Matrices, Vectors & Fourier analysis Fall 24-25

Chapter-4
Eigenvalues and Eigenvectors
Let 𝐴 = (𝑎𝑖𝑗 )𝑛×𝑛 be a square matrix. A non-zero vector 𝑽 in ℝ𝑛 is called an eigenvector of 𝐴 if 𝐴𝑽 is a
scalar multiple of 𝑽; that is 𝐴𝑽 = 𝜆𝑽 for some scalar 𝜆. The scalar 𝜆 is called an eigenvalue of 𝐴 and 𝑽 is
called the eigenvector of 𝐴 corresponding to 𝜆.
1 3 0
Example: The vector 𝑽 = ( ) is an eigenvector of 𝐴 = ( ) corresponding to the eigenvalue
2 8 −1
𝜆 = 3.
3 0 1 3 1
𝐴𝑽 = ( ) ( ) = ( ) = 3 ( ) = 𝜆𝑽.
8 −1 2 6 2
Characteristic matrix:
Provided that 𝐴 is a square matrix of order 𝑛 × 𝑛. Then the matrix 𝐴 − 𝜆𝐼 is called the characteristic matrix
where 𝜆 is a scalar and 𝐼 is an unit matrix.

Example:
2 2 1
𝐴 = (1 3 1 )
1 2 2
Characteristic matrix is
2 2 1 1 0 0 2−𝜆 2 1
𝐴 − 𝜆𝐼 = (1 3 1) − 𝜆 (0 1 0) = ( 1 3−𝜆 1 )
1 2 2 0 0 1 1 2 2−𝜆
Characteristic polynomial:
The determinant of characteristic matrix (𝐴 − 𝜆𝐼) is a polynomial of 𝜆. Characteristic polynomial is
2−𝜆 2 1
|𝐴 − 𝜆𝐼| = | 1 3−𝜆 1 |
1 2 2−𝜆
= (2 − 𝜆)(6 − 5𝜆 + 𝜆2 ) − 2(1 − 𝜆) + 1(2 − 3 + 𝜆)

= 𝜆3 − 7𝜆2 + 11𝜆 − 5.

Characteristic equation:
The equation |𝐴 − 𝜆𝐼| = 0 is called characteristic equation for matrix 𝐴.

For example, 𝜆3 − 7𝜆2 + 11𝜆 − 5 = 0 is characteristic equation for the above matrix 𝐴.

Characteristic root(s) or eigenvalue(s):

The roots of the characteristic equation |𝐴 − 𝜆𝐼| = 0 are called characteristic values or eigenvalues of
matrix A.
∴ 𝜆3 − 7𝜆2 + 11𝜆 − 5 = 0

⟹ ( 𝜆 − 1)(𝜆 − 1)(𝜆 − 5) = 0 ⟹ 𝜆 = 1,1,5

So, the characteristic values or eigenvalues are 1, 1 and 5.
1
Matrices, Vectors & Fourier analysis Fall 24-25

Example:
1 2 −1
Find the eigenvalues and eigenvectors of the matrix 𝐴 = (0 −2 0 )
0 5 2
Solution:
The characteristic matrix of 𝐴 is,
1 2 −1 1 0 0 1−𝜆 2 −1
𝐴 − 𝜆𝐼 = (0 −2 0 ) − 𝜆 (0 1 0) = ( 0 −2 − 𝜆 0 )
0 5 2 0 0 1 0 5 2−𝜆
1−𝜆 2 −1
The characteristic polynomial of matrix 𝐴 is |𝐴 − 𝜆𝐼| = | 0 −2 − 𝜆 0 |
0 5 2−𝜆
The characteristic equation of matrix 𝐴 is |𝐴 − 𝜆𝐼| = 0
1−𝜆 2 −1
⇒| 0 −2 − 𝜆 0 |=0
0 5 2−𝜆
∴ (𝜆 − 1)(𝜆 + 2)(𝜆 − 2) = 0
So, the characteristic roots or the eigenvalues of matrix 𝐴 is 𝜆 = 1, −2,2
𝑣1
Now consider 𝑽 = (𝑣2 ) is an eigenvector of matrix 𝐴 corresponding to the eigenvalue 𝜆. From the
𝑣3
definition of eigenvalue and eigenvector MatLab command for finding eigenvalues and
(𝐴 − 𝜆𝐼)𝑽 = 0 eigenvectors:

1−𝜆 2 −1 𝑣1 0 >> A=[1 2 -1;0 -2 0;0 5 2];

( 0 −2 − 𝜆 𝑣
0 ) ( 2 ) = (0)
0 5 2 − 𝜆 𝑣3 0 >> [v,d]=eig(sym(A));

For 𝜆 = 1, >> eigenvalues=eig(A)'

Augmented matrix eigenvalues =
0 2 −1 0 1 2 -2
(0 −3 0 | 0)
0 5 1 0 >> eigenvectors=double(v)
1 0 2 −1 0
𝑟2 → − 𝑟2 ~ (0 1 0 | 0) eigenvectors =
3 0 5 1 0
0 2 −1 0 1.0000 -1.0000 0.8667
𝑟3 → 𝑟3 − 5𝑟2 ~ (0 1 0 | 0)
0 0 -0.8000
0 0 1 0
For 𝜆 = 1, system of equations becomes, 0 1.0000 1.0000
2 𝑣2 − 𝑣3 = 0
𝑣2 = 0
𝑣3 = 0
Solving we get 𝑣2 = 𝑣3 = 0
2
Matrices, Vectors & Fourier analysis Fall 24-25

