Module 2

Eigenvalues, Eigenvectors and Diagonalization of Matrix

Out Lines
This present module will include the following topics:
(i) Eigenvalues and eigenvectors of matrices
(ii) Characteristic equations for matrices
(iii) Similarity between matrices
(iv) Diagonalization of matrices using the concept of eigenvalues
(v) Statement of Cayley-Hamilton theorem without proof and its application
(vi) Evaluation of the matrix’s power using Cayley Hamilton theorem

Introduction
In the study of linear algebra, the theory of eigenvalues and eigenvectors are of much
significant. An eigen vector (sometimes, it is called as characteristic vector) is a non-zero vector
concerning to a linear transformation (like, matrix) which provides sense of stretching in
particular direction by the linear transformation. There is a non-zero real value associated with
eigenvector which is used as a factor (scale) for stretching vector. This is known as eigenvalue
or characteristic value of the linear transformation corresponding to the eigenvector. Later, in
our discussion, we will provide the mathematical definitions of eigenvector and eigenvalues in
more precise sense.

Eigenvalues and eigenvectors

Let us consider a system of linear homogenous equations consisting 𝑛 equations and 𝑛
unknowns as follows:
𝑎11 𝑥1 + 𝑎12 𝑥2 + ⋯ + 𝑎1𝑛 𝑥𝑛 = 0
𝑎21 𝑥1 + 𝑎22 𝑥2 + ⋯ + 𝑎2𝑛 𝑥𝑛 = 0
……………………………
𝑎𝑛1 𝑥1 + 𝑎𝑛2 𝑥2 + ⋯ + 𝑎𝑛𝑛 𝑥𝑛 = 0
The above system of linear equation can be written in a matrix notation as follows:
𝑎11 𝑎12 … . . 𝑎1𝑛 𝑥1
𝐴𝑋 = 0, where 𝐴 = [𝑎… 21 𝑎22 … . . 𝑎2𝑛 ] is a 𝑛 × 𝑛 square matrix, 𝑋 = [ 𝑥2 ] is a column
…. …. ..
𝑎𝑛1 𝑎𝑛2 𝑎𝑛𝑛 𝑥𝑛
matrix of 𝑛 × 1 order and 0 the zero matrix of 1 × 𝑛 order.
Similarly, for a scalar 𝜆, we consider the following system of linear homogenous equations:
(𝑎11 − 𝜆)𝑥1 + 𝑎12 𝑥2 + ⋯ + 𝑎1𝑛 𝑥𝑛 = 0
𝑎21 𝑥1 + (𝑎22 − 𝜆)𝑥2 + ⋯ + 𝑎2𝑛 𝑥𝑛 = 0
……………………………
𝑎𝑛1 𝑥1 + 𝑎𝑛2 𝑥2 + ⋯ + (𝑎𝑛𝑛 − 𝜆)𝑥𝑛 = 0
Here, also the system can be written in matrix notation as follows:
𝑎11 𝑎12 … . . 𝑎1𝑛
(𝐴 − 𝜆𝐼)𝑋 = 0, where 𝜆 is a scalar, 𝐴 = [ … 𝑎 21 𝑎22 … . . 𝑎2𝑛 ] is a 𝑛 × 𝑛 square matrix, 𝐼 =
…. ….
𝑎𝑛1 𝑎𝑛2 𝑎𝑛𝑛
1 0….. 0 𝑥1
0 0…
[… ….. . …0 .] is the identity matrix of 𝑛 × 𝑛 order, 𝑋 = [ 𝑥. 2. ] is a column matrix of 𝑛 × 1
0 0 1 𝑥𝑛
order and 0 the zero matrix of 1 × 𝑛 order.
0
Now, the above system (𝐴 − 𝜆𝐼)𝑋 = 0 has trivial solution 𝑋 = [.0.], when the matrix (𝐴 − 𝜆𝐼)
0
is non-singular, i.e., 𝑑𝑒𝑡 (𝐴 − 𝜆𝐼) ≠ 0. So, for getting non-trivial (non-zero) solution 𝑋, the
matrix (𝐴 − 𝜆𝐼) should be singular, i.e., 𝑑𝑒𝑡 (𝐴 − 𝜆𝐼) = 0. For such cases, (𝐴 − 𝜆𝐼) is called
the characteristic matrix and 𝑑𝑒𝑡 (𝐴 − 𝜆𝐼) = 0 is called the characteristic equation
corresponding to the linear transformation (matrix) 𝐴. More precisely, the characteristic
equation can be written as follows:

𝑎11 − 𝜆 𝑎12 … . . 𝑎1𝑛

| 𝑎…21 𝑎22 − 𝜆 … . .
….
𝑎2𝑛 | = 0
…. (2.1)
𝑎𝑛1 𝑎𝑛2 𝑎𝑛𝑛 − 𝜆
After expanding the determinant in the left-hand side of equation (2.1), we get a polynomial
equation of 𝜆 in the following way:
(−1)𝑛 𝜆𝑛 + 𝐴1 𝜆𝑛−1 + 𝐴2 𝜆𝑛−2 + ⋯ + 𝐴𝑛 = 0 (2.2)
Equation (2.2) represents the characteristic equation as a polynomial equation for 𝜆. In equation
(2.2), 𝐴1 , 𝐴2 , …., 𝐴𝑛 are the functions of 𝑎𝑖𝑗 in Equation (2.1). Now, it is the time to define
eigenvalue using the above equations.
Definition 2.1 (Eigenvalue) Suppose, 𝐴 be a 𝑛 × 𝑛 square matrix. Then, the roots of the
polynomial equation (2.2) (known as characteristic equation of the matrix 𝐴) is called the
eigenvalues of the matrix 𝐴. In other terminologies, they are also said to be latent roots or
characteristic roots.
Definition 2.2 (Eigenvector) Suppose, 𝐴 be a 𝑛 × 𝑛 square matrix and 𝜆 be an eigenvalue of
𝐴. Then, corresponding to the eigen value 𝜆, there exists a non-zero vector 𝑋 such that 𝐴𝑋 =
𝜆𝑋. The nonzero vector 𝑋 is called eigenvector or characteristic vector of the matrix 𝐴
corresponding to the eigenvalue 𝜆.

