Linear Transformations and Matrices

Elif Tan

Ankara University

Elif Tan (Ankara University) Linear Transformations 1 / 28

Linear Transformations
De…nition (Coordinate)
Let S = fv1 , v2 , ..., vn g be an ordered basis for the n-dimensional vector
space (V , , ) . Then every vector v in V can be uniquely expressed in
the form
v = c1 v1 c2 v2 ... cn vn
where c1 , c2 , ..., cn are scalars. The coordinate vector of v with respect
to the ordered basis S is de…ned by
2 3
c1
6 c2 7
6 7
[v ]S := 6 .. 7 .
4 . 5
cn

The entries of [v ]S are called the coordinates of v with respect to the

basis S. Note that there is a one-to-one correspondence between v and
[v ]S .
Elif Tan (Ankara University) Linear Transformations 2 / 28
Linear Transformations

1 2
Example: S = , is a basis for the vector space R2 .
1 1
x
Every vector v = 2 R2 can be written uniquely as
y

x 1 2
= (2y x) (x y) .
y 1 1

Thus the coordinates of v with respect to the basis S is

2y x
[v ]S : = .
x y

If we change the order of the basis, then the coordinates of v change as

x y
.
2y x

Elif Tan (Ankara University) Linear Transformations 3 / 28

Linear Transformations

De…nition (Transition Matrix)

Let S = fv1 , v2 , ..., vn g and T = fw1 , w2 , ..., wn g be an ordered basis for
the n-dimensional vector space (V , , ) . The transition matrix from
the basis T to S is de…ned by

PS T = [[w1 ]S [w2 ]S . . . [wn ]S ]n n

and the coordinate vector of v wrt S can be written as

[ v ] S = PS T [v ]T .

Note that the transition matrix is nonsingular matrix and we have

PS 1 T = PT S.

Elif Tan (Ankara University) Linear Transformations 4 / 28

Linear Transformations

Example
Consider the ordered basis for R3
8 2 3 2 3 2 39
< 1 0 0 =
S = v1 = 0 , v2 = 1 , v3 = 0 5
4 5 4 5 4
: ;
0 0 1

and 8 3 2 2 3 2 39
< 1 2 3 =
T = w1 = 4 2 5 , w2 = 4 1 5 , w3 = 4 3 5 .
: ;
1 1 1
Find the transition matrix from the basis T to S.

Elif Tan (Ankara University) Linear Transformations 5 / 28

Linear Transformations

2 3 2 3
a1 1
w1 = a1 v1 a2 v2 a3 v3 ) [w1 ]S = 4 a2 5 = 4 2 5
a3 1
2 3 2 3
b1 2
w2 = b1 v1 b2 v2 b3 v3 ) [w2 ]S = 4 b2 5 = 4 1 5
b3 1
2 3 2 3
c1 3
w3 = c1 v1 c2 v2 c3 v3 ) [w3 ]S = 4 c2 5 = 4 3 5
c3 1
2 3
1 2 3
PS T = [[w1 ]S [w2 ]S [w3 ]S ] = 4 2 1 3 5.
1 1 1

Elif Tan (Ankara University) Linear Transformations 6 / 28

Linear Transformations

Example: Consider the ordered basis for R3

8 2 3 2 3 2 39
< 1 1 0 =
S = v1 = 4 0 5 , v2 = 4 1 5 , v3 = 4 0 5
: ;
1 0 1

and
T = fw1 , w2 , w3 g .
If the transition matrix from the basis T to S is
2 3
1 1 2
PS T = 4 2 1 1 5,
1 1 1

then …nd the basis T .

Elif Tan (Ankara University) Linear Transformations 7 / 28

Linear Transformations
2 3
1 1 2
Since PS T = [[w1 ]S [w2 ]S [w3 ]S ] = 4 2 1 1 5,
1 1 1
3 2 2 3
a1 1
w1 = a1 v1 a2 v2 a3 v3 ) [w1 ]S = 4 a2 5 = 4 2 5
a3 1
2 3 2 3 2 3 2 3
1 1 0 3
) w1 = 1 4 0 5 2 4 1 5 1 4 0 5=4 2 5
1 0 1 0
3 2 2 3
2 3
Similarly w2 = 4 1 5 , w3 = 4 1 5 .
0 3

Elif Tan (Ankara University) Linear Transformations 8 / 28

Linear Transformations

De…nition (Linear Transformation)

Let (V , , ) and (W , , ) be real vector spaces. L : V ! W is called
a linear transformation if the following conditions holds:

(i ) L (u v ) = L (u ) L (v ) for all u, v 2 V

(ii ) L (c u) = c L (u ) for all u 2 V and all c 2 R.

Elif Tan (Ankara University) Linear Transformations 9 / 28

Linear Transformations

De…nition
A linear transformation L : V ! W is called one-to-one if L (v1 ) = L (v2 )
implies that v1 = v2 for v1 , v2 2 V .
A linear transformation L : V ! W is called onto if for each
w 2 W , 9v 2 V such that L (v ) = w .

