Unit II

LINEAR TRANSFORMATION

1
OUTLINE

• Linear Transformation

• Null Space

• Range Space

• Dimension Theorem

• Matrix Linear Transformation

• Eigen Values and Eigen Vectors of a Linear Transformation

• Diagonalization of Linear Transformation

• Application of Diagonalization of Linear system of differential equation.

2
LINEAR TRANSFORMATION
DEFINITION

Let 𝑉and 𝑊 be a vector space over a field 𝐹. Then a function 𝑇: 𝑉 → 𝑊 is said to be a linear

transformation from 𝑉 to 𝑊 if for all 𝑥, 𝑦 ∈ 𝑉 𝑎𝑛𝑑 𝑐 ∈ 𝐹

(a) 𝑇 𝑥 + 𝑦 = 𝑇 𝑥 + 𝑇(𝑦)

(b) 𝑇 𝑐𝑥 = 𝑐𝑇(𝑥)

NOTE

Let 𝑉 and 𝑊 be vector space over 𝐹 .

𝐼 : 𝑉 → 𝑉 𝑏𝑦 𝐼 𝑥 = 𝑥 ∀ 𝑥 ∈ 𝑉 is called Identity transformation and

T ∶ 𝑉 → 𝑊 𝑏𝑦 𝑇 𝑥 = 0 ∀ 𝑥 ∈ 𝑉 is called zero transformation.

3
PROPERTIES

1. If 𝑇 is a linear then 𝑇 0 = 0

2. 𝑇 is linear iff 𝑇 𝑐𝑥 + 𝑦 = 𝑐𝑇 𝑥 + 𝑇(𝑦)

3. 𝑇 is linear then 𝑇 𝑥 − 𝑦 = 𝑇 𝑥 − 𝑇(𝑦), for all 𝑥, 𝑦 ∈ 𝑉

4. 𝑇 is linear iff for 𝑥 , 𝑥 … . 𝑥 ∈ 𝑉 and𝑎 , 𝑎 … 𝑎 ∈ 𝐹 we have

𝑇Σ 𝑎𝑥 =Σ 𝑎 𝑇(𝑥 )
EXAMPLE 1 :
Define 𝑇 ∶ 𝑅 → 𝑅 by 𝑇 𝑎 , 𝑎 = 2𝑎 + 𝑎 , 𝑎 . Show that T is linear Transformation.

SOLUTION
Let 𝑐 ∈ 𝑅 and 𝑥 = (𝑏 , 𝑏 ) and 𝑦 = (𝑐 , 𝑐 ) be in 𝑅 , then
𝑥+𝑦 = 𝑏 ,𝑏 + 𝑐 ,𝑐 = 𝑏 + 𝑐 ,𝑏 + 𝑐
𝑇 𝑥+𝑦 = 𝑇 𝑏 + 𝑐 ,𝑏 + 𝑐
= 2 𝑏 +𝑐 + 𝑏 + 𝑐 ,𝑏 + 𝑐
= 2𝑏 + 2𝑐 + 𝑏 + 𝑐 ,𝑏 + 𝑐
= (2𝑏 + 𝑏 + 2𝑐 + 𝑐 ,𝑏 + 𝑐 ) (1)
𝑇 𝑥 = 𝑇 𝑏 ,𝑏 = 2𝑏 + 𝑏 , 𝑏
𝑇 𝑦 = 𝑇 𝑐 ,𝑐 = 2𝑐 + 𝑐 , 𝑐
𝑇 𝑥 + 𝑇 𝑦 = 2𝑏 + 𝑏 + 2𝑐 + 𝑐 ,𝑏 + 𝑐
= (2𝑏 + 𝑏 + 2𝑐 + 𝑐 , 𝑏 + 𝑐 ) (2)
From (1) and (2) 𝑇 𝑥+𝑦 = 𝑇 𝑥 + 𝑇 𝑦
5
𝑇 𝑐𝑥 = 𝑇 𝑐 𝑏 , 𝑏
= 𝑇 ( 𝑐 𝑏 ,𝑐 𝑏 )
= 2 𝑐𝑏 + 𝑐𝑏 , 𝑐𝑏 (3)
𝑐𝑇 𝑥 = 𝑐 𝑇 𝑏 ,𝑏
= 𝑐 2𝑏 + 𝑏 , 𝑏
= 2𝑐𝑏 + 𝑐𝑏 , 𝑐𝑏 (4)
From (3) and (4)
𝑇(𝑐𝑥) = 𝑐𝑇(𝑥)
Hence T is linear.

6
EXAMPLE 2 :

Let 𝑇: 𝑃 (𝑅) → 𝑃 ( 𝑅 ) defined by 𝑇(𝑓(𝑥)) = 𝑓′(𝑥) . Prove that T is linear.

SOLUTION

Given that 𝑇: 𝑃 (𝑅) → 𝑃 ( 𝑅 )defined by 𝑇(𝑓(𝑥)) = 𝑓′(𝑥)

To prove: T is linear transformation.

Let 𝑐 ∈ 𝑅 and let 𝑓 𝑥 , 𝑔 𝑥 ∈ 𝑃 (𝑅) . Then

a) 𝑇 𝑓 + 𝑔 𝑥 = 𝑓 + 𝑔  𝑥 = 𝑓 𝑥 + 𝑔 𝑥

=𝑇 𝑓 𝑥 +𝑇 𝑔 𝑥

b) 𝑇 𝑐𝑓 𝑥 = 𝑐𝑓 𝑥 = 𝑐𝑇 𝑓 𝑥

 𝑇 𝑐𝑓 𝑥 = 𝑐𝑇 𝑓 𝑥

 T is a linear transformation
7
EXAMPLE 3

Define 𝑇 ∶ 𝑀 × 𝐹 →𝑀 × ( 𝐹) 𝑏𝑦 𝑇(𝐴) = 𝐴 where 𝐴 is the transpose of A . Then prove

that T is a linear transformation.

SOLUTION

Given 𝑇 ∶ 𝑀 × 𝐹 →𝑀 × (𝐹) 𝑏𝑦 𝑇(𝐴) = 𝐴 . Let A, 𝐵 ∈ 𝑀 × 𝐹 𝑎𝑛𝑑 𝑐 ∈ 𝐹 , then

𝑎 𝑎 𝑏 𝑏
A= 𝑎 𝑎 B= 𝑏 𝑏
𝑎 𝑎 𝑏 𝑏

𝑎 𝑎 𝑏 𝑏 𝑎 +𝑏 𝑎 +𝑏
A+ B = 𝑎 𝑎 + 𝑏 𝑏 = 𝑎 +𝑏 𝑎 +𝑏
𝑎 𝑎 𝑏 𝑏 𝑎 +𝑏 𝑎 +𝑏

8
𝑎 +𝑏 𝑎 +𝑏
A+B = 𝑎 +𝑏 𝑎 +𝑏
𝑎 +𝑏 𝑎 +𝑏

𝑎 +𝑏 𝑎 +𝑏 𝑎 +𝑏
𝑇(𝐴 + 𝐵) = (1)
𝑎 +𝑏 𝑎 +𝑏 𝑎 +𝑏

𝑎 𝑎
𝑎 𝑎 𝑎
T A = 𝑎 𝑎 = 𝑎 𝑎 𝑎
𝑎 𝑎

𝑏 𝑏
𝑏 𝑏 𝑏
T B = 𝑏 𝑏 =
𝑏 𝑏 𝑏
𝑏 𝑏

𝑎 𝑎 𝑎 𝑏 𝑏 𝑏
𝑇 𝐴 + 𝑇(𝐵) = 𝑎 𝑎 𝑎 +
𝑏 𝑏 𝑏

9
𝑎 +𝑏 𝑎 +𝑏 𝑎 +𝑏
= (2)
𝑎 +𝑏 𝑎 +𝑏 𝑎 +𝑏

From (1) and (2), we have 𝑇 𝐴 + 𝐵 = 𝑇 𝐴 + 𝑇 𝐵

𝑎 𝑎
T(A) = 𝑎 𝑎
𝑎 𝑎

𝑐𝑎 𝑐𝑎
𝑐𝑎 𝑐𝑎 𝑐𝑎
T(cA) = 𝑐𝑎 𝑐𝑎 = 𝑐𝑎 𝑐𝑎 𝑐𝑎
𝑐𝑎 𝑐𝑎

𝑎 𝑎 𝑎
=𝑐 𝑎 𝑎 𝑎

= 𝑐𝑇 𝐴 .
Hence 𝑇 is linear.

10
EXAMPLE 4 :
Define 𝑇: 𝑅 → 𝑅 𝑏𝑦 𝑇(𝑎 , 𝑎 ) = (1, 𝑎 ) show that 𝑇 is not a linear transformation.
SOLUTION
𝐿𝑒𝑡 𝑐 ∈ 𝑅 𝑎𝑛𝑑 𝑥 = (𝑏 , 𝑏 ) 𝑎𝑛𝑑 𝑦 = (𝑐 , 𝑐 ) 𝑏𝑒 𝑖𝑛 𝑅 , 𝑡ℎ𝑒𝑛
𝑥+𝑦 = 𝑏 ,𝑏 + 𝑐 ,𝑐 = 𝑏 + 𝑐 ,𝑏 + 𝑐
𝑇(𝑥 + 𝑦) = 𝑇( (𝑏 + 𝑐 , 𝑏 + 𝑐 ))
= (1, 𝑏 + 𝑐 ) (1)
𝑇 𝑥 = 𝑇(𝑏 , 𝑏 ) = (1, 𝑏 )
𝑇 𝑦 = 𝑇 𝑐 ,𝑐 = 1, 𝑐
𝑇(𝑥) + 𝑇(𝑦) = (1, 𝑏 ) + (1, 𝑐 )
= (2, 𝑏 + 𝑐 ) (2)
𝐹𝑟𝑜𝑚 1 𝑎𝑛𝑑 2 , 𝑇 (𝑥 + 𝑦) ≠ 𝑇(𝑥) + 𝑇(𝑦)
𝐻𝑒𝑛𝑐𝑒 𝑇 𝑖𝑠 𝑛𝑜𝑡 𝑙𝑖𝑛𝑒𝑎𝑟.

11
EXAMPLE 5 :

Define 𝑇 ∶ 𝑅 → 𝑅 𝑏𝑦 𝑇(𝑎 , 𝑎 ) = (sin 𝑎 , 0) show that T is not a linear transformation.