Properties of Eigenvalues and Eigenvectors

Property-1 Suppose, 𝐴 = [𝑎𝑖𝑗 ]𝑛×𝑛 be a 𝑛 × 𝑛 square matrix and 𝜆𝑖 , for 𝑖 = 1,2, . . , 𝑛 are
eigenvalues of 𝐴. Then, the sum and product of eigenvalues are equal to trace and determinant
of the matrix 𝐴. That is, ∑𝑛𝑖=1 𝜆𝑖 = 𝑇𝑟𝑎𝑐𝑒(𝐴) = ∑𝑛𝑖=1 𝑎𝑖𝑖 and ∏𝑛𝑖=1 𝜆𝑖 = 𝐷𝑒𝑡(𝐴).
Example 2.1 Here is an application as the immediate application of Property-1. Let us take a
2 0 1
matrix, 𝐴 = [3 4 5]. Then, sum of the eigenvalues, ∑𝑛𝑖=1 𝜆𝑖 = 𝑇𝑟𝑎𝑐𝑒(𝐴) = 2 + 4 + 3 =
2 1 3
9 and product of the eigenvalues, ∏𝑛𝑖=1 𝜆𝑖 = 𝐷𝑒𝑡(𝐴) = 2(4 × 3 − 5 × 1) − 0(3 × 3 −
2 × 5) + 1(3 × 1 − 2 × 4) = 14 − 0 − 5 = 9

Property-2 Suppose, 𝐴 = [𝑎𝑖𝑗 ]𝑛×𝑛 be a 𝑛 × 𝑛 square matrix and 𝜆 be an eigenvalue of 𝐴

corresponding to the eigenvector 𝑋. Then, for 𝑘 is an integer, 𝜆𝑘 will be an eigenvalue of
𝐴𝑘 corresponding to the eigenvector 𝑋.
2 0 0
Example 2.2 Suppose, we take 𝐴 = [0 4 0] , a diagonal matrix. Then, its eigenvalues are
0 0 3
2 0 0 2 0 0 4 0 0
2
2, 4 and 3. Also, 𝐴 = [0 4 0] × [0 4 0] = [0 16 0] has the eigenvalues 4, 16 and
0 0 3 0 0 3 0 0 9
9.

Property-3 The eigenvalues of a matrix 𝐴 = [𝑎𝑖𝑗 ]𝑛×𝑛 and its transpose 𝐴𝑇 = [𝑎𝑗𝑖 ]𝑛×𝑛 have
same set of eigenvalues.

Property-4 Suppose, we take two square matrices, 𝐴 = [𝑎𝑖𝑗 ]𝑛×𝑛 and 𝐵 = [𝑏𝑖𝑗 ]𝑛×𝑛 and
furthermore, 𝐴 = [𝑎𝑖𝑗 ]𝑛×𝑛 is non-singular (hence, 𝐴−1 exists), then the matrices 𝐴−1 𝐵 and
𝐵𝐴−1 have same set of eigenvalues.

Property-5 Suppose, 𝐴 = [𝑎𝑖𝑗 ]𝑛×𝑛 be a 𝑛 × 𝑛 square matrix and 𝜆𝑖 , for 𝑖 = 1,2, . . , 𝑛 are
eigenvalues of 𝐴. Then, for a scalar, 𝑘 ≠ 0, 𝑘𝜆𝑖 , (for 𝑖 = 1,2, . . , 𝑛) are eigenvalues of the matrix
𝑘𝐴.

Property-6 Suppose, 𝐴 = [𝑎𝑖𝑗 ]𝑛×𝑛 be a 𝑛 × 𝑛 square matrix and 𝜆 is an eigenvalue of 𝐴.

Then, for a scalar, 𝑘 ≠ 0, 𝜆 + 𝑘 is an eigenvalue of the matrix 𝐴 + 𝑘𝐼, where 𝐼 is the identity
matrix of order 𝑛 × 𝑛 .

Property-7 Suppose, 𝐴 = [𝑎𝑖𝑗 ]𝑛×𝑛 be a 𝑛 × 𝑛 square matrix and 𝜆𝑖 , for 𝑖 = 1,2, . . , 𝑛 are
eigenvalues of 𝐴.Then, for a scalar, 𝑘 ≠ 0, 𝜆𝑖 ± 𝑘 , (for 𝑖 = 1,2, . . , 𝑛) are eigenvalues of the
matrix 𝐴 ± 𝑘𝐼, where 𝐼 is the identity matrix of order 𝑛 × 𝑛.
Property-8 Suppose, 𝐴 = [𝑎𝑖𝑗 ]𝑛×𝑛 be a non-singular 𝑛 × 𝑛 square matrix and 𝜆𝑖 , for 𝑖 =
1,2, . . , 𝑛 are eigenvalues of 𝐴. Then, 𝜆𝑖 −1, for 𝑖 = 1,2, . . , 𝑛 , are the eigenvalues of the inverse
matrix 𝐴−1 .

Property-9 Suppose, 𝐴 = [𝑎𝑖𝑗 ]𝑛×𝑛 be a non-singular 𝑛 × 𝑛 square matrix and 𝜆 is an

𝐷𝑒𝑡 (𝐴)
eigenvalue of 𝐴. Then, is the eigenvalue of the matrix 𝐴𝑑𝑗(𝐴).
𝜆

Proof: We know that, 𝐴𝑑𝑗(𝐴) 𝐴 = 𝐷𝑒𝑡 (𝐴)𝐼 . Taking the eigenvector, we get 𝐴𝑑𝑗(𝐴)(𝐴𝑋) =
𝐷𝑒𝑡 (𝐴) 𝐷𝑒𝑡 (𝐴)
𝐷𝑒𝑡 (𝐴)𝑋. That is, 𝐴𝑑𝑗(𝐴)(𝜆𝑋) = 𝐷𝑒𝑡 (𝐴)𝑋. That is, 𝐴𝑑𝑗(𝐴)𝑋 = 𝑋. So, is the
𝜆 𝜆
eigenvalue of the matrix 𝐴𝑑𝑗(𝐴).