Here, 𝑣1 is a free variable. Let, 𝑣1 = 𝑎, where 𝑎 (𝑎 ≠ 0) is any real number. Therefore, the eigenvector of
𝑎
𝐴 corresponding to the eigenvalue 𝜆 = 1 is the non-zero vector of the form 𝑽𝟏 = (0).
0
Again, for 𝜆 = −2,
3 2 −1 0 𝑟 ↔ 𝑟 3 2 −1 0
(0 0 0 | 0) 2 ~ 3 (0 5 4 | 0)
0 5 4 0 0 0 0 0
So, System of linear equations becomes,
3 𝑣1 + 2 𝑣2 − 𝑣3 = 0
5 𝑣2 + 4 𝑣3 = 0
This system has one free variable. 𝑣3 is a free variable and let 𝑣3 = 𝑏, where 𝑏 (𝑏 ≠ 0) is any real number.
4𝑏 13𝑏
∴ 𝑣2 = − and ∴ 𝑣1 =
5 15
13𝑏
15
Therefore, 𝑽𝟐 = (− 4𝑏).
5
𝑏
For 𝜆 = 2,
−1 2 −1 0 −1 2 −1 0
( 0 −4 0 | 0) 𝑟3 → 4𝑟3 + 5𝑟2 ~ ( 0 −4 0 | 0)
0 5 0 0 0 0 0 0
So, system of linear equations becomes,
−𝑣1 + 2 𝑣2 − 𝑣3 = 0
−4 𝑣2 = 0
Hence, 𝑣2 = 0 and 𝑣3 is free variable. Let 𝑣3 = 𝑐, where 𝑐 (𝑐 ≠ 0) is any real number.
Then we have 𝑣1 = −𝑐
−𝑐
Therefore, 𝑽𝟑 = ( 0 )
𝑐
13𝑏
𝑎 15 −𝑐
Therefore, eigenvectors are (0), (− ) and ( 0 )corresponding to the eigenvalues 𝜆 = 1, −2, 2
4𝑏
0 5 𝑐
𝑏
respectively.

Example: Solve the following system of differential equation using eigenvalues and eigenvectors.
𝑥1̇ (𝑡) = −1.5𝑥1 (𝑡) + 0.5𝑥2 (𝑡)
{ with 𝑥1 (0) = 5 and 𝑥2 (0) = 4,
𝑥2̇ (𝑡) = 𝑥1 (𝑡) − 𝑥2 (𝑡)
𝑑𝑥1 𝑑𝑥2
where 𝑥1̇ (𝑡) = and 𝑥2̇ (𝑡) = .
𝑑𝑡 𝑑𝑡

3
Matrices, Vectors & Fourier analysis Fall 24-25

Solution:
𝑥1 (𝑡) 𝑥 ̇ (𝑡)
Let, 𝑋(𝑡) = ( ) and 𝑋(̇𝑡) = ( 1 )
𝑥2 (𝑡) 𝑥2̇ (𝑡)
𝑥1 (0) 5
So, 𝑋(0) = ( ) = ( ).
𝑥2 (0) 4
Now the system of differential equation can be written as
−1.5 0.5
𝑋(̇𝑡) = 𝐴𝑋(𝑡), where 𝐴 is the coefficient matrix and 𝐴 = ( ).
1 −1
Let 𝜆 and 𝑽 be the eigenvalue and eigenvector of 𝐴 and 𝐶 is an integral constant then we have the solution
of the form,
𝑋(𝑡) = 𝐶𝑽𝑒 𝜆𝑡 .
−1.5 0.5 1 0
The characteristic matrix of 𝐴 is (𝐴 − 𝜆𝐼) = ( )−𝜆( )
1 −1 0 1
−1.5 − 𝜆 0.5
=( )
1 −1 − 𝜆
−1.5 − 𝜆 0.5
The characteristic polynomial of 𝐴 is |𝐴 − 𝜆𝐼| = | |
1 −1 − 𝜆
The characteristic equation of 𝐴 is |𝐴 − 𝜆𝐼| = 0
−1.5 − 𝜆 0.5
⟹| |
1 −1 − 𝜆
⟹ (𝜆 + 1.5)(𝜆 + 1) − 0.5 = 0
⟹ (𝜆2 + 2.5𝜆 + 1) = 0
⟹ (𝜆 + 0.5)(𝜆 + 2) = 0
So the characteristic roots or the eigenvalues of 𝐴 are 𝜆 = −0.5, −2
𝑣1
Now, consider 𝑽 = (𝑣 ) is the eigenvector of 𝐴 corresponding to the eigenvalue 𝜆. From the definition of
2
eigenvalue and eigenvector, 𝐴𝑽 = 𝜆𝑽
So, (𝐴 − 𝜆𝐼)𝑽 = 0
Augmented matrix form
−1.5 − 𝜆 0.5 0
( | )
1 −1 − 𝜆 0
For 𝜆 = −0.5,
−1 0.5 0
( | )
1 −0.5 0

−1 0.5 0
𝑟2 → 𝑟2 + 𝑟1  ( | )
0 0 0
Solving the above system, we get −𝑣1 + 0.5𝑣2 = 0.
Here 𝑣2 is a free variable. Let 𝑣2 = 𝑎. 𝑣1 = 0.5𝑎
Therefore, the eigenvector of 𝐴 corresponding to the eigenvalue 𝜆 = −0.5 are the non-zero vectors of the
0.5𝑎
form 𝑽1 = ( ).
𝑎