SOLUTION

Let 𝑐 ∈ 𝑅 and 𝑥 = (𝑏 , 𝑏 ) and 𝑦 = (𝑐 , 𝑐 ) be in R , then

𝑥+𝑦 = (𝑏 , 𝑏 ) + (𝑐 , 𝑐 ) = (𝑏 + 𝑐 , 𝑏 + 𝑐 )

𝑇(𝑥 + 𝑦) = 𝑇 𝑏 + 𝑐 , 𝑏 + 𝑐 = sin 𝑏 + 𝑐 , 0 (1)

𝑇 𝑥 = 𝑇 𝑏 ,𝑏 = (𝑠𝑖𝑛 𝑏 , 0)

𝑇 𝑦 = 𝑇 𝑐 ,𝑐 = (sin 𝑐 , 0)

𝑇(𝑥) + 𝑇(𝑦) = ( sin 𝑏 , 0 ) + (sin 𝑐 , 0 ) = (sin 𝑏 + sin 𝑐 , 0) (2)

From (1) and (2) , 𝑇 𝑥+𝑦 ≠ 𝑇 𝑥 + 𝑇 𝑦

Hence, it is not linear.

12
EXAMPLE 6 :

Suppose that 𝑇 ∶ 𝑅2 → 𝑅2 is linear 𝑇(1, 0) = (1, 4) and 𝑇(1, 1) = (2, 5), what is 𝑇(2, 3) =?

SOLUTION

Given 𝑇 ∶ 𝑅2 → 𝑅2 is linear.

Now 2,3 = 𝑎 1,0 + 𝑏 1,1 = 𝑎 + 𝑏, 𝑏

𝑎 = −1 , 𝑏 = 3

Then 𝑇 2,3 = 𝑇(−1 1,0 + 3(1,1))

= −1𝑇 1,0 + 3𝑇 1,1

= −1 1,4 + 3 2,5

T 2,3 = (5,11)

13
Problem 1

Define 𝑇: 𝑅 → 𝑅 𝑏𝑦 𝑇(𝑎 , 𝑎 ) = (𝑎 , 𝑎 ) show that T is not a linear transformation.

Problem 2

Define 𝑇: 𝑅 → 𝑅 𝑏𝑦 𝑇(𝑥, 𝑦) = (𝑥 + 2𝑦, 2𝑥 − 𝑦, 𝑥 + 5𝑦) show that T is linear.

Problem 3

Define 𝑇: 𝑅 → 𝑅 𝑏𝑦 𝑇(𝑎 , 𝑎 , 𝑎 ) = 3𝑎 − 2𝑎 + 𝑎 , −𝑎 + 5𝑎 + 4𝑎 check whether that T is

linear or not.

14
NULL SPACE, RANGE SPACE AND DIMENSION
THEOREM

15
DEFINITION

Let 𝑉 and 𝑊 be vector spaces, and let 𝑇 ∶ 𝑉 → 𝑊 be linear. We define the Null space (or kernel)

𝑁 𝑇 𝑜𝑓 𝑇 𝑡𝑜 𝑏𝑒 𝑡ℎ𝑒 𝑠𝑒𝑡 𝑜𝑓 𝑎𝑙𝑙 𝑣𝑒𝑐𝑡𝑜𝑟𝑠 𝑥 𝑖𝑛 𝑉 𝑠𝑢𝑐ℎ 𝑡ℎ𝑎𝑡 𝑇 𝑥 = 0

𝑖𝑒. , 𝑁 𝑇 = {𝑥 ∈ 𝑉 ∶ 𝑇(𝑥) = 0}

DEFINITION

Let 𝑉 and 𝑊 be vector spaces, and let T : V →W be linear. We define the Range Space

(or image) R(T) of 𝑇 be the subset of 𝑊 consisting of all images (under 𝑇) of vectors in 𝑉

𝑖𝑒. , 𝑅 𝑇 = { 𝑇(𝑥) ∶ 𝑥 ∈ 𝑉 }

16
NOTE
Let I : 𝑉 → 𝑉 and T : 𝑉 → 𝑊 be the identity and zero transformation on V.

Then 𝑁 𝐼 = 𝑥 ∈ 𝑉 ∶ 𝐼 𝑥 = 0 = {0}

𝑅 𝐼 = { 𝐼(𝑥) ∶ 𝑥 ∈ 𝑉 } = 𝑉

𝑁 𝑇 = 𝑥 ∈ 𝑉∶ 𝑇 𝑥 = 0 = 𝑉

𝑅 𝑇 = 𝑇 𝑥 : 𝑥 ∈ 𝑉 = {0}

17
THEOREM 1 :

Let 𝑉 and 𝑊 be vector spaces, and 𝑇 ∶ 𝑉 → 𝑊 be linear. Then 𝑁(𝑇) and 𝑅(𝑇) are subspaces
𝑉 and 𝑊 respectively.

PROOF

Given that 𝑉 and 𝑊 be vector spaces, and 𝑇 ∶ 𝑉 → 𝑊 be linear.

To Prove: 𝑵(𝑻) is a subspace of 𝑽.

Since 𝑇 is a linear, 𝑇 𝑂 = 𝑂 , 𝑂𝑉 ∈ 𝑁 𝑇

where 𝑂𝑉 is zero vector of 𝑉 and 𝑂𝑊 is zero vector of 𝑊

𝐿𝑒𝑡 𝑥, 𝑦 ∈ 𝑁(𝑇), 𝑎𝑛𝑑 𝑐 ∈ 𝐹
⇒ 𝑇 𝑥 = 𝑂 ,
𝑇 𝑦 = 𝑂

18
S𝑖𝑛𝑐𝑒 𝑇 is linear , 𝑇(𝑥 + 𝑦) = 𝑇(𝑥) + 𝑇(𝑦)

= 𝑂 + 𝑂

= 𝑂
⇒ 𝑥 + 𝑦 ∈ 𝑁(𝑇)
𝐿𝑒𝑡 𝑥 ∈ 𝑁(𝑇), 𝑐∈ 𝐹
𝑇(𝑐𝑥) = 𝑐𝑇(𝑥) = 𝑐𝑂
=O
⇒ 𝑐𝑥 ∈ 𝑁 𝑇

∴ 𝑁(𝑇) is a subspace 𝑜𝑓 𝑉

19
To Prove: 𝑹(𝑻) is a subspace of 𝑾

𝑇(𝑂 ) = 𝑂𝑤 (since 𝑇 is linear)

⇒ 𝑂𝑤 ∈ 𝑅 𝑇

Let 𝑥, 𝑦 ∈ 𝑅 𝑇 , 𝑐 ∈ 𝐹 there exists 𝑣, 𝑤 in 𝑉 such that 𝑇(𝑣) = 𝑥 𝑇(𝑤) = 𝑦

𝑇(𝑣 + 𝑤) = 𝑇(𝑣) + 𝑇(𝑤) = 𝑥 + 𝑦

⇒ 𝑥+𝑦 ∈𝑅 𝑇

𝑇(𝑐𝑣) = 𝑐𝑇(𝑣) = 𝑐𝑥

⇒ 𝑐𝑥 ∈ 𝑅(𝑇)

is a subspace of 𝑊
20
THEOREM 2 :
Let 𝑉 and 𝑊 be vector spaces and let 𝑇: 𝑉 → 𝑊 be linear. If  = { 𝑣1, 𝑣2, … , 𝑣𝑛} is a
basis for 𝑉, then 𝑅 𝑇 = 𝑠𝑝𝑎𝑛 𝑇  = 𝑠𝑝𝑎𝑛 { 𝑇 𝑣1 , 𝑇 𝑣2 , … , 𝑇 𝑣𝑛 }

PROOF
Given 𝑇: 𝑉 → 𝑊 be linear
Then clearly 𝑇(𝑣 ) ∈ 𝑅(𝑇) for each 𝑖.
We know that, 𝑅(𝑇) is subspace of 𝑊.
By theorem
The span of any subset 𝑆 of a vector space 𝑉 is a subspace of 𝑉. Moreover , any subspace
of 𝑉 that contains 𝑆 must also contain the 𝑠𝑝𝑎𝑛 𝑜𝑓 𝑆 
𝑅 𝑇 𝑐𝑜𝑛𝑡𝑎𝑖𝑛𝑠 𝑠𝑝𝑎𝑛 𝑇 𝑣1 , 𝑇 𝑣2 , … , 𝑇 𝑣𝑛
∴ 𝑠𝑝𝑎𝑛 𝑇 𝑣1 , 𝑇 𝑣2 , … , 𝑇 𝑣𝑛 ⊆ 𝑅 𝑇
𝑠𝑝𝑎𝑛 𝑇 𝛽 ⊆𝑅 𝑇 (1)
21
To prove that: 𝑹(𝑻) ⊆ span (𝑻(𝜷))

Now, suppose that 𝑤 ∈ 𝑅(𝑇), then there exists 𝑣 ∈ 𝑉 such that 𝑇(𝑣) = 𝑤

Since  = { 𝑣1, 𝑣2, … , 𝑣𝑛} is a basis of 𝑉 then 𝑣 = 𝑎1𝑣1 + 𝑎2𝑣2 + ⋯ + 𝑎𝑛𝑣𝑛

𝑣= 𝑎𝑣 𝑓𝑜𝑟 𝑠𝑜𝑚𝑒 𝑎1, 𝑎2, … , 𝑎 ∈ 𝐹

⇒𝑇 𝑣 = 𝑇 ∑ 𝑎𝑣 ⇒ 𝑤 =∑ 𝑎 𝑇(𝑣 ) ( 𝑠𝑖𝑛𝑐𝑒 𝑇 𝑖𝑠 𝑙𝑖𝑛𝑒𝑎𝑟)

= 𝑎1𝑇 𝑣1 + 𝑎2𝑇 𝑣2 + ⋯ + 𝑎𝑛𝑇(𝑣𝑛) ∈ 𝑠𝑝𝑎𝑛{𝑇(𝑣1), 𝑇(𝑣2), … , 𝑇(𝑣𝑛)}

Hence 𝑤 ∈ 𝑠𝑝𝑎𝑛 (𝑇(𝛽))

Hence 𝑅 𝑇 ⊆ 𝑠𝑝𝑎𝑛 𝑇 𝛽 (2)

From (1) and (2) , we have 𝑅(𝑇) = span (𝑇())

NOTE

If  is infinite, then 𝑅(𝑇) = span { 𝑇 𝑣 : 𝑣 ∈ } ie.,The above theorem is true for infinite basis set. 22
DEFINITION

Let 𝑉 and 𝑊 be vector spaces, and let 𝑇 ∶ 𝑉 → 𝑊 be linear. If 𝑁 𝑇 and 𝑅 𝑇

are finite dimensional, then we define the nullity of T denoted by nullity (T) and is defined
to be the dimension of N(T) and the 𝑟𝑎𝑛𝑘 𝑇 is denoted by 𝑟𝑎𝑛𝑘 𝑇 and is defined to be
dimension of 𝑅 𝑇 respectively.