Property-9 Suppose, 𝐴 = [𝑎𝑖𝑗 ]𝑛×𝑛 be a 𝑛 × 𝑛 square matrix and 𝑃 be a non-singular 𝑛 × 𝑛

square matrix. Then, the matrices 𝐴 and 𝑃 −1 𝐴𝑃 have same set of eigenvalues.
Property-10 The eigenvalues of upper triangular, lower triangular, scalar and diagonal
matrices are the elements on the principal diagonal of the corresponding matrices.
2 0 0 4 5 1 2 0 0
Example 2.3 Suppose, we take 𝐴 = [0 4 0] , 𝐵 = [0 10 0], 𝐶 = [6 6 0] and 𝐷 =
0 0 3 0 0 7 2 5 4
5 0 0
[0 5 0]. 𝐴 is a diagonal matrix which have eigenvalues 2, 4 and 3. 𝐵 is an upper triangular
0 0 5
matrix which have eigenvalues 4, 10 and 7. 𝐶 is a lower triangular matrix which have
eigenvalues 2, 6 and 4. 𝐷 is a scalar matrix which have eigenvalues 5,5 and 5.

Property-11 Suppose, 𝐴 = [𝑎𝑖𝑗 ]𝑛×𝑛 be a real symmetric matrix, i.e., 𝐴 = 𝐴𝑇 . Then, all the
eigen values of the matrix 𝐴 are real.
Property-12 The eigen vectors corresponding to the distinct eigenvalues of a matrix are
linearly independent.

Property-13 Suppose, 𝐴 = [𝑎𝑖𝑗 ]𝑛×𝑛 be a real symmetric matrix, i.e., 𝐴 = 𝐴𝑇 . Then, the eigen
vectors corresponding to the distinct eigen values are orthogonal.

Method for finding Eigenvalues and Eigenvectors of matrices

Step 1: Find eigenvalues solving the characteristic equation 𝑑𝑒𝑡 (𝐴 − 𝜆𝐼) = 0. If 𝜆𝑖 , for 𝑖 =
1,2, . . , 𝑛 are roots of the equations, they are called eigenvalues of 𝐴.
Step 2: When the eigenvalues 𝜆𝑖 are distinct, reduce the coefficient matrix in the matrix
equation (𝐴 − 𝜆𝑖 𝐼)𝑋 = 0 to the echelon form. The rank of the echelon matrix should be less
than 𝑛, usually 𝑟 = (𝑛 − 1). So, there are 𝑛 − 𝑟 = 1 linearly independent eigen vectors
corresponding to each distinct eigenvalue.
Step 3: Suppose, two eigen values 𝜆𝑖 = 𝜆𝑖 . Then, reduce the coefficient matrix in the matrix
equation (𝐴 − 𝜆𝑖 𝐼)𝑋 = 0 to the echelon form. The rank of the echelon matrix should be less
than 𝑛, usually 𝑟 = (𝑛 − 2). So, there are 𝑛 − 𝑟 = 2 linearly independent eigen vectors
corresponding to that pair of eigenvalues.

1 6
Example 2.3 Find the eigenvalues and eigen vectors of the matrix 𝐴 = [ ].
1 2
Solution: The characteristic equation is 𝑑𝑒𝑡 (𝐴 − 𝜆𝐼) = 0
1−𝜆 6
That is | | = 0 ⇒ (1 − 𝜆)(2 − 𝜆) − 6 = 0 ⇒ 𝜆2 − 3𝜆 − 4 = 0 ⇒ (𝜆 − 4)(𝜆 +
1 2−𝜆
1) = 0 ⇒ 𝜆 = 4, −1
So, the eigenvalues are 4 and −1.
The eigenvalues are distinct.
𝑥1
Now to find eigenvectors corresponding to the eigenvalue 4, let the eigenvector be 𝑋 = [𝑥 ].
2
1−4 6 𝑥1 0
Then, we have [ ][ ] = [ ]
1 2 − 4 𝑥2 0
−3 6 𝑥1 0
⇒[ ][ ] = [ ]
1 −2 𝑥2 0
−3 6 𝑥1 0 1
⇒[ ][ ] = [ ] , 𝑅2 → 𝑅2 + 3 𝑅1
0 0 𝑥2 0