4
Matrices, Vectors & Fourier analysis Fall 24-25

Again, for 𝜆 = −2,

Augmented matrix
0.5 0.5 0
( | )
1 1 0
0.5 0.5 0
𝑟2 → 𝑟2 − 2𝑟1  ( | )
0 0 0
1 1 1 0
𝑟1 → 𝑟  ( | )
0.5 1 0 0 0
Solving the above system, we get 𝑣1 + 𝑣2 = 0.
Here 𝑣2 is a free variable. Let 𝑣2 = 𝑏, 𝑣1 = −𝑏.
Therefore, the eigenvector of 𝐴 corresponding to the eigenvalue 𝜆 = −2 are the non-zero vectors of the
−𝑏
form 𝑽2 = ( ) .
𝑏
So, the solution of the system of differential equation can be written as,
𝑋(𝑡) = 𝐶1 𝑽𝟏 𝑒 𝜆1 𝑡 + 𝐶2 𝑽𝟐 𝑒 𝜆2 𝑡
𝑥 (𝑡) 0.5𝑎 −0.5𝑡 −𝑏
⟹ ( 1 ) = 𝐶1 ( )𝑒 + 𝐶2 ( ) 𝑒 −2𝑡
𝑥2 (𝑡) 𝑎 𝑏
𝑥 (0) 0.5 −1
⟹ ( 1 ) = 𝑎𝐶1 ( ) + 𝑏𝐶2 ( )
𝑥2 (0) 1 1
5 0.5 −1
⟹ ( ) = 𝑎𝐶1 ( ) + 𝑏𝐶2 ( )
4 1 1
We can write,
0.5𝑎𝐶1 − 𝑏𝐶2 = 5
𝑎𝐶1 + 𝑏𝐶2 = 4
6 −2
Solving the system for 𝐶1 and 𝐶2 , we have, 𝐶1 = 𝑎 and 𝐶2 = . Now,
𝑏
𝑥 (𝑡) 0.5 −1
( 1 ) = 6 ( ) 𝑒 −0.5𝑡 − 2 ( ) 𝑒 −2𝑡
𝑥2 (𝑡) 1 1
𝑥 (𝑡) −0.5𝑡 −2𝑡
⇒ ( 1 ) = (3𝑒 −0.5𝑡 ) − (−2𝑒−2𝑡 )
𝑥2 (𝑡) 6𝑒 2𝑒
Therefore,
𝑥1 (𝑡) = 3𝑒 −0.5𝑡 + 2𝑒 −2𝑡
𝑥2 (𝑡) = 6𝑒 −0.5𝑡 − 2𝑒 −2𝑡

5
Matrices, Vectors & Fourier analysis Fall 24-25

Sample Exercise-4.1
1. Find the eigenvalues and eigenvectors of the following matrices

1 4
a. ( ) Ans: 𝜆1 = −1, 𝜆2 = 5, 𝑽1 = (−2𝑎, 𝑎)𝑇 , 𝑽2 = (𝑏, 𝑏)𝑇
2 3

2 1
b. ( ) Ans: 𝜆1 = 1, 𝜆2 = 3, 𝑽1 = (−𝑎, 𝑎)𝑇 , 𝑽2 = (𝑏, 𝑏)𝑇
1 2

1 0
c. ( ) Ans: 𝜆1 = 1, 𝜆2 = 3, 𝑽1 = (−2𝑎, 𝑎)𝑇 , 𝑽2 = (0, 𝑏)𝑇
1 3
4 0 1 Ans: 𝜆1 = 1, 𝜆2 = 2, 𝜆3 = 3
d. (−2 1 0)
−2 0 1 𝑽1 = (0, 𝑎, 0)𝑇 , 𝑽2 = (−𝑏, 2𝑏, 2𝑏)𝑇 , 𝑽3 = (−𝑐, 𝑐, 𝑐)𝑇
5 0 1 Ans: 𝜆1 = −2, 𝜆2 = 4, 𝜆3 = 6
e. (0 −2 0)
1 0 5 𝑽1 = (0, 𝑎, 0)𝑇 , 𝑽2 = (𝑏, 0, −𝑏)𝑇 , 𝑽3 = (𝑐, 0, 𝑐)𝑇

2. Solve the following system of differential equations using eigenvalue and eigenvector where
𝑑𝑥 𝑑𝑥
𝑥1̇ (𝑡) = 1 and 𝑥2̇ (𝑡) = 2 .
𝑑𝑡 𝑑𝑡

𝑥1̇ (𝑡) = −𝑥1 (𝑡) + 4𝑥2 (𝑡) Ans.

a. { 𝑥1 (𝑡)
𝑥2̇ (𝑡) = 3𝑥1 (𝑡) − 2𝑥2 (𝑡) 4 −1
( ) = 𝑒 2𝑡 ( ) + 𝑒 −5𝑡 ( ).
with 𝑥1 (0) = 3, 𝑥2 (0) = 4. 𝑥2 (𝑡) 3 1

𝑥1̇ (𝑡) = 5𝑥1 (𝑡) + 𝑥2 (𝑡) Ans.

b. { 𝑥1 (𝑡)
𝑥2̇ (𝑡) = 𝑥1 (𝑡) + 5𝑥2 (𝑡) 1 −1
( ) = −𝑒 6𝑡 ( ) − 2𝑒 4𝑡 ( )
with 𝑥1 (0) = 1, 𝑥2 (0) = −3. 𝑥2 (𝑡) 1 1