𝑖𝑒. , 𝑛𝑢𝑙𝑙𝑖𝑡𝑦(𝑇) = 𝑑𝑖𝑚 𝑁 𝑇

𝑟𝑎𝑛𝑘 𝑇 = 𝑑𝑖𝑚(𝑅 𝑇 )
DIMENSION THEOREM

𝐿𝑒𝑡 𝑉 𝑎𝑛𝑑 𝑊 𝑏𝑒 𝑣𝑒𝑐𝑡𝑜𝑟 𝑠𝑝𝑎𝑐𝑒𝑠, 𝑎𝑛𝑑 𝑙𝑒𝑡 𝑇 ∶ 𝑉 → 𝑊 𝑏𝑒 𝑙𝑖𝑛𝑒𝑎𝑟. 𝐼𝑓 𝑉 𝑖𝑠 𝑓𝑖𝑛𝑖𝑡𝑒 𝑑𝑖𝑚𝑒𝑛𝑠𝑖𝑜𝑛𝑎𝑙, 𝑡ℎ𝑒𝑛
𝑛𝑢𝑙𝑙𝑖𝑡𝑦 (𝑇) + 𝑟𝑎𝑛𝑘(𝑇) = 𝑑𝑖𝑚(𝑉)

PROOF

Let 𝑑𝑖𝑚 (V) = n & 𝑑𝑖𝑚 𝑁 𝑇 = 𝑘 & 𝑣 , 𝑣 , … , 𝑣 𝑖𝑠 𝑎 𝑏𝑎𝑠𝑖𝑠 𝑓𝑜𝑟 𝑁(𝑇)

Then by theorem ,

“If W is a subspace of finite dimensional vector space 𝑉 𝑡hen any basis for 𝑊 can be extended to a basis for 𝑉”

Using this , we may extend 𝑣 , 𝑣 , … , 𝑣 is a basis  = 𝑣 , 𝑣 , … , 𝑣 for 𝑉

𝑖. 𝑒. ,  = 𝑣 , 𝑣 , … , 𝑣 , 𝑣 , …,𝑣 for 𝑉

Now we claim that, 𝑆 = 𝑇 𝑣 ,𝑇 𝑣 ,…,𝑇 𝑣 𝑖𝑠 𝑎 𝑏𝑎𝑠𝑖𝑠 𝑓𝑜𝑟 𝑅(𝑇)

First we prove that 𝑆 generates 𝑅 𝑇 .

24
We know that ,

𝐿𝑒𝑡 𝑉 𝑎𝑛𝑑 𝑊 𝑏𝑒 𝑣𝑒𝑐𝑡𝑜𝑟 𝑠𝑝𝑎𝑐𝑒𝑠, 𝑎𝑛𝑑 𝑙𝑒𝑡 𝑇 ∶ 𝑉 → 𝑊 𝑏𝑒 𝑙𝑖𝑛𝑒𝑎𝑟. If  = 𝑣 , 𝑣 , … … 𝑣 is a

basis for 𝑉 then

𝑅 𝑇 = 𝑠𝑝𝑎𝑛 𝑇  = 𝑠𝑝𝑎𝑛 𝑇 𝑣 ,𝑇 𝑣 ,…,𝑇 𝑣

𝑖𝑒. , 𝑅(𝑇) = 𝑠𝑝𝑎𝑛 (𝑇  )

Using above theorem, we have 𝑅 𝑇 = 𝑠𝑝𝑎𝑛 𝑇 𝑣 ,𝑇 𝑣 ,…,𝑇 𝑣 (1)

𝑆𝑖𝑛𝑐𝑒 𝑣 , 𝑣 , … , 𝑣 𝑏𝑒 𝑡ℎ𝑒 𝑏𝑎𝑠𝑖𝑠 𝑓𝑜𝑟 N(T),

 𝑇 𝑣 = 0 𝑓𝑜𝑟 𝑖 = 1 𝑡𝑜 𝑘

(1)  𝑅 𝑇 = 𝑆𝑝𝑎𝑛( 𝑇 𝑣 ,𝑇 𝑣 ,…,𝑇 𝑣 = 𝑠𝑝𝑎𝑛(𝑆)

𝐻𝑒𝑛𝑐𝑒 𝑆 𝑔𝑒𝑛𝑒𝑟𝑎𝑡𝑒𝑠 𝑅(𝑇)

25
Now , we have to prove S is linearly independent for the scalars 𝑏 ,𝑏 , … 𝑏 ∈ 𝐹,

𝑤𝑒 ℎ𝑎𝑣𝑒 𝑏 𝑇 𝑣 +𝑏 𝑇 𝑣 + ⋯ +𝑏 𝑇 𝑣 =0 (2)

i.e, ∑ 𝑏𝑇 𝑣 =0  𝑇(∑ 𝑏 𝑣 ) = 0 [ since T is linear]

𝑠𝑜, 𝑏 𝑣 ∈ 𝑁(𝑇)

Hence there exists 𝑐 , 𝑐 … … 𝑐 ∈ 𝐹 𝑠𝑢𝑐ℎ 𝑡ℎ𝑎𝑡 ∑ 𝑏 𝑣 = 𝑐 𝑣 +𝑐 𝑣 + ⋯ … . . 𝑐 𝑣

[𝑠𝑖𝑛𝑐𝑒 𝑣 , 𝑣 , … , 𝑣 𝑖𝑠 𝑎 𝑏𝑎𝑠𝑖𝑠 𝑓𝑜𝑟 𝑁 𝑇 ]

𝑖𝑒., ∑ 𝑏𝑣 =∑ 𝑐𝑣  ∑ 𝑏 𝑣 −∑ 𝑐𝑣 =0

 ∑ −𝑐 𝑣 + ∑ 𝑏𝑣 =0

26
Since  = 𝑣 , 𝑣 , … … 𝑣 , 𝑣 ,𝑣 is a basis for 𝑉, we have 𝑣 , 𝑣 , … … 𝑣 are linearly independent.
 – 𝑐 = 0 𝑓𝑜𝑟 𝑖 + 1 𝑡𝑜 𝑘 & 𝑏 = 0 𝑓𝑜𝑟 𝑖 = 𝑘 + 1 𝑡𝑜 𝑛

In particular , 𝑖. 𝑒. , 𝑏 =𝑏 = …=𝑏 =0

From (2), we have {𝑇 𝑣 ,𝑇 𝑣 , … , 𝑇 𝑣 } is linear independent.

i.e., S is linear independent

Hence 𝑆 = {𝑇 𝑣 ,𝑇 𝑣 , … , 𝑇 𝑣 } is a basis for 𝑅(𝑇)

 𝑑𝑖𝑚 𝑅 𝑇 = 𝑛 − 𝑘, 𝑖𝑒. , 𝑟𝑎𝑛𝑘 (𝑇) = 𝑛 − 𝑘

𝑟𝑎𝑛𝑘 (𝑇) = 𝑑𝑖𝑚(𝑣) – 𝑑𝑖𝑚(𝑁(𝑇))

𝑟𝑎𝑛𝑘 𝑇 = 𝑑𝑖𝑚(𝑣)– 𝑛𝑢𝑙𝑙𝑖𝑡𝑦 𝑇

 𝑟𝑎𝑛𝑘 (𝑇) + 𝑛𝑢𝑙𝑙𝑖𝑡𝑦 (𝑇) = 𝑑𝑖𝑚(𝑣)

Hence the theorem.

27
DEFINITION :

We say that a function 𝑓: 𝐴 → 𝐵 is One-to-One if each element of the range has a unique pre-image.

i.e.., 𝑓: 𝐴 → 𝐵 is 1-1 if 𝑓 𝑥 = 𝑓 𝑦  𝑥 = 𝑦.

A function is said to Onto if 𝑓 𝐴 = 𝐵,

So, 𝑓 is onto iff the range of 𝑓 equals the co-domain of 𝑓.

THEOREM 1.1

Let 𝑉 and 𝑊 be vector spaces and let 𝑇 ∶ 𝑉 → 𝑊 be linear. Then T is one-to -one iff 𝑁 𝑇 = 0 .

PROOF

Given 𝑇: 𝑉 → 𝑊 linear.

Assume T is 1-1 then we have to prove N(T) = {0}

28
Let 𝑥 ∈ 𝑁(𝑇). Then, T(x) = 0 = T(0) [ since T is linear]

T(x) = T(0)  𝑥 = 0 [ since T is 1-1]

Hence 𝑁(𝑇) = {0}

Conversely,

Assume 𝑁(𝑇) = {0}. Then we have to prove T is 1-1.

