Chapter Five

Linear Transformation

Definition and Examples

Definition: Linear Transformation

A transformation T : Rm → Rn is called a linear transformation if, for every scalar c and
every pair of vectors u and v in Rm
1. T(𝑢 + 𝑣) = T(u) + T(v) (additivity) and
2. T(cu) = c T(u) (homogeneity).
Note that the above statement describes how a transformation T interacts with the two
operations of vectors, addition and scalar multiplication. It tells us that if we take two
vectors in the domain and add them in the domain, then transform the result; we will get the
same thing as if we transform the vectors individually first, then add the results in the
codomain. We will also get the same thing if we multiply a vector by a scalar and then
transform as we will if we transform first, then multiply by the scalar.
Example: Define T(x) = mx, where m is fixed real number. Show that T is linear
transformation.
Solution: We must show that T is additive and homogenous. For the additively, we let u and
v be in V.
i. T(u+v) = m(u+v) = mu +mv
T(u)+T(v)= mu+mv
⇒T(u+v) = T(u)+T(v)
For homogenous property,
ii. T(cu) = m(cu)=c.(mu)= cT(u)
⇒ T is linear transformation.
Example: Define T(x) = mx+b, where m is fixed real number and b≠ 0. Show that T is linear
transformation.
Solution: First we check additivity,
T(x+y) = m(x+y)+b = mx+my+b

1
and T(x)+T(y)= mx+b+my+b = mx+my+2b
⇒T(x+y)≠T(x)+T(y)
⇒ T is not linear transformation
b
Example: Let V be set of real valued functions and define T: V→ W by T(f) = ∫a f(x)dx.
Then T is linear transformation.
b b b
T(f+g)= ∫a (f(x) + g(x))dx = ∫a f(x)dx +∫a g(x)dx = T(f) +T(g)
and
b b
T(cf) = ∫a cf(x)dx = c∫a f(x)dx = c.T(f)

⇒ T is linear transformation
Example: Let V = Mnm and W = Mmn and define T(A) =At . Then show that T is linear
transformation.
Solution: For Additivity property
T(A + B) = (𝐴 + 𝐵)𝑡 = 𝐴𝑡 + 𝐵𝑡 = T(A)+T(B)
T(cA) = (𝑐𝐴)𝑡 = 𝑐. 𝐴𝑡 = cT(A)
⇒ T is linear transformation
Linear Transformations and Matrices

Definition: Matrix of a Linear Transformation

If T: 𝑅𝑚 → 𝑅𝑛 is a linear transformation, then there is a matrix A such that T(𝑥⃗ ) = A𝑥⃗ for
every 𝑥⃗ in 𝑅𝑚 . We will call A the matrix that represents the transformation
Example: Find the matrix [T] of the linear transformation T: 𝑅3 → 𝑅2 defined by

𝑥
𝑥+𝑦
T[𝑦] = [ 𝑦 − 𝑧 ]
𝑧

Solution: Based standard unit vectors

1
1+0 1
T(𝑒1 ) = T[0] = [ ]= [ ]
0−0 0
0

0
0+1 1
T(𝑒2 ) = T[1] = [ ]= [ ]
1−0 1
0

2
 0
0+0 0
T(𝑒3 ) = T[0] = [ ]= [ ]
0−1 −1
1
The matrix representation of T is
1 1 0
A = [T(𝑒1 ) T(𝑒2 ) T(𝑒3 )] = [ ]
0 1 −1

Finding the Matrix of a Linear Transformation

Let 𝑒1 , 𝑒2 , . . ., 𝑒𝑛 be the standard basis vectors of 𝑅𝑚 , and suppose that T : 𝑅𝑚 →𝑅𝑛 is a
linear transformation. Then the columns of [T] are the vectors obtained when T acts on each
of the standard basis vectors 𝑒1 , 𝑒2 , . . ., 𝑒𝑛
We indicate this by
[T] = [T(𝑒1 ) T(𝑒2 ) , … , T(𝑒𝑛 )]
Example: Each of the following linear transformation T. Give the matrix [T] of each.
𝑥 𝑥+𝑧
a. T[𝑦 ] = [ 𝑦 + 𝑧 ]
𝑧 𝑥+𝑦
Solution: To find the matrix, first we have to determine
1 1+0 1
T(𝑒1 )= T[0] = [0 + 0] = [0]
0 1+0 1
0 0+0 0
T(𝑒2 )= T[1 = 1 + 0
] [ ]= [ 1]
0 0+1 1
0 0+1 1
T(𝑒3 ) = T[0] = [0 + 1] = [1]
1 0+0 0
1 0 1
[T] = [T(𝑒1 ) T(𝑒2 ) T(𝑒3 )] = [0 1 1].
1 1 0
𝑎 3𝑎
b. T[ ] = [ ]
𝑏 𝑎−𝑏
Solution: To determine the matrix of T
1 3 3
T(𝑒1 ) =T[ ] = [ ]= [ ]
0 1−0 1
0 0 0
T(𝑒2 ) =T[ ] = [ ]= [ ]
1 0−1 −1
3 0
[T] = [T(𝑒1 ) T(𝑒2 )] = [ ]
1 −1

