T a matrix transformation, and write TA (~x ) = A~x .
Chapter 4, Linear Transformations Transformations Page 3/58

Equality of Transformations

Definition
Suppose S : Rn → Rm and T : Rn → Rm are transformations. Then
S = T if and only if S(~x ) = T (~x ) for every ~x ∈ Rn .

Chapter 4, Linear Transformations Transformations Page 4/58

 Specifying the Action of a Transformation

Example
T : R3 → R4 defined by  
 a+b 
a  b+c 
T b =  a−c 

c
c −b
 
1
is a transformation that transforms the vector 4  in R3 into the vector

7
   
  1+4 5
1  4 + 7   11 
T 4 =
 1 − 7  =  −6
  .

7
7−4 3

Chapter 4, Linear Transformations Transformations Page 5/58

Linear Transformations

Definition
A transformation T : Rn → Rm is a linear transformation if it satisfies the
following two properties for all ~x , ~y ∈ Rn and all (scalars) a ∈ R.
1 T (~x + ~y ) = T (~x ) + T (~y ) (preservation of addition)
2 T (a~x ) = aT (~x ) (preservation of scalar multiplication)

Chapter 4, Linear Transformations Linear Transformations Page 6/58

 Properties of Linear Transformations
Let T : Rn → Rm be a linear transformation, and let ~x ∈ Rn . Since T
preserves scalar multiplication,
1 T (0~x ) = 0T (~x ) implying T (~0) = ~0, so T preserves the zero vector.
2 T ((−1)~x ) = (−1)T (~x ), implying T (−~x ) = −T (~x ), so T preserves
the negative of a vector.
Suppose ~x1 , ~x2 , . . . , ~xk are vectors in Rn and

~y = a1 ~x1 + a2 ~x2 + · · · + ak ~xk

for some a1 , a2 , . . . , ak ∈ R. Then

3

T (~y ) = T (a1 ~x1 + a2 ~x2 + · · · + ak ~xk )

= a1 T (~x1 ) + a2 T (~x2 ) + · · · + ak T (~xk ),
i.e.,T preserves linear combinations.

Chapter 4, Linear Transformations Linear Transformations Page 7/58

Problem
Let T : R3 → R4 be a linear transformation such that
   
  4   4  
1 4 −7
 4 
 and T  0  =  5  . Find T  3  .
 
T 3 =  −1 
 0 
1 5 −9
−2 5

Solution
The only way it is possible to solve this problem is if
     
−7 1 4
 3  is a linear combination of  3  and  0 ,
−9 1 5
i.e., if there exist a, b ∈ R so that
     
−7 1 4
 3  = a 3  + b 0 .
−9 1 5

Chapter 4, Linear Transformations Linear Transformations Page 8/58

 Solution (continued)
To find a and b, solve the system of three equations in two variables:
   
1 4 −7 1 0 1
 3 0 3  → · · · →  0 1 −2 
1 5 −9 0 0 0

Thus a = 1, b = −2, and

     
−7 1 4
 3  =  3  − 2 0 .
−9 1 5

Chapter 4, Linear Transformations Linear Transformations Page 9/58

Solution (continued)
We now use that fact that linear transformations preserve linear combinations,
implying that
     
−7 1 4
T  3  = T  3  − 2  0 
−9 1 5
   
1 4
= T 3 − 2T 0 
  
1 5
     
4 4 −4
 − 2  5  =  −6 
 4     
=   0   −1   2 
−2 5 −12
 
 −4 
−7  −6 
Therefore, T  3 = .
 2 
−9
−12

Chapter 4, Linear Transformations Linear Transformations Page 10/58

 Problem
Let T : R4 → R3 be a linear transformation such that
     
1   0   2
 1  2  −1  5  5 
  3  and T 
T  0 =  1  = 0 . Find T  −3
    .

−1 1
−2 1 −7

Final Answer
 
2  
 5  −11
T
 −3  =
  6 .
−5
−7

Chapter 4, Linear Transformations Linear Transformations Page 11/58

Matrix Transformations
Theorem
Every matrix transformation is a linear transformation.