Suppose that 𝑇 𝑥 = 𝑇 𝑦  𝑇 𝑥 − 𝑇 𝑦  𝑇 𝑥 − 𝑦 = 0 𝑠𝑖𝑛𝑐𝑒 𝑇 is linear

 𝑥 − 𝑦 ∈ 𝑁(𝑇) = 0

x=𝑦

𝑇ℎ𝑒𝑟𝑒𝑓𝑜𝑟𝑒 T is 1-1

29
THEOREM 1.2

Let V and W be vector space of equal (finite) dimension and let T :V → 𝑊 be linear. Then the following are

equivalent : a)T is 1-1 b) T is onto c) rank (T) = dim (v)

PROOF

From the dimension theorem, we have nullity (T) +rank (T) = dim (V) ------------------ (1)

By above theorem, we have

T is 1-1 iff N(T) ={0}  nullity (T) =0

 𝑟𝑎𝑛𝑘 (𝑇) = 𝑑𝑖𝑚(𝑉) [ substitute 𝑛𝑢𝑙𝑙𝑖𝑡𝑦 (𝑇) = 0 in (1)]

 𝑑𝑖𝑚(𝑅(𝑇)) = 𝑑𝑖𝑚(𝑊)

 𝑅(𝑇) = 𝑊

 T is onto.
30
PROBLEM 1

Let 𝑇: 𝑃 𝑅 → 𝑃 𝑅 𝑏𝑒 𝑡ℎ𝑒 𝑙𝑖𝑛𝑒𝑎𝑟 𝑡𝑟𝑎𝑛𝑠𝑓𝑜𝑟𝑚𝑎𝑡𝑖𝑜𝑛 𝑑𝑒𝑓𝑖𝑛𝑒𝑑 𝑏𝑦 𝑇(𝑓 𝑥 ) = 2 𝑓  𝑥 + 3 ∫ 𝑓 𝑡 𝑑𝑡.

Check whether 𝑇 is 1 − 1 or not.

SOLUTION

Since  = {1, 𝑥, 𝑥 } is a basis for 𝑃 𝑅

We know that 𝑅 𝑇 = 𝑠𝑝𝑎𝑛 𝑇 𝛽

𝑖. 𝑒. , 𝑅 𝑇 = 𝑠𝑝𝑎𝑛 𝑇 1 ,𝑇 𝑥 ,𝑇 𝑥

If 𝑓 𝑥 = 1, 𝑓  𝑥 = 0 then

𝑇 1 = 2 0 + 3 ∫ 𝑑𝑡 = 3 𝑡 = 3𝑥

31
if 𝑓 𝑥 = 𝑥, 𝑓 𝑥 = 1, then 𝑇 𝑥 = 2 1 + 3 ∫ 𝑡𝑑𝑡 = 2 + 3 = 2+

if 𝑓 𝑥 = 𝑥 , 𝑓 𝑥 = 2𝑥, then 𝑇 𝑥 = 2 2𝑥 + 3 ∫ 𝑡 𝑑𝑡 = 4𝑥 + 3 = 4𝑥 + x

To check { 3𝑥, 2 + (3𝑥 )/2, 4𝑥 + 𝑥 } is linear independent.

For scalars, , ,  ∈ 𝐹 𝑤𝑒 ℎ𝑎𝑣𝑒

3𝑥
 3𝑥 +  2 + +  ( 4𝑥 + 𝑥 ) = 0
2

3
  𝑥 +  𝑥 + 3 + 4𝛾 𝑥 + 2 = 0
2

i.e.,  = 0,  = 0  = 0

32
Hence { 3𝑥, 2 + , 4𝑥 + x } is linear independent

Tℎ𝑒𝑟𝑒𝑓𝑜𝑟𝑒 { 3𝑥, 2 + , 4𝑥 + x } is a basis for 𝑅(𝑇)

Which implies, rank 𝑇 = 3 & dim(𝑃 (𝑅)) = 4

rank (T) = dim (R(T)) = 3  4 = dim 𝑃 𝑅

𝑖𝑒. , 𝑟𝑎𝑛𝑘 𝑇  𝑑𝑖𝑚(𝑃 (𝑅))  T is not onto.

By dimension theorem, 𝑟𝑎𝑛𝑘 𝑇 + 𝑛𝑢𝑙𝑙𝑖𝑡𝑦 𝑇 = dim 𝑣 , ℎ𝑒𝑟𝑒 𝑉 = 𝑃 𝑅

 𝑛𝑢𝑙𝑙𝑖𝑡𝑦 𝑇 + 3 = 3  𝑛𝑢𝑙𝑙𝑖𝑡𝑦 𝑇 = 0  𝑁 𝑇 = 0

By theorem 1.1 , 𝐿𝑒𝑡 𝑉 𝑎𝑛𝑑 𝑊 𝑏𝑒 𝑣𝑒𝑐𝑡𝑜𝑟 𝑠𝑝𝑎𝑐𝑒𝑠, 𝑎𝑛𝑑 𝑙𝑒𝑡 𝑇 ∶ 𝑉 → 𝑊 𝑏𝑒 𝑙𝑖𝑛𝑒𝑎𝑟. Then T is one- to – one iff
𝑁 𝑇 = 0.

 T is 1-1.

33
PROBLEM 2
Let 𝑇: 𝐹 → 𝐹 be linear transformation defined by 𝑇(𝑎 , 𝑎 ) = (𝑎 + 𝑎 , 𝑎 ), then prove that 𝑇 is onto.

SOLUTION
Given 𝑇: 𝐹 → 𝐹 defined by 𝑇(𝑎 , 𝑎 ) = (𝑎 + 𝑎 , 𝑎 )
𝑇 𝑎 ,𝑎 = 0  (𝑎 +𝑎 , 𝑎 ) = 0  𝑎 = 0 𝑎𝑛𝑑 𝑎 =0
Then 𝑁(𝑇) = {0} i.e., nullity (T) = 0
By theorem 1.1, T is 1-1
By theorem 1.2, T is onto

34
EXAMPLE PROBLEMS

PROBLEM 1
Let 𝑃 (𝑅) → 𝑅 be the linear transformation defined by 𝑇(𝑎 + 𝑎 𝑥 + 𝑎 𝑥 ) = (𝑎 , 𝑎 , 𝑎 ), then
prove that 𝑇 is linear transformation and one-one.

PROBLEM 2
Suppose that T: R → 𝑅 is linear, T(1,0) = (1,4) & T(1,1) = (2,5). Find T(2,3).

PROBLEM 3
Let 𝑇: R → 𝑅 be the linear defined by 𝑇(𝑎 , 𝑎 , 𝑎 ) = (𝑎 − 𝑎 , 2𝑎 ), then prove that 𝑇 𝑖𝑠 𝑜𝑛𝑡𝑜 .

35
THEOREM 1.3

Let V and W be vector spaces over F and suppose that {𝑣 , 𝑣 , … … 𝑣 } is a basis for V . For 𝑤 , 𝑤 , … … 𝑤 ∈ W

there exist exactly one linear transformation T : V→W such that T(𝑣 )= 𝑤 for i=1,2, ……n.

PROOF

Let 𝑥 ∈ 𝑉, then 𝑥 = ∑ 𝑎 𝑣 where 𝑎 ∈ 𝐹, where 𝑎 , 𝑎 , … … 𝑎 are unique scalars.

Define 𝑇: 𝑉 → 𝑊 by 𝑇 𝑥 = ∑ 𝑎𝑤 (1)

Then, We have to prove T is linear and is unique.

To prove : T is linear

Suppose that 𝑥, 𝑦 ∈ 𝑉 & 𝑐 ∈ 𝐹.

Then 𝑥 = ∑ 𝑎𝑣 & y=∑ 𝑏 𝑣 for some scalars an 𝑎 , 𝑎 , … 𝑎 , 𝑏 , 𝑏 , … 𝑏

Thus c𝑥 + 𝑦 = 𝑐 ∑ 𝑎 𝑣 +∑ 𝑏𝑣 = ∑ 𝑐𝑎𝑣 +∑ 𝑏𝑣 36
c𝑥 + 𝑦 = ∑ (𝑐𝑎 +𝑏 )𝑣

Now, 𝑇(c𝑥 + 𝑦) = 𝑇(∑ (𝑐𝑎 +𝑏 )𝑣 ) = ∑ (𝑐𝑎 +𝑏 )𝑇(𝑣 ) = ∑ (𝑐𝑎 +𝑏 )𝑤

=∑ 𝑐𝑎 𝑤 + ∑ 𝑏𝑤

= 𝑐∑ 𝑎 𝑤 +∑ 𝑏𝑤 (by (1))
= 𝑐𝑇 𝑥 + 𝑇 𝑦

 T is linear Transformation

Clearly 𝑇 𝑣 =𝑤 𝑓𝑜𝑟 𝑖 = 1,2, … … 𝑛.

37
Now , we have to prove T is unique.

Suppose that 𝑈: 𝑉 → 𝑊 is a another linear trasformation and 𝑈 𝑣 = 𝑤 for i = 1, 2, … , n.

Then for 𝑥 ∈ V with 𝑥 = ∑ av

we have 𝑈(𝑥) = ∑ a U(v ) = ∑ a w = T(𝑥)

Hence 𝑈 = 𝑇

 T is unique linear transformation.

NOTE

Let 𝑉 and 𝑊 be vector spaces and suppose that V has a finite basis v , v , … , v .

If U, T ∶ V → W are linear & 𝑈(𝑣 ) = 𝑇(𝑣 ) for i = 1,2, … , n then U = T.

38
PROBLEM

Show that 𝑇 ∶ 𝑅 → 𝑅 defined by 𝑇 𝑎 , 𝑎 , 𝑎 = 𝑎 − 𝑎 , 2𝑎 is a linear transformation and find bases

𝑁 𝑇 & 𝑅 𝑇 . Compute the nullity & rank of T & verify the dimension theorem, Also determine whether T is

1-1 or onto.