De…nition
Let (V , , ) and (W , , ) be real vector spaces. L is called an
isomorphism if L : V ! W is a linear transformation that is one-to-one
and onto. In this case the vector spaces V and W are called isomorphic
and denoted by V = W .

Elif Tan (Ankara University) Linear Transformations 10 / 28

Linear Transformations

De…nition
Let L : V ! W be a linear transformation.
The kernel of L is de…ned by

KerL = fv 2 V j L (v ) = 0W g .

The range of L is de…ned by

RangeL = L (V ) = fw 2 W j 9v 2 V ; L (v ) = w g .

Elif Tan (Ankara University) Linear Transformations 11 / 28

Linear Transformations

Theorem
Let L : V ! W be a linear transformation. Then we have the following
results:
1 L (0V ) = 0W
2 KerL < V
3 L is one-to-one, KerL = f0V g
4 RangeL < W
5 L is onto, L (V ) = W .

Elif Tan (Ankara University) Linear Transformations 12 / 28

Linear Transformations

Theorem (Rank-Nullity Theorem)

Let L : V ! W be a linear transformation with dimV = n, then

dim V = dim
| {zKerL} + dim RangeL.
| {z }
"nullity"of L "rank"of L

Elif Tan (Ankara University) Linear Transformations 13 / 28

Linear Transformations
Example
31 02
x1
x1 + 2x2 + 3x3
Let L : R3 ! R2 , L @4 x2 5A = be a linear
2x1 + x2 3x3
x3
transformation. Find the rank of L.
Solution:
82 3 02 31 9
< x1 x1 =
KerL = 4 x2 5 2 R3 j L @4 x2 5A = 0W
: ;
x3 x3
82 3 9
< x1 =
4 x1 + 2x2 + 3x3 0
= x2 5 2 R3 j =
: 2x1 + x2 3x3 0 ;
x3
82 3 9
< 3x3 =
= 4 3x3 5 j x3 2 R .
: ;
x3
Elif Tan (Ankara University) Linear Transformations 14 / 28
Linear Transformations

82 39
< 3 =
Thus 4 3 5 is a basis for KerL and dim KerL = 1. Since
: ;
1

dim V = dim KerL + dim RangeL,

then we have
3 = 1 + rankL.
Therefore rankL = 2.

Elif Tan (Ankara University) Linear Transformations 15 / 28

Linear Transformations

Theorem
1 Let V be an n-dimensional vector space. Then V = Rn .
2 Let V and W be …nite dimensional vector spaces.

V = W , dim V = dim W .
2 3
a
a b 6 b 7
Example: L : M22 ! R4 , L =6 7
4 c 5 is an isomorphism. So
c d
d

M22 = R4 .

Elif Tan (Ankara University) Linear Transformations 16 / 28

LINEAR TRANSFORMATIONS and MATRICES

De…nition (Matrix representation of a linear transformation)

Let L : V ! W be a linear transformation and consider the ordered basis
S = fv1 , v2 , ..., vn g and T = fw1 , w2 , ..., wm g for the vector spaces V and
W , respectively. The matrix representation of the linear transformation L
with respect to the basis S and T is de…ned by

A = [[L (v1 )]T [L (v2 )]T . . . [L (vn )]T ]m n .

Also for v 2 V , we have

[L (v )]T = A [v ]S .

Elif Tan (Ankara University) Linear Transformations 17 / 28

Linear Transformations and Matrices
Theorem
Let L : Rn ! Rm be a linear transformation and consider the standard
basis fe1 , e2 , ..., en g for Rn . Let A = [L (e1 ) L (e2 ) . . . L (en )]m n . The
matrix A is the only matrix satis…ying the property;

L (x ) = Ax, for x 2 Rn .

It is called the standard matrix representation of the linear

transformation L.

Remark: Linear transformation L $ A

If A is m n matrix, then there is a corresponding linear
transformation L : Rn ! Rm such that L (x ) = Ax, for x 2 Rn .
If L : Rn ! Rm is a linear transformation, then there is a
corresponding m n matrix A which is de…ned by
A = [L (e1 ) L (e2 ) . . . L (en )]m n .
Elif Tan (Ankara University) Linear Transformations 18 / 28
Linear Transformations and Matrices

Example
31 02
x1
x1 + 2x2 + 3x3
Let L : R3 ! R2 , L @4 x2 5A = be a linear
2x1 + x2 3x3
x3
transformation and consider the standard basis
8 2 3 2 3 2 39
< 1 0 0 =
S = v1 = 4 0 5 , v2 = 4 1 5 , v3 = 4 0 5
: ;
0 0 1

and
1 0
T = w1 = , w2 =
0 1
for the vector spaces R3 and R2 , respectively.
Find the matrix representation of the linear transformation L with respect
to the basis S and T .
Elif Tan (Ankara University) Linear Transformations 19 / 28
Linear Transformations and Matrices