Proof.
Suppose T : Rn → Rm is a matrix transformation induced by the m × n matrix A,
i.e., T (~x ) = A~x for each ~x ∈ Rn . Let ~x , ~y ∈ Rn and let a ∈ R. Then

T (~x + ~y ) = A(~x + ~y ) = A~x + A~y = T (~x ) + T (~y ),

proving that T preserves addition. Also,

T (a~x ) = A(a~x ) = a(A~x ) = aT (~x ),

proving that T preserves scalar multiplication.

Since T preserves addition and scalar multiplication T is a linear

transformation.

Chapter 4, Linear Transformations Linear Transformations Page 12/58

 Some Special Matrix Transformations
Example (The Zero Transformation)
If A is the m × n matrix of zeros, then the transformation T : Rn → Rm induced
by A is called the zero transformation because for every vector ~x in Rn

T (~x ) = A~x = 0~x = ~0.

Note that the first zero is the matrix A, while the second zero is the zero vector
of Rm .
The zero transformation is usually written as T = 0.

Example (The Identity Transformation)

The transformation of Rn induced by In , the n × n identity matrix, is called the
identity transformation because for every vector ~x in Rn ,

T (~x ) = In ~x = ~x .

The identity transformation on Rn is usually written as 1Rn .

Chapter 4, Linear Transformations Linear Transformations Page 13/58

Example (Revisited)
Recall T : R3 → R4 defined by
 
 a+b

a  b+c 
T b =

 a−c 

c
c −b

Is T a matrix 
transformation? 
1 1 0
 0 1 1 
Consider A =  , then
 1 0 −1 
0 −1 1
   
  1 1 0   a+b 
a  0 a a
1 1 
  b+c 
A b  =  b = =T b 
 1 0 −1   a−c 
c c c
0 −1 1 c −b
So in this case T is a matrix transformation!

Chapter 4, Linear Transformations Linear Transformations Page 14/58

 Not all transformations are matrix transformations!

Example
Consider T : R2 → R2 defined by
 
1
T (~x ) = ~x + for all ~x ∈ R2 .
−1

Why is T not a matrix transformation?

Chapter 4, Linear Transformations Linear Transformations Page 14/58

Example (continued)
We have T : R2 → R2 defined by
 
1
T (~x ) = ~x + for all ~x ∈ R2 .
−1

Since every matrix transformation is a linear transformation, we consider

T (~0, where ~0 is the zero vector of R2 .
         
0 0 1 1 0
T = + = 6= ,
0 0 −1 −1 0

violating one of the properties of a linear transformation.

Therefore, T is not a linear transformation, and hence is not a matrix

transformation.
Can you see any other reasons why T is not a matrix transformation?

Chapter 4, Linear Transformations Linear Transformations Page 15/58

 Matrix Transformations

Theorem
Let T : Rn → Rm be a linear transformation. Then we can find an n × m
matrix A such that

T (~x ) = A~x
In this case, we say that T is induced, or determined, by A and we write

TA (~x ) = A~x

Corollary
A transformation T : Rn → Rm is a linear transformation if and only if it is
a matrix transformation.

Chapter 4, Linear Transformations Linear Transformations Page 16/58

Problem
 
 a+b 
a  b+c 
The transformation T : R3 → R4 defined by T  b  =   a−c
 for

c
c −b
3
each ~x ∈ R is another matrix transformation, that is, T (~x ) = A~x for some
matrix A. Can you find a matrix A that works?

Solution
First, since T : R3 → R4 , we know that A must have size 4 × 3. Now
consider the product
   
? ? ?   a+b
 ? ? ?  a
 b  =  b + c ,
 

 ? ? ?   a−c 
c
? ? ? c −b

and try to fill in the values of the matrix.

Chapter 4, Linear Transformations Linear Transformations Page 17/58
 Solution (continued)
We can deduce from the product that T is induced by the matrix
 
1 1 0
 0 1 1 
A= .
 1 0 −1 
0 −1 1

Chapter 4, Linear Transformations Linear Transformations Page 18/58

Is there an easier way to find the matrix of T ?