SOLUTION

Let 𝑥 = 𝑎 , 𝑎 , 𝑎 𝑦 = 𝑏 ,𝑏 ,𝑏 ∈𝑅

𝑇 𝑥 = 𝑎 − 𝑎 , 2𝑎 , 𝑇(𝑦) = 𝑏 − 𝑏 , 2𝑏

𝑇(𝑐𝑥 + 𝑦) = 𝑇 𝑐𝑎 + 𝑏 , 𝑐𝑎 + 𝑏 , 𝑐𝑎 + 𝑏 = 𝑐𝑎 + 𝑏 − (𝑐𝑎 + 𝑏 ), 2(𝑐𝑎 + 𝑏 )

= 𝑐(𝑎 −𝑎 ) + (𝑏 −𝑏 ), 2𝑐𝑎 + 2𝑏

= 𝑐(𝑎 −𝑎 ) , 2𝑐𝑎 + (𝑏 − 𝑏 , 2𝑏 )

= 𝑐 (𝑎 −𝑎 ) , 2𝑎 ) + ((𝑏 −𝑏 ), 2𝑏

= 𝑐𝑇 𝑥 + 𝑇(𝑦)

 T is linear
39
Basis for N(T),

let 𝑎 , 𝑎 , 𝑎 ∈ 𝑁 𝑇 , then 𝑇 𝑎 , 𝑎 , 𝑎 =0

ie., 𝑎 − 𝑎 , 2𝑎 = 0,0  𝑎 = 𝑎 ,𝑎 = 0

Tℎ𝑒𝑟𝑒𝑓𝑜𝑟𝑒, 𝑁 𝑇 = 𝑎, 𝑎, 0 : 𝑎 ∈ 𝑅 = 𝑎 1,1,0 : 𝑎 ∈ 𝑅

𝐵𝑎𝑠𝑖𝑠 𝑓𝑜𝑟 𝑁 𝑇 𝑖𝑠 1,1,0 𝑎𝑛𝑑 dim 𝑁 𝑇 = 1 = 𝑛𝑢𝑙𝑙𝑖𝑡𝑦(𝑇)

Basis for 𝑹 𝑻

We know that,𝑒 = 1,0,0 , 𝑒 = 0,1,0 , 𝑒 = 0,0,1 𝑖𝑠 𝑎 𝑏𝑎𝑠𝑖𝑠 𝑓𝑜𝑟 𝑅 and 𝑅 𝑇 = 𝑠𝑝𝑎𝑛 𝑇 𝑒 , 𝑇 𝑒 , 𝑇 𝑒

Now 𝑇 𝑒 = 𝑇 1,0,0 = 1 − 0, 0 = 1,0

𝑇 𝑒 = 𝑇 0,1,0 = 0 − 1,0 = −1,0

𝑇 𝑒 = 𝑇 0,0,1 = 0 − 0,2 = 0,2

𝑅 𝑇 = 𝑠𝑝𝑎𝑛 1,0 , −1,0 , 0,2

40
1 0 1 0
𝑇ℎ𝑒 𝑚𝑎𝑡𝑟𝑖𝑥 𝑜𝑓 𝑖𝑚𝑎𝑔𝑒𝑠 𝑜𝑓 𝑒 , 𝑒 , 𝑒 𝑖𝑠 A = −1 0 = 0 0 𝑅 →𝑅 +𝑅
0 2 0 2
The non – zero rows 1,0 0,2 constitute a basis for 𝑅 𝑇

Tℎ𝑒𝑟𝑒𝑓𝑜𝑟𝑒, Basis for 𝑅 𝑇 is 1,0 , 0,2

𝑑𝑖𝑚 𝑅 𝑇 = 2 = 𝑟𝑎𝑛𝑘 𝑇

𝑅 𝑇 =𝑅

𝐵𝑦 𝑑𝑖𝑚𝑒𝑛𝑠𝑖𝑜𝑛 𝑡ℎ𝑚, 𝑟𝑎𝑛𝑘 𝑇 + 𝑛𝑢𝑙𝑙𝑖𝑡𝑦 𝑇 = dim V

rank (T) + nullity (T) = 2+1 = 3 = dim 𝑅

 Dimension theorem is verified.

Here, 𝑁 𝑇 ≠ 0  𝑇 𝑖𝑠 𝑛𝑜𝑡 1 − 1

Also, dim (𝑅 𝑇 ) = 2 = 𝑑𝑖𝑚 𝑅  T is onto.

41
PROBLEM 1

Prove that T is a linear transformation and find bases 𝑁 𝑇 & 𝑅 𝑇 . Compute the nullity & rank of T and
verify the dimension theorem, Also determine whether T is 1-1 or onto for the following.

𝑇: 𝑃 (𝑅) → 𝑃 (𝑅) defined by 𝑇 𝑓 𝑥 = 𝑥𝑓 𝑥 + 𝑓  (𝑥)

PROBLEM 2

If 𝑇: 𝑅 → 𝑅 is a linear Transformation defined by

𝑇 𝑥 ,𝑥 ,𝑥 ,𝑥 = 𝑥 − 𝑥 + 𝑥 + 𝑥 , 𝑥 + 2𝑥 − 𝑥 , 𝑥 + 𝑥 + 3𝑥 − 3𝑥 𝑓𝑜𝑟 𝑡ℎ𝑒 𝑥 , 𝑥 , 𝑥 , 𝑥 ∈ 𝑅 then
verify that dimension theorem.

42
DEFINITION

Let V be a finite-dimensional vector space. An ordered basis for V is a basis for V with a specific order.

That is, an ordered basis for V is a finite sequence of linearly independent vectors in V that generatres V .

𝐸. 𝑔

1)  = {𝑒 , 𝑒 , … 𝑒 } is the standard ordered basis for vector space 𝐹 .

2)  = {1, 𝑥, … . 𝑥 }is the standard ordered basis for vector space 𝑃 (𝐹).

3) In 𝑅 ,  = {𝑒 , 𝑒 , 𝑒 } can be considered an ordered basis . Also ,  ={𝑒 , 𝑒 , 𝑒 } is an ordered basis but

   as ordered bases.

43
DEFINITION

Let  = 𝑢 , 𝑢 , … . . 𝑢 be an ordered basis for a finite - dimensional vector space 𝑉.

For 𝑥 ∈ 𝑉, let 𝑎 , 𝑎 , … … 𝑎 be the unique scalars such that 𝑥 = ∑ 𝑎 𝑢 . Then the co-ordinate vector of 𝑥

𝑎
𝑎
relative to , denoted 𝑥  = .
.
𝑎

E.g.,

Let 𝑉 = 𝑃 𝑅 and let 𝛽 = 1, 𝑥, 𝑥 be the standard ordered basis for 𝑉,

4
𝐼𝑓 𝑓 𝑥 = 4 + 6𝑥 − 7𝑥 then 𝑓= 6
−7

44
MATRIX REPRESENTATION OF A LINEAR TRANSFORMATION
𝐃𝐄𝐅𝐈𝐍𝐈𝐓𝐈𝐎𝐍

Let 𝑉 and 𝑊 are finite − dimensional vector space over 𝐹, with ordered bases

𝛽 = 𝑣 ,𝑣 ,….𝑣 and  = 𝑤 , 𝑤 , … . 𝑤 respectively. let 𝑇: 𝑉 → 𝑊 be linear transformation.

Then for each 𝑗, 1  𝑗 𝑛, 𝑡here exist unique scalars 𝑎 ∈ 𝐹, 1  𝑖 𝑚 such that

𝑇 𝑣 = ∑ 𝑎 𝑤 𝑓𝑜𝑟 1  𝑗 𝑛

ie., 𝑇 𝑣 =∑ 𝑎 𝑤 = 𝑎 𝑤 + 𝑎 𝑤 + ⋯+ 𝑎 𝑤

𝑇 𝑣 =∑ 𝑎 𝑤 = 𝑎 𝑤 + 𝑎 𝑤 + ⋯+ 𝑎 𝑤

𝑇 𝑣 =∑ 𝑎 𝑤 = 𝑎 𝑤 + 𝑎 𝑤 + ⋯+ 𝑎 𝑤
45
The transpose of the above matrix of co-efficient, denoted by 𝑇 is called the matrix representation of 𝑇 is the

ordered basis  and  .

𝑎 ⋯ 𝑎
ie., 𝑇 = ⋮ ⋱ ⋮
𝑎 ⋯ 𝑎

𝑵𝒐𝒕𝒆: 𝐼𝑓 𝛽 = 𝛾, then 𝑇 = 𝑇

46
PROBLEM:

1 2 1
Find the linear transformation 𝑇: 𝑉 𝑅 → 𝑉 𝑅 determined by the matrix 0 1 1
−1 3 4
with respect to standard basis 𝑒 , 𝑒 , 𝑒 .

SOLUTION:
𝑇𝑒 = 𝑒 + 0𝑒 − 𝑒 = 1,0, −1
𝑇𝑒 = 2𝑒 + 𝑒 + 3𝑒 = 2,1,3

𝑇𝑒 = 𝑒 + 𝑒 + 4𝑒 = 1,1,4

𝑎, 𝑏, 𝑐 = 𝑎 1,0,0 + 𝑏 0,1,0 + 𝑐 0,0,1

= 𝑎𝑒 + 𝑏𝑒 + 𝑐𝑒
𝑇 𝑎, 𝑏, 𝑐 = 𝑎𝑇 𝑒 + 𝑏𝑇 𝑒 + 𝑐𝑇 𝑒
= 𝑎 1,0, −1 + 𝑏 2,1,3 + 𝑐 1,1,4

= (𝑎 + 2𝑏 + 𝑐, 𝑏 + 𝑐, −𝑎 + 3𝑏 + 4𝑐) 47
PROBLEM:

Let 𝛽 and 𝛾 be the standard bases for 𝑅 and 𝑅 respectively. Then for each linear transformation

𝑇: 𝑅 → 𝑅 , compute 𝑇 .

1)𝑇: 𝑅 → 𝑅 defined by 𝑇 𝑎 , 𝑎 = (2𝑎 − 𝑎 , 3𝑎 + 4𝑎 , 𝑎 )

2) 𝑇: 𝑅 → 𝑅 defined by 𝑇 𝑎 , 𝑎 … 𝑎 = (𝑎 , 𝑎 ,…𝑎 )

𝑓(0) 2𝑓(1)
3)𝑇: 𝑃 𝑅 → 𝑀 × (𝑅) defined by 𝑇 𝑓 𝑥 =
0 𝑓(3)

SOLUTION:

1) Here  = 1,0 , 0,1 𝑎𝑛𝑑  = 1,0,0 , 0,1,0 , 0,0,1

Given 𝑇 𝑎 , 𝑎 = (2𝑎 − 𝑎 , 3𝑎 + 4𝑎 , 𝑎 )
Then 𝑇 𝑒 = 𝑇 1,0 = 2,3,1
𝑇 𝑒 = 𝑇 0,1 = −1,4,0 .
2 −1
Therefore 𝑇 =3 4
1 0
48
2) 𝑇: 𝑅 → 𝑅 defined by 𝑇 𝑎 , 𝑎 … 𝑎 = 𝑎 ,𝑎 ,…𝑎

Here 𝛽 = 𝛾 = {𝑒 , 𝑒 , … . 𝑒 }

Then 𝑇 𝑒 = 𝑇 1,0,0 … .0 = 0,0, … 1

𝑇 𝑒 = 𝑇 0,1,0 … .0 = 0,0, … 1,0

𝑇 𝑒 = 𝑇 0,0,0 … .1 = 1,0, … 0,0

0 ⋯ 1
𝑇= ⋮ ⋱ ⋮
1 ⋯ 0

49
𝑓(0) 2𝑓(1)
3) 𝑇: 𝑃 𝑅 → 𝑀 × (𝑅) defined by 𝑇 𝑓 𝑥 =
0 𝑓(3)

1 0 0 1 0 0 0 0
Given 𝛽 = 1, 𝑥, 𝑥 and 𝛼 = , , ,
0 0 0 0 1 0 0 1
Then
0 2 1 0 0 1 0 0 0 0
𝑇 1 = =0 +2 +0 +0
0 0 0 0 0 0 1 0 0 1

1 2 1 0 0 1 0 0 0 0
𝑇 𝑥 = =1 +2 +0 +0
0 0 0 0 0 0 1 0 0 1

0 2 1 0 0 1 0 0 0 0
𝑇 𝑥 = =0 +2 +0 +2
0 2 0 0 0 0 1 0 0 1

Hence the required matrix is

0 1 0
2 2 2
𝑇 =
0 0 0
0 0 2
50
PROBLEM

Let 𝑇: 𝑅 → 𝑅 be defined by T a , a = a − a , a , 2a + a . Let 𝛽 be the standard ordered basis

for 𝑅 , given that 𝛾 = { 1,1,0 , 0,1,1 , 2,2,3 } then compute 𝑇 If ={(1,2),(2,3)} then compute 𝑇 .