Therefore, the rank of the coefficient matrix is 𝑟 = 1. Also, the number of unknowns is 𝑛 = 2.
Therefore, there are 𝑛 − 𝑟 = 1 linearly independent eigenvectors corresponding to the
eigenvalue 4. In such case, the above written matrix equation can be written as
−3𝑥1 + 6𝑥2 = 0 ⇒ 𝑥1 = 2𝑥2
2
If we take 𝑥2 = 𝑘, then 𝑥1 = 2𝑘 and hence the eigen vector can be written as 𝑋 = 𝑘 [ ].
1
𝑥1
Now to find eigenvectors corresponding to the eigenvalue −1, let the eigenvector be 𝑋 = [𝑥 ].
2
1+1 6 𝑥1 0
Then, we have, [ ][ ] = [ ]
1 2 + 1 𝑥2 0
2 6 𝑥1 0
⇒[ ][ ] = [ ]
1 3 𝑥2 0
2 6 𝑥1 0 1
⇒[ ] [𝑥 ] = [ ] , 𝑅2 → 𝑅2 + 2 𝑅1
0 0 2 0
Therefore, the rank of the coefficient matrix is 𝑟 = 1. Also, the number of unknowns is 𝑛 = 2.
Therefore, there are 𝑛 − 𝑟 = 1 linearly independent eigenvectors corresponding to the
eigenvalue −1. In such case, the above written matrix equation can be written as
2𝑥1 + 6𝑥2 = 0 ⇒ 𝑥1 = −3𝑥2
−3
If we take 𝑥2 = 𝑘, then 𝑥1 = −3𝑘 and hence the eigen vector can be written as 𝑋 = 𝑘 [ ].
1
 1 6 1 6
Exercise 2.1 Find the eigenvalues and eigen vectors of the matrix 𝐴 = [ ] and 𝐵 = [ ].
0 2 6 2
Example 2.4 Find the eigenvalues of the matrix, 3𝐴3 + 5𝐴2 − 6𝐴 + 2𝐼, where 𝐴 =
1 2 −3
[0 3 2 ].
0 0 −2
1 2 −3
Solution: Since, 𝐴 = [0 3 2 ] is an upper triangular matrix, the elements in its principal
0 0 −2
diagonal are the eigenvalues of this matrix. Therefore, the eigenvalues are 1, 3 and -2.
When 1 is the eigenvalue of 𝐴 , the eigenvalue of 3𝐴3 + 5𝐴2 − 6𝐴 + 2𝐼 will be 3 × 13 +
5 × 12 − 6 × 1 + 2 × 1 = 4.
When 3 is the eigenvalue of 𝐴 , the eigenvalue of 3𝐴3 + 5𝐴2 − 6𝐴 + 2𝐼 will be 3 × 33 +
5 × 32 − 6 × 3 + 2 × 1 = 110.
When −2 is the eigenvalue of 𝐴 , the eigenvalue of 3𝐴3 + 5𝐴2 − 6𝐴 + 2𝐼 will be 3 × (−2)3 +
5 × (−2)2 − 6 × (−2) + 2 × 1 = 10.
Then, eigenvalues of the matrix 3𝐴3 + 5𝐴2 − 6𝐴 + 2𝐼 are 4, 110 and 10.
1 2 −3
3 −1
Exercise 2.2 Find the eigenvalues of the matrices 𝐴 and 𝐴 , where 𝐴 = [0 0 5 ].
0 0 −2
1 0 −1
Exercise 2.3 Find the eigenvalues of the matrix 𝐴−1, where 𝐴 = [1 2 1 ].
2 2 3
1
Exercise 2.4 Find the eigenvalues and eigen vectors of the matrix 𝐵 = 2𝐴2 − 2 𝐴 + 3𝐼, where
8 −4
𝐴=[ ].
2 2
1 0 −1
Exercise 2.5 Find product and sum the eigenvalues of the matrix 𝐴 = [1 2 1 ].
2 2 3
1 1 3
−1
Exercise 2.6 Find the eigenvalues of the matrix 𝐴 , where 𝐴 = [1 5 1].
3 1 1
8 −6 2
Example 2.5 Find the eigenvalues and eigen vectors of the matrix 𝐴 = [−6 7 −4].
2 −4 3
8 −6 2
Solution: The characteristic equation of 𝐴 = [−6 7 −4] can be written as
2 −4 3
𝑑𝑒𝑡 (𝐴 − 𝜆𝐼) = 0
8−𝜆 −6 2
⟹ | −6 7−𝜆 −4 | = 0
2 −4 3−𝜆
 ⟹ −𝜆3 + 18𝜆2 − 45𝜆 = 0
⟹ 𝜆(𝜆 − 15)(𝜆 − 3) = 0
⟹ 𝜆 = 0, 15, 3
So, the eigenvalues are 0, 15, 3.
𝑥1
Now to find eigenvectors corresponding to the eigenvalue 0, let the eigenvector be 𝑋 = [𝑥2 ].
𝑥3
8−0 −6 2 𝑥1 0
[ −6 7−0 𝑥
−4 ] [ 2 ] = [0]
2 −4 3 − 0 𝑥3 0
8 −6 2 𝑥1 0
⇒ [0 𝑥
10 −10] [ 2 ] = [0], 𝑅2 → 4𝑅2 + 3𝑅1 𝑎𝑛𝑑 𝑅3 → 4𝑅3 − 𝑅1
0 −10 10 𝑥3 0
8 −6 2 𝑥1 0
⇒ [0 10 −10] [𝑥 2 ] = [0], 𝑅3 → 𝑅3 + 𝑅2
0 0 0 𝑥 3 0
Clearly, the rank of the coefficient matrix in the above system is 𝑟 = 2 < 3 = 𝑛, the number
of unknown. So, there are 𝑛 − 𝑟 = 3 − 2 = 1 linearly independent solutions. Then solution
can be derived by solving the system of equations
8𝑥1 − 6𝑥2 + 2𝑥3 = 0 (2.3)
10𝑥2 − 10𝑥3 = 0 (2.4)
𝑘
From Equation (2.4), 𝑥2 = 𝑥3 . Let, 𝑥2 = 𝑥3 = 𝑘. Then, from Equation (2.3), 𝑥1 = 2.
𝑘
2 𝑘
1
Therefore, the corresponding eigen vector can be written as 𝑋 = [𝑘] = 2 [2].
𝑘 2
𝑥1
Now to find eigenvectors corresponding to the eigenvalue 15, let the eigenvector be 𝑋 = [𝑥2 ].
𝑥3
8 − 15 −6 2 𝑥1 0
[ −6 7 − 15 𝑥
−4 ] [ 2 ] = [0]
2 −4 3 − 15 𝑥3 0
−7 −6 2 𝑥1 0
𝑥
⇒ [−6 −8 −4 ] [ 2 ] = [0]
2 −4 −12 𝑥3 0