𝑥1̇ (𝑡) = −2𝑥1 + 3𝑥2 (𝑡) Ans.

c. { 𝑥1 (𝑡)
𝑥2̇ (𝑡) = 4𝑥1 (𝑡) − 𝑥2 (𝑡) 3 −1
( ) = 𝑒 2𝑡 ( ) − 𝑒 5𝑡 ( )
with 𝑥1 (0) = 4, 𝑥2 (0) = 3. 𝑥2 (𝑡) 4 1

6
Matrices, Vectors & Fourier analysis Fall 24-25

Cayley-Hamilton Theorem:
Every square matrix satisfies its characteristic equation, if the characteristic equation is

𝜆𝑛 + 𝑎𝑛−1 𝜆𝑛−1 + 𝑎𝑛−2 𝜆𝑛−2 + ⋯ … . . + + 𝑎1 𝜆 + 𝑎0 𝐼 = 0

Then 𝐴𝑛 + 𝑎𝑛−1 𝐴𝑛−1 + 𝑎𝑛−2 𝐴𝑛−2 + ⋯ … . . + + 𝑎1 𝐴 + 𝑎0 𝐼 = 0
1 2 3
Example: Verify the Cayley-Hamilton theorem for the matrix 𝐴 = (2 −1 1)
3 1 1
1 2 3 1 0 0
Solution: The characteristic matrix of 𝐴 is 𝐴 − 𝜆𝐼 = (2 −1 1) − 𝜆 (0 1 0)
3 1 1 0 0 1
1−𝜆 2 3
=( 2 −1 − 𝜆 1 )
3 1 1−𝜆

Therefore, the characteristic equation of the matrix 𝐴 is |𝐴 − 𝜆𝐼| = 0

1−𝜆 2 3
| 2 −1 − 𝜆 1 |=0
3 1 1−𝜆
⟹ 𝜆3 − 𝜆2 − 15𝜆 − 15 = 0
According to Cayley-Hamilton theorem

𝐴3 − 𝐴2 − 15𝐴 − 15𝐼 = 0
1 2 3
Now, 𝐴 = (2 −1 1)
3 1 1
1 2 3 1 2 3 14 3 8
𝐴2 = (2 −1 1) (2 −1 1) = ( 3 6 6)
3 1 1 3 1 1 8 6 11
14 3 8 1 2 3 44 33 53
𝐴3 = 𝐴2 𝐴 = ( 3 6 6 ) (2 −1 1) = (33 6 21)
8 6 11 3 1 1 53 21 41
∴ 𝐴3 − 𝐴2 − 15𝐴 − 15𝐼
44 33 53 14 3 8 1 2 3 1 0 0
= (33 6 21) − ( 3 6 6 ) − 15 (2 −1 1) − 15 (0 1 0)
53 21 41 8 6 11 3 1 1 0 0 1
0 0 0
= (0 0 0) = 0
0 0 0
∴ 𝐴3 − 𝐴2 − 15𝐴 − 15𝐼 = 0
Hence Cayley-Hamilton theorem is verified.

7
Matrices, Vectors & Fourier analysis Fall 24-25

1 2 3
Example: Using Cayley-Hamilton theorem find the inverse of the matrix 𝐴 = (2 −1 1)
3 1 1
1 2 3 1 0 0
Solution: The characteristic matrix of A is 𝐴 − 𝜆𝐼 = (2 −1 1) − 𝜆 (0 1 0)
3 1 1 0 0 1

1−𝜆 2 3
=( 2 −1 − 𝜆 1 )
3 1 1−𝜆
Therefore, the characteristic equation of the matrix A is |𝐴 − 𝜆𝐼| = 0
1−𝜆 2 3
| 2 −1 − 𝜆 1 |=0
3 1 1−𝜆
⟹ 𝜆3 − 𝜆2 − 15𝜆 − 15 = 0
Now according to the Cayley-Hamilton theorem, we have

𝐴3 − 𝐴2 − 15𝐴 − 15𝐼 = 0

⟹ 𝐴2 − 𝐴 − 15𝐼 − 15𝐴−1 = 0 [Multiplying by 𝐴−1]

1
⟹ 𝐴−1 = 15 [𝐴2 − 𝐴 − 15𝐼]

Here,
1 2 3 1 2 3 14 3 8
2
𝐴 = (2 −1 1) (2 −1 1) = ( 3 6 6)
3 1 1 3 1 1 8 6 11

14 3 8 1 2 3 1 0 0
−1 1
∴𝐴 = 15 (( 3 6 6 ) − (2 −1 1) − 15 (0 1 0))
8 6 11 3 1 1 0 0 1

−2 1 5
1
= 15 ( 1 −8 5 ).
5 5 −5
1 2 2
Example: Using Cayley-Hamilton theorem find the inverse of the matrix 𝐴 = (3 1 0)
1 1 1
1 2 2 1 0 0
The characteristic matrix of 𝐴 is 𝐴 − 𝜆𝐼 = (3 1 0) − 𝜆 (0 1 0)
1 1 1 0 0 1
1−𝜆 2 2
=( 3 1−𝜆 0 ).
1 1 1−𝜆
Therefore, the characteristic equation of the matrix 𝐴 is |𝐴 − 𝜆𝐼| = 0