For some transformations guess and check will work, but this is not an
efficient method. The next theorem gives a method for finding the matrix
of T .

Definition
e1 , e~2 , . . . , e~n } of In is called the standard basis of Rn .
The set of columns {~

Chapter 4, Linear Transformations Linear Transformations Page 19/58

 Matrix and Linear Transformations

Good News!
There is an easy way to find the matrix of T !

Definition
e1 , e~2 , . . . , e~n } of In is called the standard basis of Rn .
The set of columns {~

Theorem
Let T : Rn → Rm be a linear transformation.
Then T is a matrix transformation.
Furthermore, T is induced by the unique matrix
 
A = T (~ e2 ) · · · T (~
e1 ) T (~ en ) ,

where e~j is the j th column of In , and T (~

ej ) is the j th column of A.

Chapter 4, Linear Transformations Linear Transformations Page 20/58

Problem
Let T : R2 → R2 be a linear transformation defined by
   
x x + 2y
T =
y x −y

for each ~x ∈ R2 . Find the matrix, A, of T .

Solution
To find A, we must find T (~
e1 ) and T (~
e2 ), where e~1 and e~2 are the standard basis
2
vectors of R .
           
1 1 + 2(0) 1 0 0 + 2(1) 2
T = = and T = =
0 1−0 1 1 0−1 −1

The columns T (~
e1 ) and T (~
e2 ) become the columns of A, so
 
1 2
A= ,
1 −1

and T (~x ) = A~x for every ~x ∈ R2 , so A is the matrix for T .

Chapter 4, Linear Transformations Finding the Matrix Page 21/58
 Find the Matrix of T
Problem
Sometimes T is not defined so nicely for us. Suppose T is given as
       
1 1 1 3
T = , T =
5 2 4 2

Find the matrix A of T .

Solution
We need to write e~1 and e~2 as a linear combination of the vectors provided.
So we reduce the augmented matrix having e~1 and e~2 as the third and
fourth columns:
     
1 1 1 1 1 0 1 0 −4 1
e~ e~ = → ··· →
5 4 1 2 5 4 0 1 0 1 5 −1

Chapter 4, Linear Transformations Finding the Matrix Page 22/58

Solution (continued)
   
1 1 1 0 1 0 −4 1
→ ··· →
5 4 0 1 0 1 5 −1
From this we can see that
     
1 1 1
e~1 = = −4 +5
0 5 4

and      
0 1 1
e~2 = = −
1 5 4

Chapter 4, Linear Transformations Finding the Matrix Page 23/58

 Solution (continued)
and
So 
0

T (e~2 ) = T
 
1
 
1
 1
T (e~1 ) = T −4 +5
5 4    
1 1
=T −

1
 
1
 5 4
= −4T + 5T
5 4  
1
 
1
=T −T

1
 
3
 5 4
= −4 +5
2 2    
1 3
= −

11
 2 2
=
2 
−2

=
0

Chapter 4, Linear Transformations Finding the Matrix Page 24/58

Solution (continued)
The matrix of T is the matrix whose first column is T (e~1 ), and second
column is T (~
e2 ):
 
  11 −2
A = T (~e1 ) T (~
e2 ) =
2 0

Chapter 4, Linear Transformations Finding the Matrix Page 25/58

 Determining if a Transformation is Linear
Example
 
2x  
x
Let T : R2 → R3 be a transformation defined by T = y .
y
−x + 2y
One way to show that T is a linear transformation is to show that it preserves
addition and scalar multiplication. However, now that we know that linear
transformations are matrix transformations, we can use this to our advantage.
If T were a linear transformation, then T would be induced by the matrix
 
      2 0
  1 0
A = T (~ e1 ) T (~e2 ) = T T =  0 1 .
0 1
−1 2
Since    
  2 0   2x  
x x x
A = 0 1  = y =T ,
y y y
−1 2 −x + 2y
T is a matrix transformation, and is therefore a linear transformation.