−7 −6 2 𝑥1 0
⇒ [ 0 −20 −40] [𝑥2 ] = [0], 𝑅2 → 7𝑅2 − 6𝑅1 𝑎𝑛𝑑 𝑅3 → 7𝑅3 + 2𝑅1
0 −40 −80 𝑥3 0
 −7 −6 2 𝑥1 0
⇒ [ 0 −20 −40] [𝑥 2 ] = [ 0], 𝑅3 → 𝑅3 − 2𝑅2
0 0 0 𝑥 3 0
Clearly, the rank of the coefficient matrix in the above system is 𝑟 = 2 < 3 = 𝑛, the number
of unknown. So, there are 𝑛 − 𝑟 = 3 − 2 = 1 linearly independent solutions. Then solution
can be derived by solving the system of equations
−7𝑥1 − 6𝑥2 + 2𝑥3 = 0 (2.5)
−20𝑥2 − 40𝑥3 = 0 (2.6)

From Equation (2.6), 𝑥2 = −2𝑥3 . Let 𝑥3 = 𝑘, then 𝑥2 = −2𝑘 and from Equation (2.5), 𝑥1 =
2
2𝑘. Hence, the corresponding eigen vector can be written as 𝑋 = 𝑘 [−2].
1
𝑥1
𝑥
Now to find eigenvectors corresponding to the eigenvalue 3, let the eigenvector be 𝑋 = [ 2 ].
𝑥3
8−3 −62 𝑥1 0
[ −6 7−3 𝑥
−4 ] [ 2 ] = [0]
2 3 − 3 𝑥3
−4 0
5 −6 2 𝑥1 0
⇒ [−6 4 −4] [𝑥2 ] = [0]
2 −4 0 𝑥3 0

5 −6 2 𝑥1 0
⇒ [0 𝑥
−16 −8] [ ] = [0],
2 𝑅2 → 5𝑅2 + 6𝑅1 𝑎𝑛𝑑 𝑅3 → 5𝑅3 − 2𝑅1
0 −8 −4 𝑥3 0
5 −6 2 𝑥1 0
⇒ [0 −16 −8] [𝑥2 ] = [0], 𝑅3 → 2𝑅3 − 𝑅2
0 0 0 𝑥3 0
Clearly, the rank of the coefficient matrix in the above system is 𝑟 = 2 < 3 = 𝑛, the number
of unknown. So, there are 𝑛 − 𝑟 = 3 − 2 = 1 linearly independent solutions. Then solution
can be derived by solving the system of equations
5𝑥1 − 6𝑥2 + 2𝑥3 = 0 (2.7)
−16𝑥2 − 8𝑥3 = 0 (2.8)
1 𝑘
From Equation (2.4), 𝑥2 = − 2 𝑥3 . Let, 𝑥3 = 𝑘. Then, 𝑥2 = − 2 and from Equation (2.7),
𝑘
−2
𝑥1 = −𝑘. Therefore, the corresponding eigen vector can be written as 𝑋 = 2 [−1].
2
Definition 2.3 (Orthogonal vector) Two vectors 𝑋1 and 𝑋2 are said to orthogonal, when
𝑋1 𝑇 𝑋2 = 𝑋2 𝑇 𝑋1 = 0, where 𝑋1 𝑇 and 𝑋2 𝑇 are transposes of the vectors 𝑋1 and 𝑋2, respectively.
 1 0 0
Example 2.6 Let us consider two vectors 𝑋1 = [ ] and 𝑋2 = [ ]. Then, 𝑋1 𝑇 𝑋2 = [1 0] [ ] =
0 1 1
𝑇 1
0 and 𝑋2 𝑋1 = [0 1] [ ] = 0. Therefore, 𝑋1 and 𝑋2 are orthogonal vectors.
0
1 2 2
Example 2.7 Let us consider three vectors 𝑋1 = [2], 𝑋2 = [ 1 ] and 𝑋2 = [−2] Then,
2 −2 1
2
𝑋1 𝑇 𝑋2 = [1 2 2] [ 1 ] = 2 + 2 − 4 = 0
−2
2
𝑋2 𝑇 𝑋3 = [2 1 − 2] [−2] = 4 − 2 − 2 = 0 and
1
1
𝑋3 𝑇 𝑋1 = [2 − 2 1] [2] = 2 − 4 + 2 = 0. Therefore, 𝑋1 , 𝑋2 and 𝑋3 are orthogonal vectors.
2
Remark 2.1 If two eigenvalues of a matrix are equal, we cannot conclude whether the
corresponding eigen vectors are linearly independent or not.
Definition 2.4 (Linearly independent vectors) We consider a set of non-zero vectors
{𝑋1 , 𝑋2 , … , 𝑋𝑛 } such that 𝑎1 𝑋1 + 𝑎2 𝑋2 + … + 𝑎𝑛 𝑋𝑛 = 0. If there exists only the trivial
solution 𝑎1 = 𝑎2 =. . = 𝑎𝑛 = 0 of the equation, then the given vectors 𝑋1 , 𝑋2 , … , 𝑋𝑛 are called
the linearly independent vectors. If there exists at least one non-zero 𝑎𝑖 such that the equation
holds, then the vectors are called linearly dependent vectors.
1 0 0
Example 2.8 Let us consider three vectors 𝑋1 = [0], 𝑋2 = [1] and 𝑋2 = [0]. Then, we
0 0 1
consider the equation 𝑎1 𝑋1 + 𝑎2 𝑋2 + 𝑎3 𝑋3 = 0
1 0 0 0
This implies 𝑎1 [0] + 𝑎2 [1] + 𝑎3 [0] = [0]
0 0 1 0
1 0 0 𝑎1 0
𝑎
That is [0 1 0] [ ] = [0].
2
0 0 1 𝑎3 0
Here the rank of the coefficient matrix is 𝑟 = 3 = 𝑛, the number of unknowns. So, the system
1 0 0
has only the trivial solution 𝑎1 = 𝑎2 = 𝑎3 = 0. Therefore, 𝑋1 = [0], 𝑋2 = [1] and 𝑋2 = [0]
0 0 1
are linearly independent vectors.
1 2 2
Exercise 2.7 Show that, three vectors 𝑋1 = [2], 𝑋2 = [ 1 ] and 𝑋2 = [−2] are linearly
2 −2 1
independent vectors.
(Hint: Try to make an upper triangular matrix from the coefficient matrix).
 3 −1 1
Example 2.9 Verify that the sum of the eigenvalues of the matrix 𝐴 = [−1 5 −1] is equal
1 −1 3
to the 𝑇𝑟𝑎𝑐𝑒(𝐴). Also, find the corresponding eigen vectors.
Solution: We have the characteristic equation
𝑑𝑒𝑡 (𝐴 − 𝜆𝐼) = 0
3−𝜆 −1 1
⟹ | −1 5−𝜆 −1 | = 0
1 −1 3−𝜆
⟹ −𝜆3 + 11𝜆2 − 36𝜆 + 36 = 0
⟹ (𝜆 − 2)(𝜆 − 3)(𝜆 − 6) = 0
⟹ 𝜆 = 2, 3, 6
Therefore, eigenvalues are 2, 3 and 6 and hence their sum is 11. Again, 𝑇𝑟𝑎𝑐𝑒(𝐴) is the sum
of the elements in the principal diagonal of the matrix and is equal to 11. Therefore, the
statement is verified.
𝑥1
𝑥
Next, to find eigenvectors corresponding to the eigenvalue 2, let the eigenvector be 𝑋 = [ 2 ].
𝑥3
3−2 −11 𝑥1 0
[ −1 −1 ] [𝑥2 ] = [0]
5−2
1 3 − 2 𝑥3
−1 0
1 −1 1 𝑥1 0
⇒ [−1 3 −1] [𝑥2 ] = [0]
1 −1 1 𝑥3 0

