Linear transformation

Definition Let V and W be two vector spaces with the same sets of scalars. Then a function
T: V  W is called a linear transformation of V into W if it satisfies the following properties:

i. T(u + v) =T(u) + T(v), for all u and v in V

ii. T(αv) = αT(v), for every v in V and every scalars 

[For the Euclidean vector spaces]

The function T: Rn  Rm is called linear transformation if

T(x1, x2,.. xn) = (u1, u2,.., um)

where u1 = a11x1 + a12x2 +….+ a1nxn

u2 = a21x1 + a22x2 +….+ a2nxn

… … …

um = am1x1 + am2x2 +….+ amnxn 1 i  m,1  j  n

Linear operator If V = W, then the linear transformation T: V→V is called a linear operator
on V.

Theorem Let T: V→W be a linear transformation. If x1, x2, x3,…, xn are vectors in V and

α1, α2, α3,…., αn are scalars, then

T(α1x1 + α2x2 + α3x3+….+ αnxn) = α1T(x1) + α2T(x2)+ α3T(x3)+….+ αnT(xn).

Corollary Let T: V→W be a linear transformation and {v1, v2, v3,…,vn} be a basis for V.

If v = α1v1 + α2v2 + α3v3+….+ αnvn is a vector of V, then

T(v) = α1T(v1)+ α2T(v2)+ α3T(v3)+….+ αnT(vn).

Corollary [Properties of linear transformation]

If T: V→W is a linear transformation, then

(i) T(0) = 0 (ii) T(−u) = −T(u)

(iii) T(u − v) = T(u) – T(v) for all u and v in V.

1
Algorithm to find matrix associated to linear transformation

Let V and W be vector spaces. Suppose the sets B ={e1, e2,..., en}and B′ ={w1, w2, …, wn}are
bases for the vector spaces V and W respectively and T is linear transformation from V to W.
Then the matrix associated with the linear transformation can be calculated using following
algorithm.

Step 1 Find the images of basis vectors of B under T, that is, find T(e1), T(e2), …, T(en)

Step 2 Find the coordinates of images with respect to basis B', that is, expressed the images as a
linear combinations of vectors of basis B'. In other words,

T(ek) = α1w1+ α2w2+ …+ αnwn

[T(ek)]B' = (α1, α2,…, αn)

Find α’s for each k, 1≤ k ≤n.

Step 3 Put them as columns of m×n matrix

A =  T e1 B T e2 B  T en B 

.

Addition and scalar multiplication of linear transformations Let S: V→W and T: V→W
be linear transformation. If α is any scalar, then addition S+T and scalar multiplication (αT) is
defined by the formula;
(S+T) (v) = S (v) + T (v), for v  V

(αT) (v) = α T(v), for all v  V and any scalar α.

Theorem The set L(V, W) of all Linear transformation from V into W is a vector space
with operations,

Addition : (T+S)(v) = T(v) + S(v), for S,T  L(V, W)

Scalar multiplication : (αT)(v) = αT(v), for all scalars α.

Composition operation Let V, U, W be vector spaces. If T1: V→U and T2: U→W are
linear transformations, then composition of T2 with T1 is the function T2oT1: V→W defined by
the equation

(T2oT1) (v) = T2(T1(v)), for all v in V.

Theorem If T1: V→U and T2: U→W are linear transformations, then T2oT1: V→W is
also a linear transformation.

2
One-to-one linear transformation If the linear transformation T: V→W maps distinct
vectors in V into distinct vectors in W, and then T is called one-to-one linear transformation.

Theorem Let T: V→W be a linear transformation and dim(V) = n. then the following
statement and equivalent.

i. T is one-to-one on V.
ii. If v₁, v₂,…,vn are independent elements in V, then T(v₁) ,T(v₂),…,T(vn) are
independent elements in T(V).
iii. If {e₁, e₂,…,en} is a basis for V, then {T(e₁),…..T(en)} is a basis for T(V).
iv. Dim(T(V)) = n.
Inverse of Linear Transformation Let T: V→W be a one-to-one linear transformation then the
inverse of T is a map T⁻¹: T(V) → V such that T⁻¹ (w) = v if T(v)= w for all v in V and w in
T(V).

Remark As we have seen in section 4.2, every Euclidean linear operator T: Rⁿ → Rⁿ can
be expressed by the formula T(x) = Ax where A is an n×n matrix. Therefore, the definition of
inverse of such linear transformation is reduced as follows

T⁻ ¹(x) = A−¹x if it exists.

Therefore we have following theorem for the Euclidean linear operator.

Nullity and Rank If T: V → W is a linear transform, then the dimension of kernel of T is

called nullity of T and the dimension of range of T is called the rank of T.

Notation: nullity of T = nullity(T); Rank of T = rank(T)

Theorem If T : Rn → Rm is a linear transformation defined by the formula, T(x) = Ax, A is

on m×n matrix, then

i. nullity(T) = nullity(A)
ii. rank(T) = rank(A)

Theorem [Rank – nullity theorem]

Let V be a finite dimensional vector space. If T: V → W be a linear transformation, then

rank (T) + nullity(T) = dim V.

3
Change of basis If B and B' are two bases for vector space V and P is a matrix that relates
[v]B and [v]B' by the equation [v]B = P[v]B` for every vϵV then P is called a transition matrix
from B' to B.
Moreover the transition matrix P can be expressed as

P = [I]B,B' = [ [u'1]B [u'2]B … [u'n]B ] where B= {u1, u2, …,un} and B̀'={u'1, u'2,…, u'n}
Theorem If V is a finite dimensional vector space with bases B and B' and P is transition
matrix from B` to B, then P is invertible and P-1 is the transition matrix from B to B', that is,

[v]B' = P–1 [v]B

Theorem If B and B' are the bases for the vector space V and T: V→V is a linear operator
on V, then [T]B` = P−1[T]BP where P is the transition matrix from B' to B.

Similar matrices Two square matrices A and B of order n are called similar matrices if there
is an invertible matrix C such that B =C-1AC.
Theorem Two matrices A and B are similar if and only if they represent the same linear
transformation.

Similarity invariant A property of square matrices is said to be a similarity invariant or

invariant under similarity if it is shared by any two similar matrices.