8
Matrices, Vectors & Fourier analysis Fall 24-25

1−𝜆 2 2
| 3 1−𝜆 0 |=0
1 1 1−𝜆
⟹ 𝜆3 − 3𝜆2 − 5𝜆 + 1 = 0
Now according to the Cayley –Hamilton theorem, we have

𝐴3 − 3𝐴2 − 5𝐴 + 𝐼 = 0

⟹ 𝐴2 − 3𝐴 − 5𝐼 + 𝐴−1 = 0

𝐴−1 = 3𝐴 + 5𝐼 − 𝐴2
1 2 2
𝐴 = (3 1 0 )
1 1 1
1 2 2 1 2 2 9 6 4
𝐴2 = (3 1 0) (3 1 0) = (6 7 6)
1 1 1 1 1 1 3 4 3
1 2 2 1 0 0 9 6 4 −1 0 2
∴ 𝐴−1 = 3 (3 1 0) + 5 (0 1 0) − (6 7 6) = ( 3 1 −6)
1 1 1 0 0 1 3 4 3 −2 −1 5
Sample Exercise-4.2
State the Cayley-Hamilton theorem. Hence find the inverse of the following matrices using Cayley-
Hamilton theorem and verify your result.

1 2 1 −4 2
a. A  Ans. 𝐴−1 = ( )
3 4 2 3 −1

2 1 1 2 −1 −1
  1
b. A   0 1 0  Ans. 𝐴−1 = 3 ( 0 3 0)
1 1 2 −1 −1 2
 

3 −3 4 1 −1 0
c. 𝐴 = (2 −3 4) Ans. 𝐴−1 = (−2 3 −4)
0 −1 1 −2 3 −3

1 3 3 7 −3 −3
d. 𝐴 = (1 4 3) Ans. 𝐴−1 = (−1 1 0)
1 3 4 −1 0 1

1 −1 1 1 2 −1
e. 𝐴 = (4 1 0) Ans. 𝐴−1 = (−4 −7 4)
8 1 1 −4 −9 5

9
Matrices, Vectors & Fourier analysis Fall 24-25

Vector
Vector Spaces

Vectors in ℝ𝟐 :

The set of all ordered pairs of real numbers is called two-dimensional vector space and is denoted by ℝ2 .

ℝ2 = {(𝑥, 𝑦): 𝑥, 𝑦 ∈ ℝ}, where ℝ2 = ℝ × ℝ

Example: (1,2); (2, −1); (3,4); (√3, 5); (0, e) ∈ ℝ2

Vectors in ℝ𝟑 :

The set of all ordered triplets of real numbers is called three-dimensional vector space and is denoted by ℝ3

ℝ3 = {(𝑥, 𝑦, 𝑧): 𝑥, 𝑦, 𝑧 ∈ ℝ}, where ℝ3 = ℝ × ℝ × ℝ

Vectors in ℝ𝐧 :
If 𝑛 is a positive integer then the set of all ordered n tuples of real numbers is called vector in 𝑛-dimensional
space and is denoted by ℝ𝑛 and if 𝒖 ∈ ℝn ; 𝑢 = (𝒖1 , 𝒖2 , … … , 𝒖𝑛 ), then 𝒖 is called a 𝑛-dimensional vector
in ℝ𝑛 . In Cartesian co-ordinate system a particular n tuples in ℝ𝐧 is called co-ordinates of a point.

Addition of two vectors in ℝ𝟐 :

If a= (𝑎1 , 𝑎2 ) and 𝒃 = (𝑏1 , 𝑏2 ) be two vectors in ℝ𝟐 then 𝒂 + 𝒃 = (𝑎1 + 𝑏1 , 𝑎2 + 𝑏2 )

Fig: addition of two vectors.

Subtraction of two vectors in ℝ𝟐 :

If a= (𝑎1 , 𝑎2 ) and 𝒃 = (𝑏1 , 𝑏2 ) be two vectors in ℝ𝟐 then 𝒂 − 𝒃 = (𝑎1 − 𝑏1 , 𝑎2 − 𝑏2 )

Fig: Subtraction of two vectors.

10
Matrices, Vectors & Fourier analysis Fall 24-25

Scalar multiplication of vectors in ℝ𝟐 :

Let 𝑘 ∈ ℝ be a scalar and 𝒖 ∈ ℝ2 where 𝒖 = (𝑢1 , 𝑢2 ) is a vector in ℝ2 then

𝑘𝒖 = 𝑘(𝒖1 , 𝒖2 ) = (𝑘𝒖1 , 𝑘𝒖2 )

Zero Vector:
The vector whose components are all zero is called the zero vector and is denoted by 0 and defined by 𝟎 =
(0,0, … . . ,0)

Dot or Inner product of two non-zero vectors:

Let 𝒖(𝑢1 , 𝑢2 ) and 𝒗(𝑣1 , 𝑣2 ) be two non-zero vectors and 𝜃 be the angle between them. The dot product of
𝒖 and 𝒗 is denoted by 𝒖 ∙ 𝒗 and defined by 𝒖 ∙ 𝒗 = ‖𝒖‖‖𝒗‖cosθ = 𝑢1 𝑣1 + 𝑢2 𝑣2 .

Parallel vectors in ℝ𝟐 :

Let 𝒖, 𝒗𝜖ℝ2 , then the vectors 𝒖 and 𝒗 are parallel if 𝒖 = 𝑘𝒗. If 𝒌 > 𝟎 then they are in the same direction
and if 𝒌 < 𝟎 then they are in opposite direction.