DEFINITION

𝑇 is invertible  𝑇 𝑖𝑠 1 − 1 and 𝑇 is onto

DEFINITION

Let 𝐴 be an 𝑚 × 𝑛 matrix with entries from a field 𝐹. Then the mapping 𝐿 : 𝐹 → 𝐹 defined by𝐿 𝑥 =

𝐴𝑥 for each column vector 𝑥 ∈ 𝐹 is called the left-multiplication transformation. It is denoted by𝐿

1 2 1
EXAMPLE Let 𝐴 = Then 𝐴 ∈ 𝑀 × (𝑅) and 𝐿 : 𝑅 → 𝑅
0 1 2

1 1
1 2 1 6
If 𝑥 = 3 then, 𝐿 = 𝐴𝑥 = 3 =
0 1 2 1 51
−1 −1
DIAGONALIZABLE - EIGEN VALUES , EIGEN VECTORS &

APPLICATION OF SYSTEM OF DIFFERENTIAL EQUATIONS

52
DEFINITION:

A linear operate 𝑇 on a finite − dimensional vector space 𝑉 is called diagonalizable if there is an ordered
basis  for 𝑉 such that 𝑇 is a diagonal matrix.

NOTE:

1) A square matrix A is said to be diagonalizable if A is similar to a diagonal matrix.

2) A matrix 𝐴 𝑖𝑠 called diagonalizable if 𝐿 is diagonalizable .

53
THEOREM:

A linear operator 𝑇 of the finite dimensional vector space 𝑉 is diagonalizable iff there exists an basis 𝛽 =
{𝑥 , 𝑥 … 𝑥 } for V and scalars 𝜆 , 𝜆 , … 𝜆 (not necessarily distinct) such that 𝑇 𝑥 = 𝜆 𝑥 , 𝑓𝑜𝑟 1𝑗𝑛. Under these
circumstances
𝜆 0 … 0
0 𝜆 … 0
𝑇 =
⋮ ⋮ ⋯ ⋮
0 0 ⋯ 𝜆
SKETCH OF PROOF:
Suppose that 𝑇 is diagonalizable. Then there is a basis 𝛽 for 𝑉 such that 𝑇  = 𝐷 is a diagonal matrix.

Let 𝜆 = 𝐷 and 𝛽 = 𝑥 , 𝑥 … . . 𝑥 . Then for each 𝑗,

𝑇 𝑥 = 𝐷 𝑥 =𝐷 𝑥 =𝜆𝑥

Conversely, suppose there exists a basis 𝛽 = {𝑥 , 𝑥 … . . 𝑥 } and scalars 𝜆 , 𝜆 … 𝜆 such that 𝑇 𝑥 = 𝜆 𝑥 . Then clearly
𝜆 0 … 0
0 𝜆 … 0
𝑇 =
⋮ ⋮ ⋯ ⋮ 54
0 0 ⋯ 𝜆
DEFINITION:
Let 𝑇 be an operator on a vector space 𝑉. A non-zero vector 𝑣 ∈ 𝑉 is called an Eigen vector of 𝑇 if there
exist a scalar 𝜆 sucℎ that 𝑇 𝑣 = 𝜆𝑣. The scalar 𝜆 is called Eigen value corresponding to the Eigen vector 𝑣.

Let 𝐴 be in 𝑀 × 𝐹 . A non − zero vector 𝑣 ∈ 𝐹 is called an Eigen vector of A if 𝑣 is an eigenvector of 𝐿 .

𝑖𝑒. , 𝐴𝑣 = 𝜆𝑣 for some scalar 𝜆.
The scalar 𝜆 is called the Eigen value of 𝐴 corresponding to the eigenvector 𝑣.

NOTE:
:
1) Eigen vector is also called characteristic vector or proper vector. Similarly Eigen value is also called as
characteristic value, proper value.

2) A linear operator 𝑇 on a finite-dimensional vector space 𝑉 is diagonalizable if and only if there exists a basis
𝛽 for 𝑉 consisting of Eigen vectors of 𝑇. Furthermore, if 𝑇 is diagonalizable, 𝛽 = {𝑥 , 𝑥 … 𝑥 } is a basis of
eigenvectors of 𝑇, and if 𝐷 = 𝑇 , 𝑡ℎ𝑒𝑛 𝐷 is a diagonal matrix and 𝐷 is the eigenvalues corresponding to 𝑥

55
DEFINITION:

Let 𝐴 ∈ 𝑀 × (𝐹) then the polynomial 𝑓 𝑡 = det(𝐴 − 𝑡𝐼 ) is called the characteristic polynomial of 𝐴.

THEOREM

Let 𝐴 ∈ 𝑀 × (𝐹), then a scalar 𝜆 is an eigen value 𝑜𝑓 𝐴 iff det 𝐴 − 𝜆𝐼 =0

PROOF:

A scalar 𝜆 is an eigenvalue of 𝐴 iff there exists a non-zero vector 𝑣 ∈ 𝐹 such that 𝐴𝑣 = 𝜆𝑣

 𝐴 − 𝜆𝐼 𝑣 = 0

 𝑣 ∈ 𝑁(𝐴 − 𝜆𝐼 )

 𝐴 − 𝜆𝐼 is not invertible

 𝐴 − 𝜆𝐼  = 0

NOTE: The above theorem states that the Eigen values of a matrix are the zeros of its characteristic polynomial.

56
EXAMPLE:

1 1
To find the eigenvalues of 𝐴 = ∈𝑀 × 𝑅
4 1

SOLUTION:

The characteristic polynomial is 𝑓 𝑡 =𝐴 − 𝑡𝐼 

1−𝑡 1
𝑓 𝑡 = = 𝑡 − 2𝑡 − 3 = 𝑡 − 3 𝑡 + 1
4 1−𝑡

By above theorem, the eigenvalues are the zero of f(t).

 The eigenvalue of 𝐴 are 3,-1

57
PROBLEM:
Let 𝑇 be the linear operator on 𝑃 (𝑅) defined by 𝑇 𝑓 𝑥 = 𝑓 𝑥 + 𝑥 + 1 𝑓(𝑥). Let 𝐵 be the standard
ordered basis for 𝑃 𝑅 . Compute characteristic polynomial and Eigen value of 𝑇.
SOLUTION
The standard basis of P R is 𝐵 = 1, 𝑥, 𝑥
𝑇 1 =1+ 1+1 0 =1
𝑇 𝑥 = 𝑥 + 𝑥 + 1 1 = 1 + 2𝑥
𝑇 𝑥 = 𝑥 + 𝑥 + 1 2𝑥 = 3𝑥 + 2𝑥

1 1 0
𝐴 = 𝑇 = 0 2 2
0 0 3

1−𝑡 1 0
The characteristic polynomial is 𝑓 𝑡 = 𝐴 − 𝑡𝐼 = 0 2−𝑡 2
0 0 3−𝑡
𝑓 𝑡 = −𝑡 + 6𝑡 − 11𝑡 + 6
Therefore, the eigenvalues of A are 1,2,3. 58
THEOREM

Let 𝐴 ∈ 𝑀 × 𝐹 . 𝑡ℎ𝑒𝑛

1. The characteristic polynomial of A is a polynomial of degree n with leading co-efficient −1

2. A has at most n distinct eigenvalues.

THEOREM

Let 𝑇 be a linear operator on a vector space 𝑉 and let  be an eigenvalue of 𝑇. Then a vector 𝑣 ∈ 𝑉is an
eigenvector of 𝑇 corresponding to  iff 𝑣 0 & 𝑣 ∈ 𝑁 (𝑇 − 𝐼)

PROOF :

Let T be a linear operator on a vector space 𝑉 & let  be an eigenvalue of 𝑇.

Let 𝑣 ∈ 𝑉 be an eigen vector of 𝑇 corresponding to .

59
Then we have to prove 𝑣 ∈ 𝑁 𝑇 − 𝐼

Since v is an eigenvector of T corresponding to . We have 𝑇𝑣 = 𝑣  𝑇 − 𝐼 𝑣 = 0

Thus, 𝑣 ∈ 𝑁 𝑇 − 𝐼

Conversely,

Assume 𝑣 ∈ 𝑁 𝑇 − 𝐼

Then we have to prove 𝑣 is an eigen vector of 𝑇.

if 𝑣 ∈ 𝑁 𝑇 − 𝐼 , 𝑡ℎ𝑒𝑛 𝑇 − 𝐼 𝑣 = 0
 𝑇𝑣 − 𝑣 = 0

 𝑇𝑣 = 𝑣

Thus, 𝑣 is an eigenvector of T corresponding to .