1 −1 1 𝑥1 0
⇒ [0 𝑥
2 0] [ ] = [0],
2 𝑅2 → 𝑅2 + 𝑅1 𝑎𝑛𝑑 𝑅3 → 𝑅3 − 𝑅1
0 0 0 𝑥3 0
Clearly, the rank of the coefficient matrix in the above system is 𝑟 = 2 < 3 = 𝑛, the number
of unknown. So, there are 𝑛 − 𝑟 = 3 − 2 = 1 linearly independent solutions. Then solution
can be derived by solving the system of equations
𝑥1 − 𝑥2 + 𝑥3 = 0 (2.9)
2𝑥2 = 0 (2.10)
From Equation (2.10), 𝑥2 = 0. Let, 𝑥3 = 𝑘. Then, from Equation (2.9), 𝑥1 = −𝑘. Therefore,
−1
the corresponding eigen vector can be written as 𝑋 = 𝑘 [ 0 ].
1
𝑥1
𝑥
Next, to find eigenvectors corresponding to the eigenvalue 3, let the eigenvector be 𝑋 = [ 2 ].
𝑥3
 3−3 −1
1 𝑥1 0
[ −1 5−3
−1 ] [ 𝑥2 ] = [0]
1 −1
3−3 3 𝑥 0
0 −1 1 𝑥1 0
⇒ [−1 2 −1] [𝑥2 ] = [0]
1 −1 0 𝑥3 0

1 −1 0 𝑥1 0
⇒ [−1 2 −1] [𝑥2 ] = [0], 𝑅3 ↔ 𝑅1
0 −1 1 𝑥3 0
1 −1 0 𝑥1 0
⇒ [0 1 −1] [𝑥2 ] = [0], 𝑅2 → 𝑅2 + 𝑅1
0 −1 1 𝑥3 0
1 −1 0 𝑥1 0
⇒ [0 1 −1] [𝑥2 ] = [0] , 𝑅3 → 𝑅3 + 𝑅2
0 0 0 𝑥3 0

Clearly, the rank of the coefficient matrix in the above system is 𝑟 = 2 < 3 = 𝑛, the number
of unknown. So, there are 𝑛 − 𝑟 = 3 − 2 = 1 linearly independent solutions. Then solution
can be derived by solving the system of equations
𝑥1 − 𝑥2 = 0 (2.11)
𝑥2 − 𝑥3 = 0 (2.12)
Let us take 𝑥3 = 𝑘. Then, from Equation (2.11) and (2.12), we get 𝑥2 = 𝑥1 = 𝑘. Therefore,
1
the corresponding eigen vector can be written as 𝑋 = 𝑘 [1].
1
𝑥1
Next, to find eigenvectors corresponding to the eigenvalue 6, let the eigenvector be 𝑋 = [𝑥2 ].
𝑥3
3−6 −11 𝑥1 0
[ −1 5−6 𝑥
−1 ] [ 2 ] = [0]
1 3 − 6 𝑥3
−1 0
−3 −1 1 𝑥1 0
⇒ [−1 −1 −1] [𝑥2 ] = [0]
1 −1 −3 𝑥3 0