Perpendicular vectors in ℝ𝟐 :
Let 𝒖, 𝒗 ∈ ℝ2 then 𝒖 and 𝒗 are said to be perpendicular (or orthogonal) if 𝒖 ∙ 𝒗 = 0.

Example: Let 𝒖, 𝒗 ∈ ℝ2 where = (3, −4), 𝒗 = (8,6); 𝒖 ∙ 𝒗 = 3(8) + (−4)(6) = 0.

Hence 𝒖 and 𝒗 are perpendicular vectors.

Distance between two vectors in ℝ𝟐 :

Let 𝒖, 𝒗 ∈ ℝ2 where 𝒖(𝑢1 , 𝑢2 ) and 𝒗(𝑣1 , 𝑣2 ), then the distance between 𝒖 and 𝒗, denoted by 𝑑(𝒖, 𝒗) is
defined by 𝑑(𝒖, 𝒗) = √(𝑣1 − 𝑢1 )2 + (𝑣2 − 𝑢2 )2

Example: If 𝒖 = (5,4), 𝒗 = (1,1) then 𝑑(𝒖, 𝒗) = √(5 − 1)2 + (4 − 1)2 = √16 + 9 = 5.

Norm or Length in ℝ𝟐 :
Let 𝒖 ∈ ℝ2 , where 𝒖 = (𝑢1 , 𝑢2 ) then the norm or length of 𝒖 denoted by ‖𝒖‖ and is defined by ‖𝒖‖ =
√𝒖 ∙ 𝒖 = √𝑢12 +𝑢22
Example: Let 𝒖, 𝒗 ∈ ℝ3 ; where 𝒖 = (1,0, −1), 𝒗 = (2, −3,1) then find
(𝑖)2𝒖 + 3𝒗 (𝑖𝑖)𝒖 ∙ 𝒗 (iii) ‖𝒗‖ (𝑖𝑣) ‖𝒖 − 𝒗‖ (𝑣)𝑑(𝒖, 𝒗)
Solution:
(𝑖) 2𝒖 + 3𝒗 = 2(1,0, −1) + 3(2, −3,1) = (2,0, −2) + (6, −9,3) = (8, −9,1)
(𝑖𝑖) 𝒖 ∙ 𝒗 = 1(2) + 0(−3) + (−1)1 = 2 − 1 = 1

(𝑖𝑖𝑖) ‖𝒗‖ = √4 + 9 + 1 = √14

(𝑖𝑣) 𝒖 − 𝒗 = (1 − 2,0 + 3, −1 − 1) = (−1,3, −2)

∴ ‖𝒖 − 𝒗‖ = √(−1)2 + 32 + (−2)2 = √14

(𝑣)𝑑(𝒖, 𝒗) = √(1 − 2)2 + (0 + 3)2 + (−1 − 1)2 = √14 .

11
Matrices, Vectors & Fourier analysis Fall 24-25

Vector Spaces:
Let 𝐹 be a field of scalars and 𝑽 be a non-empty set of vectors. If 𝑽 satisfies the following axioms of vector
addition and scalar multiplication, then 𝑽 is called vector space over 𝑭.
In vector addition:
𝐴1 : 𝒖, 𝒗 ∈ 𝑉 ⇒ (𝒖 + 𝒗) ∈ 𝑽
𝐴2 : 𝒖, 𝒗, 𝒘 ∈ 𝑽 ⇒ (𝒖 + 𝒗) + 𝒘 = 𝒖 + (𝒗 + 𝒘)
𝐴3 : 𝒖, 𝟎 ∈ 𝑽 ⇒ (𝒖 + 𝟎) = (𝟎 + 𝒖) = 𝒖, here 𝟎 is the zero vector
𝐴4 : 𝒖, 𝒗 ∈ 𝑽 ⇒ 𝒖 + 𝒗 = 𝒗 + 𝒖
𝐴5 : 𝒖 ∈ 𝑽 ⇒ −𝒖 ∈ 𝑽 ⇒ 𝒖 + (−𝒖) = (−𝒖) + 𝒖 = 𝟎
In scalar multiplication:
𝑀1 : 𝑎 ∈ 𝐹 and 𝒖 ∈ 𝑽 ⇒ 𝑎𝒖 ∈ 𝑽
𝑀2 : 𝑎 ∈ 𝐹 and 𝒖, 𝒗 ∈ 𝑽 ⇒ 𝑎(𝒖 + 𝒗) = 𝑎𝒖 + 𝑎𝒗
𝑀3 : 𝑎, 𝑏 ∈ 𝐹 and 𝒖 ∈ 𝑽 ⇒ (𝑎 + 𝑏)𝒖 = 𝑎𝒖 + 𝑏𝒖
𝑀4 : 𝑎, 𝑏 ∈ 𝐹 and 𝒖 ∈ 𝑽 ⇒ (𝑎𝑏)𝒖 = 𝑎(𝑏𝒖)
𝑀5 : 1 ∈ 𝐹 ⇒ 1. 𝒖 = 𝒖. 1 = 𝒖; 𝒖∈𝑽

Subspace:
Let 𝑾 be a non empty subset of a vector space 𝑽 over the field 𝐹. We call 𝑾 a subspace of 𝑽 if and only if
𝑾 is a vector space over the field 𝐹 under the laws of vector addition and scalar multiplication defined on
𝑽, or equivalently, 𝑾 is a subspace of 𝑽 wherever 𝒘𝟏 , 𝒘𝟐 ∈ 𝑾 and 𝛼, 𝛽 ∈ 𝐹 implies that 𝛼𝒘𝟏 + 𝛽𝒘𝟐 ∈
𝑾.