3
Rank and Nullity of linear transformation

Definition: Let T: V → W be a linear transformation between vector spaces over a field F.

i. The image of T is
T(V ) = imT = { T(v) | v ∈ V }.
ii. The kernel or null space of T is
ker T = { v ∈ V | T(v) = 0W }.
Proposition: Let T: V → W be a linear transformation between vector spaces V and W over
the field F. The image and kernel of T are subspaces of W and V, respectively.
Definition: Let T: V → W be a linear transformation between vector spaces over the field F.
I. The rank of T, which we shall denote rank T, is the dimension of the image of T.
II. The nullity of T, which we shall denote null T, is the dimension of the kernel of T.
Example: Find ker(T), where T: 𝑅3 → 𝑅2 is defined by T(x, y, z) =(x+y, y-z).
Solution: Since ker(T) = {v∈ 𝑉| 𝑇(𝑣) = 0}
We must solve T(x, y, z) = (0, 0)
⇒ (x+y, y-z) =(0, 0)
The resulting equations
𝑥+𝑦 =0
⇒{
𝑦−𝑧=0
⇒ x= -y and y = z
−𝑦 −1
which have solution ( 𝑦 ) = 𝑦 ( 1 )
𝑦 1
−1 −1
Therefore, KerT = {v ∈ 𝑅3 | 𝑣 = 𝑦 ( 1 )} = span{( 1 )}
1 1
The kernel of the given linear transformation was a subspace of the domain. In fact, a basis
−1
for kerT was{( 1 )}. The kernel of a linear transformation is always a vector space.
1
⇒dim(kerT) = 1.
Example: Calculate nullity for linear transformation T: 𝑅3 → 𝑅2 defined by
T((a, b, c)) = (a+2b+c, -a+3b +c)

4
Solution: We must find the set of all vectors (a, b, c) in 𝑅3 that T(a, b, c) =(0, 0)
The equation becomes
𝑎 + 2𝑏 + 𝑐 0
( )= ( )
−𝑎 + 3𝑏 + 𝑐 0
must be solved. The solution is
𝑎 −𝑘
(𝑏) = (−2𝑘 )
𝑐 −5𝑘
and kerT = span{(-1, -2, -5)}.
Therefore, dim(kerT) = 1, so nullity(T) = 1.
Example: Define T: 𝑅3 → 𝑅3 by T((x, y, c)) = (x-y+z, 2x+y-z, -x-2y+2z). Determine range T
and dim(range T).
Solution: Let v = (a, b, c) be in range of T. Then v = T(x, y,z).
Then the equation becomes
𝑎 𝑥−𝑦+𝑧
(𝑏 ) = ( 2𝑥 + 𝑦 − 𝑧 )
𝑐 −𝑥 − 2𝑦 + 2𝑧
and they reduce to
𝑥−𝑦+𝑧 =𝑎
{3𝑦 − 3𝑧 = 𝑏 − 2𝑎
0 = −𝑎 + 𝑏 + 𝑐
So if v = (a, b, c) is to be the range of T, then –a+b+c = 0. That is
Range T = {(a, b, c)| a = b+c}
The conditions on a, b and c gives a criterion for inclusion in range T. some vectors in
range T are (-2, -1, -1) and (0, -1, 1).
The dimension of range T is 2. So rank (T) =2.
Theorem (Rank-Nullity Theorem): Let V and W be vector spaces over the field F with V
finite-dimensional and let T: V → W be a linear transformation. Then
rank T + null T = dimV.
Example: Define T: 𝑅3 → 𝑅3 by T((x, y, c)) = (x-y+z, 2x+y-z, -x-2y+2z). Find null (T).
Solution: Since dim(V) = 3 and Rank(T) = 2.
Then by above theorem
dim(V) = rank T +null T

5
 ⇒3 = 2+null T
⇒null T = 1.
Algebra of Linear Transformation
Let V and W be vector spaces, and let L and T be linear transformation from V to W. We
can define the scalar multiple rL to L and the sum L+T of L and T as linear transformation
from V to W by the rules
a. (rL)(v) = r(L(v)), r a number, v𝜖 V
b. (L+T)(v) = L(v)+T(v), v𝜖 V
Example: Consider T: 𝑅2 → 𝑅2 defined by T((a, b)) = (a+2b, a-b) and L: 𝑅2 → 𝑅2 is defined
by L((x, y)) = (x, y). Then find T+L.
Solution: S+L is determine as
(T+L)((x, y)) = T(x, y)+L(x, y)
= (x+2y, x-y)+(x, y)
= (x+3y, 2x-y)
To find rT
(rT)((x, y))= r(T((x, y))) = r(x+2y, x-y)
= (rx+2ry, rx-ry)
The standard matrix for T+L is
1 3
( )
2 −1
and the standard matrix rT is
1 3
( )
2 −1
Definition: Composition of Transformations
Let S:𝑅𝑝 → 𝑅𝑛 and T: 𝑅𝑚 → 𝑅𝑝 be transformations. The composition of S and T, denoted by
S∘T, is the transformation S∘T: 𝑅𝑚 → 𝑅𝑛 defined by
⃗⃗⃗⃗))
(S o T) (𝑥⃗)= S(T(𝑥
for all vectors in 𝑅𝑚 .