Chapter 4, Linear Transformations Finding the Matrix Page 26/58

Example
   
x xy
Let T : R2 → R2 be a transformation defined by T = .
y x +y
If T were a linear transformation, then T would be induced by the matrix
       
  1 0 0 0
A = T (~ e1 ) T (~
e2 ) = T T = .
0 1 1 1

However, 
x
 
0 0

x
 
0

A = = .
y 1 1 y x +y
   
x x
We see from this that if x = 0 or y = 0, then xy = 0, so A =T .
y y
But if we take x = y = 1, then
           
x 1 0 x 1 1
A =A = while T =T = ,
y 1 2 y 1 2
   
1 1
i.e., A 6= T . So T is not a linear transformation.
1 1
(Any more reasons?)
Chapter 4, Linear Transformations Finding the Matrix Page 27/58
 Rotations in R2

Definition
The transformation
Rθ : R2 → R2
denotes counterclockwise rotation about the origin through an angle of θ.

Rotation through an angle of θ preserves scalar multiplication.

Rotation through an angle of θ preserves vector addition.

Chapter 4, Linear Transformations 5.4: Rotations Page 28/58

Rθ is a linear transformation
Since Rθ preserves addition and scalar multiplication, Rθ is a linear
transformation, and hence a matrix transformation.

The matrix that induces Rθ can be found by computing Rθ (E1 ) and

Rθ (E2 ), where    
1 0
E1 = and E2 = .
0 1
   
1 cos θ
Rθ (E1 ) = Rθ = ,
0 sin θ
and    
0 − sin θ
Rθ (E2 ) = Rθ =
1 cos θ

Chapter 4, Linear Transformations 5.4: Rotations Page 29/58

 The Matrix for Rθ
The rotation Rθ : R2 → R2 is a linear transformation, and is induced by the
matrix  
cos θ − sin θ
.
sin θ cos θ

Chapter 4, Linear Transformations 5.4: Rotations Page 30/58

Example (Rotation through π)

We denote by
Rπ : R2 → R2
counterclockwise rotation about the origin through an angle of π.
y

(a, b)

0 x

(−a, −b)

      
a −a −1 0 a
We see that Rπ = = , so Rπ is a matrix
b −b 0 −1 b
transformation.

Chapter 4, Linear Transformations 5.4: Rotations Page 31/58

 Example (Rotation through π/2)
We denote by
Rπ/2 : R2 → R2
counterclockwise rotation about the origin through an angle of π/2.
y
(−b, a)

(a, b)

0 x

      
a −b 0 −1 a
We see that Rπ/2 = = ,
b a 1 0 b

Chapter 4, Linear Transformations 5.4: Rotations Page 32/58

Reflection in R2
Example
   
a a
In R2 , reflection in the x-axis, which transforms to , is a
b −b
matrix transformation because
    
a 1 0 a
= .
−b 0 −1 b

Example
   
a −a
In R2 , reflection in the y -axis transforms to . This is a
b b
matrix transformation because
    
−a −1 0 a
= .
b 0 1 b

Chapter 4, Linear Transformations Reflections Page 33/58

 Example
   
a b
Reflection in the line y = x transforms to .
b a
y
(b, a)
y =x

x
(a, b)

This is a matrix transformation because

    
b 0 1 a
= .
a 1 0 b

Chapter 4, Linear Transformations Reflections Page 34/58

Range of a Transformation

Let T : Rn 7→ Rm be a linear transformation given by T (~x ) = A~x .

Consider all the vectors of the form A~x for some ~x ∈ Rn . This set of
vectors is called the range or image of T .
We denote this set as T Rn , T (Rn ) or Im(T ). Notice that these vectors A~x
are in Rm .

Chapter 4, Linear Transformations Page 35/58

 The Form A~x

Theorem
Let A be an m ×n matrix
 where A1 , ..., An denote the columns of A. Then
x1
 .. 
for a vector ~x =  .  in Rn ,
xn
n
X
A~x = xk Ak
k=1

Therefore A(Rn ) is the collection of all linear combinations of these

products.