−3 −1 1 𝑥1 0
⇒ [ 0 −2 −4] [𝑥2 ] = [0], 𝑅2 → 3𝑅2 − 𝑅1
0 −4 −8 𝑥3 0
−3 −1 1 𝑥1 0
⇒ [ 0 −2 −4] [𝑥2 ] = [0], 𝑅3 → 𝑅3 − 2𝑅2
0 0 0 𝑥3 0
Clearly, the rank of the coefficient matrix in the above system is 𝑟 = 2 < 3 = 𝑛, the number
of unknown. So, there are 𝑛 − 𝑟 = 3 − 2 = 1 linearly independent solutions. Then solution
can be derived by solving the system of equations
−3𝑥1 − 𝑥2 + 𝑥3 = 0 (2.13)
−2𝑥2 − 4𝑥3 = 0 (2.14)
Let us take 𝑥3 = 𝑘. Then, from Equation (2.14), we get 𝑥2 = −2𝑘 and from Equation (2.13),
1
we get, 𝑥1 = 𝑘. Therefore, the corresponding eigen vector can be written as 𝑋 = 𝑘 [−2].
1
Exercise 2.8 Find the eigenvalues and the corresponding eigen vectors for the matrix 𝐴 =
6 −2 2 1 1 3 3 1 4
[−2 3 −1], 𝐵 = [1 5 1] and 𝐶 = [0 2 6]
2 −1 3 3 1 1 0 0 5
Exercise 2.9 Find the eigenvalues and the corresponding eigen vectors for the matrix 𝐴 =
1 1 1 −2 2 −3 2 2 0 5 −2 0
[1 1 1], 𝐵 = [ 2 1 −6 ], 𝐶 = [ 2 5 0 ] and 𝐷 = [−2 6 2].
1 1 1 −1 −2 0 0 0 3 0 2 7
Definition 2.5 (Similar matrices) Two matrices 𝑆 and 𝑇 are said to be similar to each other
when there exists a non-singular matrix 𝑃 such that 𝑇 = 𝑃−1 𝑆𝑃.
In this context, it is to be noted that a matrix is said to be non-singular if it has non-zero
determinant and thus its inverse exists.
Definition 2.4 (Diagonalizable matrix) A square matrix is said be diagonalizable if it is
similar to diagonal matrix. So, a square matrix 𝑆 is said to be diagonalizable, if there exists a
non-singular matrix 𝑃 and a diagonal matrix 𝐷 = 𝐷𝑖𝑎𝑔(𝑑11 , 𝑑22 , … . 𝑑𝑛𝑛 ) such that 𝐷 =
𝑃−1 𝑆𝑃.
Remark 2.2 The eigenvalues of the diagonal matrix 𝐷 are its diagonal entries 𝑑11 , 𝑑22 , … . 𝑑𝑛𝑛 .
The non-singular matrix 𝑃 can be formed by corresponding eigen vectors.
Remark 2.3 In Definition 2.4, the non-singular matrix 𝑃 is makes the matrix 𝑆 to be diagonal.
Here, 𝑃 is called the modal matrix and resulting diagonal matrix 𝐷 = 𝐷𝑖𝑎𝑔(𝑑11 , 𝑑22 , … . 𝑑𝑛𝑛 )
is called the spectral matrix.
Remark 2.4 The transformation of the matrix 𝑆 to 𝑃 −1 𝑆𝑃 is called the similarity
transformation.
1 0
Example 2.10 Diagonalize the square matrix 𝐴 = [ ].
2 −1
1 0
Solution: The characteristic equation of the given matrix 𝐴 = [ ] can be written as
2 −1
𝑑𝑒𝑡 (𝐴 − 𝜆𝐼) = 0
1−𝜆 0
⟹| |=0
2 −1 − 𝜆
 ⟹ 𝜆 = ±1
Therefore, the eigenvalues are 1 and -1.
𝑥1
For the eigenvalue 1, let the eigen vector 𝑋 = [𝑥 ]. Then we have the matrix equation
2

1−1 0 𝑥1 0
[ ] [𝑥 ] = [ ]
2 −1 − 1 2 0
0 0 𝑥1 0
⇒[ ] [𝑥 ] = [ ]
2 −2 2 0
Clearly, the rank of the coefficient matrix in the above system is 𝑟 = 1 < 2 = 𝑛, the number
of unknown. So, there are 𝑛 − 𝑟 = 2 − 1 = 1 linearly independent solutions. Then solution
can be derived by solving the system of equation
2𝑥1 − 2𝑥2 = 0 (2.15)
From Equation (2.15), we get 𝑥1 = 𝑥2 = 𝑘 and hence the corresponding eigen vector is
1
𝑋 = 𝑘[ ]
1
𝑥1
For the eigenvalue −1, let the eigen vector 𝑋 = [𝑥 ]. Then we have the matrix equation
2

1+1 0 𝑥1 0
[ ] [𝑥 ] = [ ]
2 −1 + 1 2 0
2 0 𝑥1 0
⇒[ ] [𝑥 ] = [ ]
2 0 2 0
Clearly, the rank of the coefficient matrix in the above system is 𝑟 = 1 < 2 = 𝑛, the number
of unknown. So, there are 𝑛 − 𝑟 = 2 − 1 = 1 linearly independent solutions. Then solution
can be derived by solving the system of equation
2𝑥1 = 0 (2.16)
From Equation (2.16), we get 𝑥1 = 0. Here, 𝑥2 = 𝑘 is arbitrary and hence the corresponding
eigen vector is
0
𝑋 = 𝑘[ ]
1
Using the eigen vectors, the matrix 𝑃 can be formed as follows:
1 0
𝑃=[ ]
1 1
Since, Det(𝑃) = 1 ≠ 0, we can say that 𝑃 −1 and is given by
1 0
𝐴𝑑𝑗(𝑃) [−1 1] 1 0
𝑃−1 = = =[ ]
𝐷𝑒𝑡 (𝑃) 1 −1 1
1 0 1 0 1 0 1+0 0+0 1 0
Therefore, 𝑃−1 𝐴𝑃 = [ ][ ][ ]=[ ][ ]
−1 1 2 −1 1 1 −1 + 2 0−1 1 1
 1 0 1 0 1+0 0+0 1 0
=[ ][ ]=[ ]=[ ] = 𝐷, a diagonal matrix.
1 −1 1 1 1 − 1 0 − 1 0 −1
Hence, diagonalization of the given matrix is completed.
1 1 3
Example 2.11 Diagonalize the square matrix 𝐴 = [1 5 1].
3 1 1
1 1 3
Solution: The characteristic equation of the given matrix 𝐴 = [1 5 1] can be written as
3 1 1
𝑑𝑒𝑡 (𝐴 − 𝜆𝐼) = 0
1−𝜆 1 3
⟹| 1 5−𝜆 1 |=0
3 1 1−𝜆
⟹ −𝜆3 + 7𝜆2 − 36 = 0
⟹ (𝜆 + 2)(𝜆 − 3)(𝜆 − 6) = 0
⟹ 𝜆 = −2, 3 and 6
Therefore, eigenvalues are −2, 3 and 6.
𝑥1
𝑥
Next, to find eigenvectors corresponding to the eigenvalue −2, let the eigenvector be 𝑋 = [ 2 ].
𝑥3
1+2 1 3 𝑥1 0
[ 1 5+2 𝑥
1 ] [ 2 ] = [0]
3 1 1 + 2 𝑥3 0
3 1 3 𝑥1 0
⇒ [1 7 1] [𝑥2 ] = [0]
3 1 3 𝑥3 0