60
PROBLEM

1 1
Find that eigenvector of the matrix 𝐴 = , check whether it is diagonalizable or not.
4 1

SOLUTION

The characteristic polynomial is 𝑓 𝑡 = 𝐴 − 𝑡𝐼

1−𝑡 1
𝑓(𝑡) =
4 1−𝑡
= 1−𝑡 −4
= 1 + 𝑡 − 2𝑡 − 4
= 𝑡+1 𝑡−3

The Eigen value of 𝐴 are 3, -1

To find eigenvector

61
𝐴 − 𝜆𝐼 𝑥 = 0

1− 1 𝑥 0
 =
4 1− 𝑥 0

1−3 1 𝑥 0
When λ = 3  𝑥 =
4 1−3 0

−2 1 𝑥 0
𝑥 =
4 −2 0

 − 2𝑥 + 𝑥 = 0, 4𝑥 − 2𝑥 = 0

 2𝑥 = 𝑥

Let 𝑥 = 𝑘, then 𝑥 = 2𝑘
𝑥 𝑘 1
X = 𝑥 = =𝑘
2𝑘 2
1
is an eigenvetor corresponding 𝜆 = 3
2
1
The set of all solution to this equation are 𝑁 𝐴 − 3 𝐼 = {𝑘 , 𝑘 ∈ 𝑅}
2
62
When 𝜆 = −1

1+1 1 𝑥 0
𝑥 =
4 1+1 0

2 1 𝑥 0
 𝑥 =
4 2 0
 2𝑥 + 𝑥 = 0 , 4𝑥 + 2𝑥 = 0

 𝑥 = −2𝑥

Let 𝑥 = 𝑘 then 𝑥 = −2𝑘

𝑥 𝑘 1
X = 𝑥 = =𝑘
−2𝑘 −2

1
X = 𝑘
−2
1
is an eigenvector corresponding to 𝜆 = −1
−2

1 1
Therefore, 𝐵 = , is basis of 𝑅 and consisting of eingenvectors of A.
2 −2
63
 𝐴 is diagonalizable .
To find the diagonal matrix
1 1
𝑄 =
2 −2

1
𝑄 = 𝑎𝑑𝑗 A
𝑄

1 −2 −1
=
−4 −2 1

Similarity transformation :

−2 −1 1 1 1 1
𝑄 𝐴𝑄 =
−2 1 4 1 2 −2

−2 − 4 −2 − 1 1 1
=
−2 + 4 −2 + 1 2 −2

−6 −3 1 1
=
2 −1 2 −2

−6 − 6 −6 − 6
=
2−2 2+2
−12 0 −3 0
= =
0 4 0 1
64
THEOREM:

Let T be a linear operator on a vector space 𝑉 and let  ,  , … . .  be distinct eigen values of 𝑇.

If 𝑣 , 𝑣 , 𝑣 , … . 𝑣 are eigen vectors of 𝑇 such that  corresponds to 𝑣 ; 𝑖 = 1 𝑡𝑜 𝑘 then 𝑣 , 𝑣 , … . . 𝑣 is linear

independent.

PROOF:

The proof is by mathematical induction on k.

Suppose that on k =1, then 𝑣 ≠ 0, since 𝑣 is an Eigen vector corresponding to 

Hence 𝑣 is linear independent

Now assume that the theorem holds for k-1 distinct eigen value.

i.e., 𝑣 , 𝑣 , … , 𝑣 is linear independent

Now , we have to prove 𝑣 , 𝑣 , … , 𝑣 is linear independent.

65
Let 𝑎 𝑣 + 𝑎 𝑣 + … + 𝑎 𝑣 = 0, where 𝑎 ′𝑠 are scalars (1)

Applying 𝑇 −  𝐼 on both sides.

𝑇 −  𝐼 𝑎 𝑣 + 𝑎 𝑣 + ⋯+ 𝑎 𝑣 +𝑎 𝑣 = 𝑇− 𝐼 0

𝑇 −  𝐼 𝑎 𝑣 ) + 𝑇 −  𝐼 (𝑎 𝑣 ) + ⋯ + 𝑇 −  𝐼 𝑎 𝑣 + 𝑇 −  𝐼 (𝑎 𝑣 =0

𝑇(𝑎 𝑣 ) −  𝑎 𝑣 + 𝑇 𝑎 𝑣 −  𝑎 𝑣 + ⋯+ 𝑇 𝑎 𝑣 − 𝑎 𝑣 + 𝑇(𝑎 𝑣 ) −  𝑎 𝑣 = 0

𝑎 𝑇 𝑣 − 𝑎 𝑣 +𝑎 𝑇 𝑣 −  𝑎 𝑣 + ⋯+ 𝑎 𝑇 𝑣 −  𝑎 𝑣 +𝑎 𝑇 𝑣 − 𝑎 𝑣 =0

𝑎 𝜆 𝑣 −  𝑎 𝑣 + 𝑎 𝜆 𝑣 −  𝑎 𝑣 + ⋯+ 𝑎 𝜆 𝑣 −𝜆 𝑎 𝑣 + 𝑎  𝑣 − 𝑎  𝑣 = 0 (since 𝑇 𝑣 = 𝜆𝑣)

𝑎 𝑣 𝜆 −𝜆 +𝑎 𝑣 𝜆 −𝜆 + …+𝑎 𝑣 𝜆 −𝜆 =0

66
By the induction hypothesis 𝑣 , 𝑣 , … , 𝑣 is linear independent and hence

𝑎  − =𝑎  − =⋯=𝑎  − =0

Since  ,  , … ,  are distinct

𝑎 =𝑎 =⋯=𝑎 =0

From (1)
𝑎 𝑣 =0

𝑎 =0

{𝑣 , 𝑣 , … , 𝑣 , 𝑣 } is Linearly independent.

NOTE:

Let T be a linear operator on a n-dimensional vector space V if 𝑇 has n- distinct eigenvalues then T is
diagonalizable .

67
PROBLEM
𝑎 4𝑎 + 𝑎
Let T be the linear operator on 𝑅 defined by 𝑇 𝑎 = 2𝑎 + 3𝑎 + 2𝑎 .
𝑎 𝑎 + 4𝑎
Determine the eigen space of 𝑇 corresponding to each eigenvalue. Let 𝐵 be the standard ordered basis for 𝑅 & also
check whether T is diagonalizable or not.

SOLUTION:

Basis 𝐵 = 1,0,0 , 0,1,0 , 0,0,1

4 0 1
𝑇 = 2 3 2
1 0 4

4− 0 1
𝑇 − 𝐼 = 2 3− 2
1 0 4−
= 4 −  3 −  4 −  + 1(0 − 3 −  )

= −𝜆 + 11𝜆 − 39𝜆 + 45
68
The characteristic polynomial is 

The eigenvalues are

That is, the eigenvalues of are with multiplicity 1

with multiplicity 2

If

69
When 𝜆 = 5 (𝐴 − 𝜆 𝐼)X = 0

4 0 1 1 0 0 𝑥 0
 2 3 2 −5 0 1 0 𝑥 = 0
1 0 4 0 0 1 𝑥 0

−1 0 1 𝑥 0
 2 −2 2 𝑥 = 0
1 0 −1 𝑥 0
 −𝑥 + 𝑥 = 0

2𝑥 − 2𝑥 + 2𝑥 = 0

𝑥 −𝑥 = 0

𝑥 1
 𝑥 = 2
𝑥 1

1
Basis for 𝐸𝜆 = 2
1
dim Eλ =1 70
When
𝐴−𝜆 𝐼 𝑋 =0

4 0 1 1 0 0 𝑥 0
 2 3 2 −3 0 1 0 𝑥 = 0
1 0 4 0 0 1 𝑥 0

1 0 1 𝑥 0
 2 0 2 𝑥 = 0
1 0 0 𝑥 0

 𝑥 +𝑥 =0

2𝑥 + 2𝑥 = 0
𝑥 +𝑥 = 0

 𝑥 = −𝑥

All equation are same

Put 𝑥 = 𝑠, 𝑥 = 𝑡, then
𝑥 = −𝑡 𝑥 = 𝑠 𝑥 = 𝑡 71
−𝑡 0𝑠 − 𝑡 0𝑠 −𝑡 0 −1
X = 𝑠 = 𝑠 + 0𝑡 = 𝑠 + 0 = 𝑠 1 +t 0
𝑡 0𝑠 + 𝑡 0𝑠 𝑡 0 1

0 −1
𝐸𝜆 = 𝑠 1 + 𝑡 0 , 𝑠, 𝑡 ∈ 𝑅
0 1
0 −1
Basis for 𝐸𝜆 = 1 , 0
0 1
dim (𝐸𝜆 ) = 2

1 0 −1
Note that 2 , 1 0 is linear independent and hence basis for 𝑅 , consisting of Eigen vectors T
1 0 1

T is diagonalizable.

Note: The multiplicity of each Eigen value  is equal to the dimension of the corresponding Eigen space 𝐸𝜆

72
NOTE :-

The set 𝐸 is called the eigen space of T, corresponding to the Eigen value .

THEOREM:

Let T be a linear operator on a finite dimensional vector space V, and let  be an eigenvalue of T having
multiplicity M, then 1  dim (𝐸 ) m.

THEOREM

Let 𝑇 be a linear operator on vector space 𝑉 and  ,  , … ,  be distinct Eigen values of 𝑇.

For each 𝑖 = 1,2, … , 𝑘. Let 𝑆 be a finite linear independent subset of the Eigen space 𝐸

then 𝑆 = 𝑆  𝑆  … …  𝑆 is a linear independent subset of 𝑉

73
DEFINITION:

A polynomial 𝑓 𝑡 in 𝑃(𝐹) splits over 𝐹 if there are scalars 𝑐, 𝑎 , 𝑎 … , 𝑎 (not necessary distinct) in 𝐹
such that 𝑓 𝑡 = 𝑐 𝑡 − 𝑎 𝑡−𝑎 … 𝑡−𝑎

EXAMPLE:

𝑡 − 1 = (𝑡 + 1)(𝑡 − 1) splits over 𝑅

𝑡 + 1 = 𝑡 + 𝑖 𝑡 − 𝑖 splits over 𝐶 but not R

TEST FOR DIAGONALIZATION

Let T be a linear operator on an n-dimensional vector space V. Then 𝑇 is diagonalizable iff both of the
following conditions satisfy.

1) The characteristic polynomial of 𝑇 splits

2) For each eigenvalue  of T, the multiplicity of λ equals n − [(rank T − λI ]

74
3 1 0
PROBLEM:- Test diagonalizability of the matrix A = 0 3 0 ∈ M × R
0 0 4

3−t 1 0
SOLUTION: f t = 0 3−t 0 = −1 t−3 t−4 =− t−3 t−4
0 0 4−t
Which splits.