Chapter 4, Linear Transformations Page 36/58

Injections
Definition
Let T : Rn → Rm be a linear transformation, and let ~x1 and ~x2 be in Rn .
We say that T is an injection or is one-to-one (sometimes written as 1-1) if
~x1 6= ~x2 implies that
T (~x1 ) 6= T (~x2 ) .
Equivalently, if T (~x1 ) = T (~x2 ), then ~x1 = ~x2 . Thus, T is one-to-one if
two distinct vectors are never transformed into the same vector.

Theorem
Let A be an m × n matrix and let ~x be a vector of length n. Then the
transformation induced by A, TA , is one-to-one if and only if A~x = 0
implies ~x = 0 (equivalently, iff A has full column rank).

Since every linear transformation is induced by a matrix A, in order to show

that T is one to one, it suffices to show that A~x = 0 has a unique solution.

Chapter 4, Linear Transformations One to One Page 37/58

 Problem
Show that the transformation defined by
    
x 1 1 x
T =
y 0 1 y

is one-to-one.

Solution
 
1 1
Since T is a matrix transformation induced by A = , and A has full
0 1
column rank (=2), it follows that A~x = 0 has the unique solution ~x = 0. From
the previous theorem we conclude that T is one-to-one.
Alternatively, observe that
   
1 1 0 1 0 0
→ .
0 1 0 0 1 0
 
0
has unique solution ~x = , so T is a one-to-one. (Any other ways?)
0

Chapter 4, Linear Transformations One to One Page 38/58

Surjections

Definition
Let T : Rn → Rm be a linear transformation. We say that T is a surjection
or onto if, for every ~b ∈ Rm there exists an ~x ∈ Rn so that T (~x ) = ~b.

Example
Let T : R2 → R2 be the linear transformation defined by
     
a a+b a
T = for all ∈ R2 .
b 0 b
 
0
Then T is not onto. To see why, choose ~b = ∈ R2 . Then there is no
1
vector ~x ∈ R2 so that T (~x ) = ~b; applying T to any vector results in a
vector whose second entry is 0, and the second entry of ~b is 1.

Chapter 4, Linear Transformations Onto Page 39/58

 Example (continued)
Consider the system A~x = ~b, where A is the matrix induced by T ,
 
1 1
A= .
0 0
Then the augmented matrix is
 
1 1 0
,
0 0 1

which is already in reduced row-echelon form. The fact that this system is
inconsistent implies that T is not onto.

Theorem
Let A be an m × n matrix. Then the transformation TA , induced by A, is
onto if and only if A~x = ~b is consistent for every vector ~b in Rm .

Chapter 4, Linear Transformations Onto Page 40/58

Problem
Show that the transformation defined by
    
x 1 2 x
T =
y 3 5 y

is onto.

Solution
 
1 2
T is a matrix transformation, T (~x ) = A~x , where A = .
3 5
Since det(A) = −1 6= 0, A is invertible.
Therefore, for every choice of ~b, the system A~x = ~b has an unique solution
(namely ~x = A−1 ~b) . So T is onto.

Chapter 4, Linear Transformations Onto Page 41/58

 Not one-to-one

Problem
 
1 2 −1
Let T be the linear transformation induced by A = . Show
3 5 0
that TA is not one-to-one.

Solution
Let R be a row-echelon form of A.
   
1 2 −1 1 2 −1
A= → =R
3 5 0 0 1 −3

Since A has rank two, A~x = 0 has infinitely many solutions, so ~x = 0 is not
the only solution. Therefore, TA is not one-to-one.

Chapter 4, Linear Transformations Examples Page 42/58

One-to-one

Problem
 
1 −1
Let T be the linear transformation induced by A =  2 2 . Show that TA
−1 2
is one-to-one.

Solution
Let R be a row-echelon form of A.
   
1 −1 1 −1
A=  2 2 → 0
  1 =R
−1 2 0 0
Since A has rank two, every variable in A~x = 0 is a leading variable, so ~x = 0 is
the unique solution. Therefore, TA is one-to-one.

Chapter 4, Linear Transformations Examples Page 43/58

 One-to-one and onto

Problem
 
1 −1
Let T be the linear transformation induced by A = . Show that
2 −1
TA is one-to-one and onto.