Example: Suppose that S: 𝑅3 → 𝑅3 and T: 𝑅2 → 𝑅3 are linear transformations that can be

represented by the matrices.

6
 3 −1 5 2 7
[S] = [0 2 1 ] and [T] = [ −6 1 ] respectively.
4 0 −3 1 −4
𝑎
For the transformations S and T just defined, find (S ◦ T) (𝑥⃗) = (S◦T)[ ]. Then find the
𝑏
matrix of the transformation S ◦ T.

𝑎 𝑎 2 7 𝑎
Solution: (S◦T)[ ] = S(T[ ]) = S ([−6 1 ] [ ])
𝑏 𝑏 𝑏
1 −4
2𝑎 + 7𝑏
= S[−6𝑎 + 𝑏]
𝑎 − 4𝑏
3 −1 5 2𝑎 + 7𝑏
=[0 2 1 ] [−6𝑎 + 𝑏]
4 0 −3 𝑎 − 4𝑏
3(2𝑎 + 7𝑏) − (−6𝑎 + 𝑏) + 5(𝑎 − 4𝑏)
=[ 0(2𝑎 + 7𝑏) + 2(−6𝑎 + 𝑏) + (𝑎 − 4𝑏) ]
4(2𝑎 + 7𝑏) + 0(−6𝑎 + 𝑏) − 3(𝑎 − 4𝑏)
17𝑎 + 0𝑏
=[−11𝑎 − 2𝑏]
5𝑎 + 40𝑏
17 0
[S◦T] = [−11 −2].
5 40
Theorem Matrix of a Composition:
Let S:𝑅𝑝 → 𝑅𝑛 and T: 𝑅𝑚 → 𝑅𝑝 be linear transformations with matrices [S] and [T]. Then
[S ◦ T] = [S][T]

Eigenvalues and Eigenvectors

Definition: Eigenvalues and Eigenvectors
A scalar λ is called an eigenvalue of a matrix A if there is a nonzero vector 𝑥⃗ such that
A𝑥⃗ = 𝜆𝑥⃗
The vector 𝑥 is an eigenvector corresponding to the eigenvalue λ.
4 −3
Example: Determine whether either of ⃗⃗⃗⃗⃗
𝑤1 = [ ], and ⃗⃗⃗⃗⃗⃗
𝑤2 = [ ] are Eigen vector of A =
−1 3
−4 −6
[ ]. If either is, give the corresponding eigenvalue.
3 5

7
 −4 −6 4 −10
Solution: We see that A𝑤 1 [
⃗⃗⃗⃗⃗= ][ ] = [ ],
3 5 −1 7
𝑤1 is not Eigen vector of A because there is no scalar λ such that A𝑤
⇒ ⃗⃗⃗⃗⃗ ⃗⃗⃗⃗⃗1 = 𝜆1 𝑥⃗.
−4 −6 −3 −6
A𝑤 2 [
⃗⃗⃗⃗⃗⃗= ][ ] = [ ]
3 5 3 6
𝑤1 is Eigen vector of A, with corresponding Eigen value λ = 2 because A𝑤
⇒ ⃗⃗⃗⃗⃗ ⃗⃗⃗⃗⃗⃗2 = 2𝑥⃗.
Theorem: The number of eigenvalues of an n × n matrix is at most n.
Definition: Eigenspace Corresponding to an Eigenvalue
For a given eigenvalue 𝜆𝑗 of an n×n matrix A, the Eigen space 𝐸𝑗 corresponding to λ is the set
of all eigenvectors corresponding to𝜆𝑗 . It is a subspace of 𝑅𝑛 .
Or
The Eigen space Ej corresponding to 𝜆𝑗 is the set of all solutions to the equation

(A − 𝜆𝑗 I)𝑥⃗ =0.
−4 −6
Example: Find the Eigen space E1 of the matrix A =[ ] corresponding to the
3 5
eigenvalue 𝜆1 = 2.
−4 −6 1 0 −4 − 2 −6
Solution: For 𝜆1 = 2 we have A − 𝜆1 I = [ ] − 𝜆1 [ ]=[ ]=
3 5 0 1 3 5−2
−6 −6
[ ]
3 3
The set of all solutions to the equation

−6 −6 𝑥 0
(A − 𝜆𝑗 I)𝑥⃗ =0⇒ [ ] [𝑦 ] = [ ]
3 3 0
⇒-6x-6y = 0
⇒x = y
−1
The eigen space corresponding to 𝜆1 = 2 can then be described by either of 𝐸1 = {𝑦 [ ]} or
1
−1
𝐸1 = 𝑠𝑝𝑎𝑛 ([ ]).
1