Solution:
a1 1
L ( v 1 ) = a1 w1 a2 w2 ) [L (v1 )]T = =
a2 2
b1 2
L (v2 ) = b1 w1 b2 w2 ) [L (v2 )]T = =
b2 1
c1 3
L (v3 ) = c1 w1 c2 w2 ) [L (v3 )]T = =
c2 3

1 2 3
A = [[L (v1 )]T [L (v2 )]T [L (v3 )]T ]2 3 = .
2 1 3

Elif Tan (Ankara University) Linear Transformations 20 / 28

Linear Transformations and Matrices
Example
Find the linear transformation which corresponds to the matrix
1 2 3
A= .
2 1 3 2 3

Solution: The corresponding linear transformation is de…ned by

L : R3 ! R2 ,
02 31 2 3
x1 x1
L @4 x2 5A = A 4 x2 5
x3 x3
2 3
x1
1 2 3 4 x2 5
=
2 1 3
x3
x1 + 2x2 + 3x3
= .
2x1 + x2 3x3
Elif Tan (Ankara University) Linear Transformations 21 / 28
Linear Transformations and Matrices

To …nd the rank of the linear transformation L : Rn ! Rm , it is

enough to check the rank of the matrix A. The rank of an m n
matrix A is the number of nonzero rows in the reduced row echelon
form of the matrix A.

dim Rn = dim KerL + dim L (Rn )

n = nullity A + rank A

Elif Tan (Ankara University) Linear Transformations 22 / 28

Linear Transformations and Matrices

To …nd the rank of the linear transformation L : Rn ! Rm , it is

enough to check the rank of the matrix A. The rank of an m n
matrix A is the number of nonzero rows in the reduced row echelon
form of the matrix A.

dim Rn = dim KerL + dim L (Rn )

n = nullity A + rank A

If A is an n n matrix, then we have

1
rank A = n , nullity A = 0 , det A 6= 0 , A exists.

Elif Tan (Ankara University) Linear Transformations 22 / 28

Linear Transformations and Matrices

Example
1 2 3
Find the rank and the nullity of the matrix A = .
2 1 3

Solution: If we transform the matrix A to the reduced row echelon form,

we have
1 2 3 1 0 3
A=
2 1 3 0 1 3

rankA = number of nonzero rows of the matrix A in the reduced ref.

= 2

nullityA = n rankA = 3 2 = 1.

Elif Tan (Ankara University) Linear Transformations 23 / 28

Linear Transformations and Matrices

Theorem
Let L : V ! W be a linear transformation and consider the ordered basis
0
S = fv1 , v2 , ..., vn g and S = fv10 , v20 , ..., vn0 g for the vector space V , and
T = fw1 , w2 , ..., wm g and T 0 = fw10 , w20 , ..., wm0 g for the vector space W .
Let the transition matrix from basis S 0 to S be P, and the transition
matrix from basis T 0 to T be Q. If A is the matrix representation for the
linear transformation L with respect to the basis S and T , then Q 1 AP is
the matrix representation for the linear transformation L with respect to
the basis S 0 and T 0 .

Elif Tan (Ankara University) Linear Transformations 24 / 28

Linear Transformations and Matrices

De…nition
Let A and B are n n matrices, if there exist nonsingular matrix P such
that B = P 1 AP, then it is called B is similar to A.

Theorem
If A and B are similar n n matrices, then rankA = rankB.

Elif Tan (Ankara University) Linear Transformations 25 / 28

Remark

For n n matrix A, the followings are equivalent:

1 A is nonsingular, that is, A 1 exists.
2 A is row equivalent to In .
3 The linear system Ax = b has a unique solution.
4 The homogenous linear system Ax = 0 has only zero (trivial) solution.
5 A is a product of elementary matrices.
6 det (A) 6= 0.
7 The rank of A is n.
8 The nullity of A is zero.(Then the corresponding linear transformation
is 1 1)
9 The columns of A form a linearly independent set of vectors in Rn .

Elif Tan (Ankara University) Linear Transformations 26 / 28

Exercises
1 1 0
1. Let A = . Then L : R3 ! R2 , L (x ) = Ax is a linear
0 1 1
transformation.
Since
1 1 0 1 0 1
A= ,
0 1 1 0 1 1
The columns of A are not linearly independent. So L is not 1 1.
2 3
1 0
2. Let A = 4 0 1 5 . Then L : R2 ! R3 , L (x ) = Ax is a linear
1 0
transformation.
Since 2 3 2 3
1 0 1 0
A=4 0 1 5 4 0 1 5,
1 0 0 0
The columns of A are linearly independent. So L is 1 1.
Elif Tan (Ankara University) Linear Transformations 27 / 28
Exercises

2 3
1 0 0
3. Let A = 4 0 1 0 5 . Then L : R3 ! R3 , L (x ) = Ax is a linear
0 0 0
transformation.
Since A is in reduced row echelon form, it can easily be seen that The
columns of A are linearly dependent. So L is not 1 1.

Elif Tan (Ankara University) Linear Transformations 28 / 28