Example: Show that 𝑆 = {(𝑎, 𝑏, 𝑐) ∣ 𝑎, 𝑏, 𝑐 ∊ ℝ and 𝑎 + 𝑏 + 𝑐 = 0 } is a subspace of the vector space ℝ𝟑 .

Solution: Here 𝟎 ∊ ℝ𝟑 , 𝟎 = (0,0,0) ∊ 𝑆 ∴0+0+0= 0

Hence 𝑆 is non empty set. Again, let 𝒖 = (𝑎1 , 𝑏1 , 𝑐1 ) ∊ 𝑆 and 𝒗 = (𝑎2 , 𝑏2 , 𝑐2 ) ∊ 𝑆

𝑎1 + 𝑏1 + 𝑐1 = 0 and 𝑎2 + 𝑏2 + 𝑐2 = 0

For any scalars α, β we get 𝛼𝒖 + 𝛽𝒗 = 𝛼(𝑎1 , 𝑏1 , 𝑐1 ) + 𝛽(𝑎2 , 𝑏2 , 𝑐2 )

= (𝛼𝑎1 + 𝛽𝑎2 , 𝛼𝑏1 + 𝛽𝑏2 , 𝛼𝑐1 + 𝛽𝑐2 )

Now, 𝛼𝑎1 + 𝛽𝑎2 + 𝛼𝑏1 + 𝛽𝑏2 + 𝛼𝑐1 + 𝛽𝑐2

= 𝛼(𝑎1 + 𝑏1 + 𝑐1 ) + 𝛽(𝑎2 + 𝑏2 + 𝑐2 ) = 𝛼. 0 + 𝛽. 0 = 0

∴ 𝛼𝒖 + 𝛽𝒗 ∈ 𝑆. Hence 𝑆 is a subspace of ℝ𝟑 .

Linear combination, Dependency and Independency of Vector

Linear combination:
Let 𝑽(𝐹) be a vector space, where 𝒗1 , 𝒗2 , 𝒗3 … 𝒗𝑛 ∊ 𝑽 and 𝛼1 , 𝛼2 , 𝛼3 … 𝛼𝑛 ∊ 𝐹. If 𝛼1 𝒗1 + 𝛼2 𝒗2 + 𝛼3 𝒗3 +
⋯ + 𝛼𝑛 𝒗𝑛 = 𝒖 ∊ 𝑽, then 𝒖 is called linear combination of 𝒗1 , 𝒗2 , 𝒗3 … 𝒗𝑛 .
12
Matrices, Vectors & Fourier analysis Fall 24-25

Linear dependency of vectors:

Let 𝑽(𝐹) be a vector space, where 𝒗1 , 𝒗2 , 𝒗3 … 𝒗𝑛 ∊ 𝑽 and 𝛼1 , 𝛼2 , 𝛼3 … 𝛼𝑛 ∊ 𝐹. If 𝛼1 𝒗1 + 𝛼2 𝒗2 + 𝛼3 𝒗3 +
⋯ + 𝛼𝑛 𝒗𝑛 = 𝟎, and at least one of the element of the set {𝛼1 , 𝛼2 , 𝛼3 … 𝛼𝑛 } is not zero, then the vectors
𝒗1 , 𝒗2 , 𝒗3 … 𝒗𝑛 are linearly dependent.

Linear independency of vectors:

Let 𝑽(𝐹) be a vector space and 𝒗1 , 𝒗2 , 𝒗3 , … 𝒗𝑛 ∊ 𝑽 and 𝛼1 , 𝛼2 , 𝛼3 , … 𝛼𝑛 ∊ 𝐹. If 𝛼1 𝒗1 + 𝛼2 𝒗2 + 𝛼3 𝒗3 +
⋯ + 𝛼𝑛 𝒗𝑛 = 𝟎, and all of the elements of the set {𝛼1 , 𝛼2 , 𝛼3 … 𝛼𝑛 } are zero except the trivial solution (𝛼1 =
𝛼2 = ⋯ = 𝛼𝑛 = 0), then the vectors 𝒗1 , 𝒗2 , 𝒗3 … 𝒗𝑛 are linearly independent.

Example: Write the vector 𝒖 = (1, −2,5) as a linear combination of the vectors

𝒖𝟏 = (1,1,1), 𝒖𝟐 = (1,2,3) and 𝒖𝟑 = (2, −1,1).

Solution: Let 𝛼1 𝒖𝟏 + 𝛼2 𝒖𝟐 + 𝛼3 𝒖𝟑 = 𝒖 ; where 𝛼1 , 𝛼2 , 𝛼3 are scalars.