Solution
Let R be a row-echelon form of A.
   
1 −1 1 −1
A= → =R
2 −1 0 1

In this case, A is invertible, so A~x = ~b has a unique solution ~x for every ~b

in R2 . Therefore TA is both one-to-one and onto.

Chapter 4, Linear Transformations Examples Page 44/58

Neither one-to-one nor onto

Problem
 
1 −1 1
Let T be the linear transformation induced by A =  −1 2 1 . Show that TA is
1 0 3
neither one-to-one nor onto.

Solution
Let R be a row-echelon form of A.
   
1 −1 1 1 −1 1
A=  −1 2 1  →  0 1 2 =R
1 0 3 0 0 0
Since A has rank two, the augmented matrix [A|~ b] will have rank three for some choice
~ 3
of b ∈ R , resulting in A~ ~
x = b being inconsistent. Therefore, TA is not onto.
The augmented matrix [A|0] has rank two, so the system A~ x = 0 has a non-leading
variable, and hence does not have unique solution ~
x = 0. Therefore, TA is not
one-to-one.

Chapter 4, Linear Transformations Examples Page 45/58

 Matrix of a One to One or Onto Transformation

Theorem (5.34, Matrix of a One to One or Onto Transformation)

Let T : Rn → Rm be a linear transformation induced by the m × n matrix A. Then T is
one to one if and only if the rank of A is n. T is onto if and only if the rank of A is m.

Chapter 4, Linear Transformations Examples Page 46/58

Kernel and Image

Definition (Kernel)
Let V be a subspace of Rn and W a subspace of Rm , and let T : V 7→ W be a linear
transformation.
v ) = ~0.
v ∈ V such that T (~
Then the kernel of T , ker(T ), consists of all ~
n o
ker(T ) = ~ v ) = ~0
v ∈ V : T (~
If A is the matrix corresponding to T , then ker(T ) is the null-space of A.

Definition (Image)
Let V be a subspace of Rn and W a subspace of Rm , and let T : V 7→ W be a linear
transformation.
~ ∈ W such that w
Then the image of T , im(T ), consists of all w v ) for some
~ = T (~
v ∈ V.
~
im(T ) = {T (~ v ∈ V}
v) : ~
If A is the matrix corresponding to T , then im(T ) is the column space of A.

Chapter 4, Linear Transformations Examples Page 47/58

 Example
Let T : R3 7→ R2 be defined by
 
a  
a + b + c
T  b  =
c −a
c

Then T is a linear transformation. Find a basis for ker(T ) and im(T ).

Solution
You can (and should!) verify that T is a linear transformation.

Chapter 4, Linear Transformations Finding the Kernel and Image Page 48/58

Solution (continued)
Kernel of T : We look for all vectors ~x ∈ R3 such that T (~x ) = ~0.


a    
a + b + c 0
T  b  = =
c −a 0
c

This gives a system of equations:

a+b+c = 0
c −a = 0

The general solution is

       
a  t   1 
 b  =  −2t  : t ∈ R = t  −2  : t ∈ R
c t 1
   

 
 1 
And therefore a basis for the kernel is  −2  .
1
 
Chapter 4, Linear Transformations Finding the Kernel and Image Page 49/58
 Solution (continued)
Image of T : We can write the image as
  
a+b+c
im(T ) = : a, b, c ∈ R
 c − a
     
a b c
= + + : a, b, c ∈ R
  −a  0  c   
1 1 1
= a +b +c : a, b, c ∈ R
−1 0 1

    
1 1 1
Thus im(T ) = span , , .
−1 0 1
These vectors are not linearly independent, but the first two are so a basis
for the image of T is    
1 1
, .
−1 0

Chapter 4, Linear Transformations Finding the Kernel and Image Page 50/58

Kernel and One to One

The kernel of a linear transformation gives important information about

whether the transformation is one to one. Recall that a linear
transformation T is one to one if and only if T (~x ) = ~0 implies ~x = ~0.