Eigen values of A

λ = 4 with multiplicity 1

λ = 3 with multiplicity 2

75
For 𝜆 = 4

−1 1 0
𝐴 − 4𝐼 = 0 −1 0
0 0 0
∴ 𝑟𝑎𝑛𝑘 𝐴 − 4𝐼 = 2
𝑛 − 𝑟𝑎𝑛𝑘 𝐴 − 4𝐼 = 3 − 2 = 1 = 𝑚𝑢𝑙𝑡𝑖𝑝𝑙𝑖𝑐𝑖𝑡𝑦 𝑜𝑓 𝜆
∴ 𝑐𝑜𝑛𝑑𝑖𝑡𝑖𝑜𝑛 2 𝑖𝑠 satisfied for 𝜆

For 𝜆 = 3

0 1 0
𝐴 − 3𝐼 = 0 0 0
0 0 1
∴ 𝑟𝑎𝑛𝑘 𝐴 − 3𝐼 = 2

Condition (2) fails for 𝜆

𝑛 − 𝑟𝑎𝑛𝑘 𝐴 − 3𝐼 = 3 − 2 = 1 ≠ 2 = multiplicity of 𝜆

𝑇ℎ𝑒𝑟𝑒𝑓𝑜𝑟𝑒, 𝐴 is not diagonalizable. 76

PROBLEM:2

Let 𝑇 be a linear operator on 𝑃 (𝑅) defined by 𝑇 𝑓 𝑥 = 𝑓 1) + 𝑓´(0 𝑥 + (𝑓 0 + 𝑓 0 )𝑥 .

Test for diagonalisability. Let α be the standard basis for 𝑃 𝑅 .

SOLUTION:

Standard ordered basis for 𝑃 𝑅 , 𝛼 = 1, 𝑥, 𝑥

𝑇 1 = 1 + 0 𝑥 + 0 + 0 𝑥 = 1 + 0𝑥 + 0𝑥
𝑇 𝑥 =1+ 1 𝑥+ 1+0 𝑥 = 1+𝑥+𝑥
𝑇 𝑥 = 1 + 2 0 𝑥 + 2(0) + 2 𝑥 = 1 + 0𝑥 + 2𝑥

1 1 1
𝐵 = 𝑇= 0 1 0
0 1 2
The characteristic polynomial for 𝐵 𝑖𝑠 𝐵 − 𝜆𝐼 = 0

1−𝜆 1 1
f(𝜆) = 𝐵 − 𝜆𝐼 = 0  0 1−𝜆 0 =0
77
0 1 2−𝜆
𝑓 𝜆 =− 𝜆−1 𝜆−2
Eigen values are  = 1 with 𝑀 = 2

 = 2 𝑤𝑖𝑡ℎ 𝑀 = 1

0 1 1
For  = 1, rank 𝐵 −  𝐼 = rank 0 0 0 = 1
0 1 1
𝑛 − (rank 𝐵 −  𝐼 )=3 -1= 2 = M

−1 1 1
 = 2, 𝐵− 𝐼 = 0 −1 0
0 1 0
rank 𝐵 −  𝐼 = 2

n- rank 𝐵 −  𝐼 =3-2=1=M

The two condition are satisfied

T is diagonalizable.

78
Find the corresponding basis

𝐴 − 𝜆𝐼 𝑋 = 0

1− 1 1 𝑥 0
0 1− 0 𝑥 = 0
0 1 2− 𝑥 0
For 𝜆 = 2

−1 1 1 𝑥 0
0 −1 0 𝑥 = 0
0 1 0 𝑥 0

 −𝑥 + 𝑥 + 𝑥 = 0 , −𝑥 = 0 −𝑥 + 𝑥 = 0

𝑥 =0, 𝑥 =𝑥

79
𝑘 1
If x = 𝑘 then X = 0 = 𝑘 0
𝑘 1
The eigen space corresponding to 𝜆 = 2 𝑖𝑠 𝑔𝑖𝑣𝑒𝑛 𝑏𝑦

1
𝐸 = 𝑘 0 |𝑘 ∈𝑅
1
𝑑𝑖𝑚 𝐸 = 1

80
For 𝜆 = 1

0 1 1 𝑥 0
0 0 0 𝑥 = 0
0 1 1 𝑥 0
ie., 𝑥 = −𝑥

Put 𝑥 = 𝑠, 𝑥 = 𝑘

𝑥 𝑠 1 0
𝑋 = 𝑥 = −𝑘 = 𝑠 0 + 𝑘 −1
𝑥 𝑘 0 1
𝑑𝑖𝑚𝐸 = 2

1 1 0
𝐵 = 0 , 0 , −1 is Linearly independent
1 0 1
𝐵 = {1 + 𝑥 , 1, −𝑥 + 𝑥 } is a basis consisting of Eigenvectors of 𝑇 for 𝑃 𝑅

𝑇 is diagonalizable.

81
APPLICATION SYSTEM OF DIFFERENTIAL EQUATION

Consider the system of differential equation

𝑥 = 3𝑥 + 𝑥 + 𝑥
𝑥 = 2𝑥 + 4𝑥 + 2𝑥
𝑥 = −𝑥 − 𝑥 + 𝑥

Where , for each 𝑖, 𝑥 = 𝑥 𝑡 is a differentiable real – valued function of the real variable t.

Aim: To solve to this system of linear differential equation using tools of linear algebra.

𝑥 (𝑡)
Note: Let 𝑋: 𝑅 → 𝑅 be a function defined by X 𝑡 = 𝑥 (𝑡) & the derivative of 𝑋 is defined as the function X ,
𝑥 (𝑡)

𝑥 (𝑡)
where 𝑋 𝑡 = 𝑥 (𝑡)
𝑥 (𝑡)
82
Then 𝑋 𝑡 = 𝐴𝑋, where A is the coefficient matrix of the given system.
PROBLEM :-

Find the general solution of system of differential equation

𝑥 = 3𝑥 + 𝑥 + 𝑥
𝑥 = 2𝑥 + 4𝑥 + 2𝑥
𝑥 = −𝑥 − 𝑥 + 𝑥

SOLUTION :

3 1 1
Given 𝐴 = 2 4 2 is the co-efficient matrix of the given system.
−1 −1 1
To find Eigen values & Eigen vectors
𝑓 𝑡 = 𝐴 − 𝑡𝐼 = 0

3−𝑡 1 1
 2 4−𝑡 2 =0
−1 −1 1 − 𝑡
𝑡 − 8𝑡 + 20𝑡 − 16 = 0
 𝑡−2 𝑡−2 𝑡−4 =0 83
The eigenvalues are
 = 2,  = 2,  = 4

𝑧
Let 𝑍 = 𝑧 0 be Eigen vector corresponding to Eigen values then 𝐴 − 𝐼 𝑍 = 0
𝑧

3− 1 1 𝑧 0
2 4− 2 𝑧 = 0 (1)
−1 −1 1− 𝑧 0

84
Case (1)  = 2

From (1)

1 1 1 𝑧 0
2 2 2 𝑧 = 0
−1 −1 −1 𝑧 0
𝑧 + 𝑧 + 𝑧 = 0

Put 𝑧 = 0, 𝑧 = −𝑧

−1
𝑍 = 0
1
Put 𝑧 = 0, 𝑧 = −𝑧

0
Z = −1
1

85
Case (2)  = 4

From (1)

−1 1 1 𝑧 0
2 0 2 𝑧 = 0
−1 −1 −3 𝑧 0
−𝑧 +𝑧 +𝑧 =0

𝑧 +𝑧 =0
−𝑧 − 𝑧 − 3𝑧 = 0

1
𝑍 = 2
−1

−1 0 1
𝑄 = 0 −1 2
1 1 −1

2 0 0
Then 𝑄 𝐴𝑄 = 𝐷 = 0 2 0
86
0 0 4
Now, put 𝐴 = 𝑄𝐷𝑄 in X, we get

𝑋 𝑡 = 𝑄𝐷𝑄 𝑋
𝑄 𝑋 𝑡 = 𝐷𝑄 𝑋 (2)

Define 𝑌: 𝑅 → 𝑅 by Y = 𝑄 𝑋 𝑡 , then by equation (2) 𝑌 = 𝐷𝑌


𝑦 𝑡 2 0 0 𝑦 𝑡
𝑦 𝑡 = 0 2 0 𝑦 𝑡
𝑦 𝑡 0 0 4 𝑦 𝑡

2𝑦 𝑡
= 2𝑦 𝑡
4𝑦 𝑡

87
 𝑦 (𝑡) = 2𝑦 (𝑡)

𝑦 (𝑡) = 2𝑦 (𝑡)
𝑦 (𝑡) = 4𝑦 (𝑡)

The general solution of these equation are

𝑦 𝑡 =𝑐 𝑒 ,𝑦 𝑡 =𝑐 𝑒 ,𝑦 𝑡 =𝑐 𝑒 where 𝑐 , 𝑐 𝑎𝑛𝑑 𝑐 are arbitrary constants.

Finally,

𝑥 𝑡 −1 0 1 𝑐 𝑒
𝑥 𝑡 = 𝑋 𝑡 = 𝑄𝑌 𝑡 = 0 −1 2 𝑐 𝑒
𝑥 𝑡 1 1 −1 𝑐 𝑒

88
−𝑐 𝑒 + 𝑐 𝑒
= −𝑐 𝑒 + 2𝑐 𝑒
𝑐 𝑒 +𝑐 𝑒 −𝑐 𝑒

−1 0 1
𝑋 𝑡 = 𝑒 𝑐 0 +𝑐 −1 + 𝑒 [𝑐 2 ]
1 1 −1
The general solution is
𝑋 𝑡 = 𝑒 𝑧 + 𝑒 𝑧 , 𝑤ℎ𝑒𝑟𝑒 𝑧 ∈ 𝐸𝜆 𝑎𝑛𝑑 𝑧 ∈ 𝐸𝜆

89
PROBLEM:

Find the general solution by the system of differential equations.

𝑥  𝑡 =𝑥 𝑡 +𝑥 𝑡
𝑥  𝑡 = 3𝑥 𝑡 − 𝑥 (𝑡)

90