𝛼1 (1,1,1) + 𝛼2 (1,2,3) + 𝛼3 (2, −1,1) = (1, −2,5)

(𝛼1 + 𝛼2 + 2𝛼3 , 𝛼1 + 2𝛼2 − 𝛼3 , 𝛼1 + 3𝛼2 + 𝛼3 ) = (1, −2,5)

Equating corresponding components

𝛼1 + 𝛼2 + 2𝛼3 = 1

𝛼1 + 2𝛼2 − 𝛼3 = −2

𝛼1 + 3𝛼2 + 𝛼3 = 5
The augmented matrix of the above system is
1 1 2 1
(1 2 −1| −2)
1 3 1 5
Reducing the system to REF by the elementary row operations

1 1 2 1
𝑟2 → 𝑟2 − 𝑟1
~ (0 1 −3| −3)
𝑟3 → 𝑟3 − 𝑟1
0 2 −1 4
1 1 2 1
𝑟3 → 𝑟3 − 2𝑟2 ~ (0 1 −3| −3)
0 0 5 10
1 1 1 2 1
𝑟3 → 𝑟 ~ (0 1 −3| −3)
5 3 0 0 1 2
So, system of linear equations becomes,

𝛼1 + 𝛼2 + 2𝛼3 = 1

𝛼2 − 3𝛼3 = −3

𝛼3 = 2
Solving, we get 𝛼1 = −6, 𝛼2 = 3 and 𝛼3 = 2.

∴−6 𝒖𝟏 + 3𝒖𝟐 + 2𝒖𝟑 = 𝒖. Hence 𝒖 is a linear combination of the vectors 𝒖𝟏 , 𝒖𝟐 and 𝒖𝟑 .

13
Matrices, Vectors & Fourier analysis Fall 24-25

Example: Test whether the vectors 𝒖𝟏 = (1, 0, 1), 𝒖𝟐 = (−3, 2, 6) & 𝒖𝟑 = (4, 5, −2) are linearly
dependent or independent.

Solution: Let 𝛼1 𝒖𝟏 + 𝛼2 𝒖𝟐 + 𝛼3 𝒖𝟑 = 𝟎 ; where 𝛼1 , 𝛼2 , 𝛼3 are scalars.

𝛼1 (1, 0, 1) + 𝛼2 (−3, 2, 6) + 𝛼3 (4, 5, −2) = (0, 0, 0)

(𝛼1 − 3𝛼2 + 4𝛼3 , 0 + 2𝛼2 + 5𝛼3 , 𝛼1 + 6𝛼2 − 2𝛼3 ) = (0, 0, 0)

We can write from above equation

𝛼1 − 3𝛼2 + 4𝛼3 = 0

0 + 2𝛼2 + 5𝛼3 = 0

𝛼1 + 6𝛼2 − 2𝛼3 = 0
The augmented matrix of the above system is
1 −3 4 0
(0 2 5 | 0)
1 6 −2 0
Reducing the system to REF by the elementary row operations

1 −3 4 0
𝑟3 → 𝑟3 − 𝑟1 ~ (0 2 5 | 0)
0 9 −6 0
1 −3 4 0
𝑟3 → 2𝑟3 − 9𝑟2 ~ (0 2 5 | 0)
0 0 −57 0
1 1 −3 4 0
𝑟3 → 𝑟 ~ (0 2 5| 0)
−57 3 0 0 1 0
So, system of linear equations becomes,

𝛼1 − 3𝛼2 + 4𝛼3 = 0

2𝛼2 + 5𝛼3 = 0

𝛼3 = 0
Solving the above linear system, we get 𝛼1 = 0, 𝛼2 = 0 and 𝛼3 = 0.

Hence the vectors 𝒖𝟏 , 𝒖𝟐 and 𝒖𝟑 are linearly independent.

14
Matrices, Vectors & Fourier analysis Fall 24-25

Sample Exercise-4.3
Write the vector 𝒖 as a linear combination of the vectors 𝒖𝟏 , 𝒖𝟐 and 𝒖𝟑 , where

1. 𝒖 = (5, −1, 9) , 𝒖𝟏 = (1,2, 1), 𝒖𝟐 = (2, 9, 0), 𝒖𝟑 = (3, 3, 4). 𝐀𝐧𝐬: (𝟏, −𝟏, 𝟐)
2. 𝒖 = (2, −1, 3) , 𝒖𝟏 = (1, 0, 0), 𝒖𝟐 = (2, 2, 0), 𝒖𝟑 = (3, 3,3). 𝑨𝐧𝐬: (𝟑, −𝟐, 𝟏)
3. 𝒖 = (4,2, 1,0) ,𝒖𝟏 = (6, −1, 2, 1), 𝒖𝟐 = (1,7, −3, −2), 𝒖𝟑 = (3, 1, 0,0),
𝒖𝟒 = (𝟑, 𝟑, −𝟐, −𝟏). 𝐀𝐧𝐬: (𝟐, 𝟏, −𝟑, 𝟎).
Test whether the following vectors are linearly independent or dependent.

4. 𝒖𝟏 = (1, −2, 3), 𝒖𝟐 = (5, 6, −1), 𝒖𝟑 = (3, 2, 1). 𝐀𝐧𝐬: Linearly dependent.
5. 𝒖𝟏 = (1, 2, 2, −1), 𝒖𝟐 = (4, 9, 9, −4), 𝒖𝟑 = (5, 8, 9, −5). A𝐧𝐬: Linearly independent.
6. 𝒖𝟏 = (−3, 0, 4), 𝒖𝟐 = (5, −1, 2), 𝒖𝟑 = (1, 1, 3). 𝐀𝐧𝐬: Linearly independent.
7. 𝒖𝟏 = (2, 2, 1), 𝒖𝟐 = (−1, − 1 ,0), 𝒖𝟑 = (3, 3, 4). 𝐀𝐧𝐬: Linearly dependent.
8. 𝒖𝟏 = (−2, 0, 1), 𝒖𝟐 = (3, 2, 5), 𝒖𝟑 = (6, −1,1), 𝒖𝟒 = (7, 0, −2).
𝐀𝐧𝐬: Linearly dependent.

15