Theorem
Let T : V 7→ W be a linear transformation where V is a subspace of Rn
and W a subspace of Rm . n o
Then T is one to one if and only if ker(T ) = ~0 .

Chapter 4, Linear Transformations Finding the Kernel and Image Page 51/58
 Dimension of the Kernel and Image

Theorem (Dimension Theorem)

Let T : V 7→ W be a linear transformation where V is a subspace of Rn
and W is a subspace of Rm . Suppose further that the dimension of V is k.
Then
k = dim(ker(T )) + dim(im(T ))

Chapter 4, Linear Transformations Dimension Page 52/58

Example (Revisited)
Let T : R3 7→ R2 be defined by
 
a  
a + b + c
T  b  =
c −a
c

Find the dimension of ker(T ) and im(T ).

Solution
We already know that a basis for the kernel of T is given by
 
 1 
 −2 
1
 

Therefore dim(ker(T )) = 1.

Chapter 4, Linear Transformations Dimension Page 53/58

 Solution (continued)
We also found a basis for the image of T as
   
1 1
,
−1 0

and this of course shows dim(im(T )) = 2.

But we could have found the dimension of im(T ) without finding a basis.
That’s because since the dimension of R3 is 3, and the dimension of
ker(T ) is 1, we get by the Dimension Theorem that:

dim(im(T )) = dim(R3 ) − dim(ker(T ))

= 3−1=2

Chapter 4, Linear Transformations Dimension Page 54/58

Definition
Suppose a linear system of equations can be written in the form

T (~x ) = ~b

where T is a linear transformation determined by T (~x ) = A~x .

If T (~xp ) = ~b, then ~xp is called a particular solution of the linear system

Example
Suppose that a system of linear equations can be written as
   
x1 5 + 2x2 − x4
 x2   x2 
 x3  =  −4 − 3x4
T   

x4 x4

Setting
 x2 = 2 and
 x4  −1 gives us the particular solution
=
x1 10
 x2   2 
~xp =  x3  =  −1
  .

x4 −1
Chapter 4, Linear Transformations The General Solution to a Linear System Page 55/58
 Null Space and Associated Homogeneous System

Definition
Let T be a linear transformation. Define:
n o
~
ker (T ) = ~x : T (~x ) = 0

The kernel, ker (T ) consists of the set of all vectors ~x such that T (~x ) = ~0.
This is also called the null space of T or the solution space of T (~x ) = ~0.

Definition
Let A~x = ~b be a system of equations. Then, the corresponding system
given by A~x = ~0, found by replacing ~b with ~0 is the associated
homogeneous system.

Chapter 4, Linear Transformations The General Solution to a Linear System Page 56/58

Associated Homogeneous System

Example
x1 + x2 − x3 + 3x4 = 2
Consider the system given by −x1 + 4x2 + 5x3 − 2x4 = 3 .
x1 + 6x2 + 3x3 + 4x4 = 4
We can write this as
 
  x1  
1 1 −1 3  2
 −1 4 x2

5 −2  
 x3  = 3
  
1 6 3 4 4
x4

whose associated homogeneous system can be written as:

 
  x1  
1 1 −1 3  0
 −1 4 x2

5 −2    x3  = 0
  
1 6 3 4 0
x4
Chapter 4, Linear Transformations The General Solution to a Linear System Page 57/58
 Solutions to Systems as Linear Combinations of Vectors

Basic Solutions to Homogeneous Systems

In our earlier treatment of homogeneous systems of linear equations, we
saw how to write the solution to a homogeneous system of linear equations
as a linear combination of basic solutions.

Example
x1 + x2 − x3 + 3x4 = 0
The system −x1 + 4x2 + 5x3 − 2x4 = 0 has general solution
x1 + 6x2 + 3x3 + 4x4 = 0
9 14
x1 = 5
s − 5
t    9  
− 14

x1 5 5
− 45 s − 15 t
 4
x2 =  x2   −5  − 51
  
or 
 x3  = s  +t .
  
 1   0 
x3 = s x4
x4 = t 0 1

Chapter 4, Linear Transformations The General Solution to a Linear System Page 58/58