3 1 3 𝑥1 0
⇒ [0 𝑥
20 0] [ ] = [0],
2 𝑅2 → 3𝑅2 − 𝑅1 𝑎𝑛𝑑 𝑅3 → 𝑅3 − 𝑅1
0 0 0 𝑥3 0
Clearly, the rank of the coefficient matrix in the above system is 𝑟 = 2 < 3 = 𝑛, the number
of unknown. So, there are 𝑛 − 𝑟 = 3 − 2 = 1 linearly independent solutions. Then solution
can be derived by solving the system of equations
3𝑥1 + 𝑥2 + 3𝑥3 = 0 (2.17)
20𝑥2 = 0 (2.18)
From Equation (2.18), we get 𝑥2 = 0. Let, 𝑥3 = 𝑘. Then, from Equation (2.17), 𝑥1 = −𝑘.
−1
Hence, the corresponding eigen vector will be 𝑋 = 𝑘 [ 0 ].
1
 𝑥1
𝑥
Next, to find eigenvectors corresponding to the eigenvalue 3, let the eigenvector be 𝑋 = [ 2 ].
𝑥3
1−3 1 3 𝑥1 0
[ 1 5−3 𝑥
1 ] [ 2 ] = [0]
3 1 1 − 3 𝑥3 0
−2 1 3 𝑥1 0
⇒[ 1 2 1 ] [ 𝑥2 ] = [0]
3 1 −2 3 𝑥 0
−2 1 3 𝑥1 0
⇒[ 0 5 5] [𝑥2 ] = [0], 𝑅2 → 2𝑅2 + 𝑅1 𝑎𝑛𝑑 𝑅3 → 2𝑅3 + 3𝑅1
0 5 5 𝑥3 0
−2 1 3 𝑥1 0
⇒[ 0 5 5] [𝑥2 ] = [0] 𝑅3 → 𝑅3 − 𝑅2
0 0 0 𝑥3 0
Clearly, the rank of the coefficient matrix in the above system is 𝑟 = 2 < 3 = 𝑛, the number
of unknown. So, there are 𝑛 − 𝑟 = 3 − 2 = 1 linearly independent solutions. Then solution
can be derived by solving the system of equations
−2𝑥1 + 𝑥2 + 3𝑥3 = 0 (2.19)
5𝑥2 + 5𝑥3 = 0 (2.20)
From Equation (2.20), we get 𝑥2 = −𝑥3 . Let, 𝑥3 = 𝑘. Then, from Equation (2.20), 𝑥2 = −𝑘
1
and from Equation (2.19), 𝑥1 = 𝑘. Hence, the corresponding eigen vector will be 𝑋 = 𝑘 [−1].
1
𝑥1
𝑥
Next, to find eigenvectors corresponding to the eigenvalue 6, let the eigenvector be 𝑋 = [ 2 ].
𝑥3
1−6 1 3 𝑥1 0
[ 1 5−6 𝑥
1 ] [ 2 ] = [0]
3 11 − 6 𝑥3 0
−5 1 3 𝑥1 0
⇒ [ 1 −1 1 ] [𝑥2 ] = [0]
3 1 −5 𝑥3 0
−5 1 3 𝑥1 0
⇒ [ 0 −4 𝑥
8 ] [ 2 ] = [0], 𝑅2 → 5𝑅2 + 𝑅1 𝑎𝑛𝑑 𝑅3 → 5𝑅3 + 3𝑅1
0 8 −16 𝑥3 0
−5 1 3 𝑥1 0
⇒ [ 0 −4 8] [𝑥2 ] = [0] 𝑅3 → 𝑅3 ∓ 2
0 0 0 𝑥3 0
Clearly, the rank of the coefficient matrix in the above system is 𝑟 = 2 < 3 = 𝑛, the number
of unknown. So, there are 𝑛 − 𝑟 = 3 − 2 = 1 linearly independent solutions. Then solution
can be derived by solving the system of equations
−5𝑥1 + 𝑥2 + 3𝑥3 = 0 (2.21)
 −4𝑥2 + 8𝑥3 = 0 (2.22)
From Equation (2.22), we get 𝑥2 = 2𝑥3 . Let, 𝑥3 = 𝑘. Then, from Equation (2.22), 𝑥2 = 2𝑘 and
1
from Equation (2.21), 𝑥1 = 𝑘. Hence, the corresponding eigen vector will be 𝑋 = 𝑘 [2].
1
Using the eigen vector the modal matrix can be obtained as
−1 1 1
𝑃 = [ 0 −1 2]
1 1 1
Since, Det(𝑃) = −1(−1 − 2) − 1(0 − 2) + 1(0 + 1) = 6 ≠ 0, we can say that 𝑃−1 and is
given by

𝐴𝑑𝑗(𝑃) 1 −3 0 3
𝑃−1 = = [ 2 −2 2]
𝐷𝑒𝑡 (𝑃) 6
1 2 1
−3 0 3 1 1 3 −1 1 1
1
Therefore, 𝑃−1 𝐴𝑃 = 6 [ 2 −2 2] [1 5 1] [ 0 −1 2]
1 2 1 3 1 1 1 1 1
1 −3 0 3 2 3 6
= [ 2 −2 2] [ 0 −3 12]
6
1 2 1 −2 3 6
−12 0 0 −2 0 0
1
= 6[ 0 18 0 ] = [ 0 3 0] = 𝐷, a diagonal matrix.
0 0 36 0 0 6
Hence, diagonalization of the given matrix is completed.
1 0 0
Exercise 2.10 Diagonalize the square matrix 𝐴 = [0 3 −1].
0 −1 3
Exercise 2.11 Find a non-singular matrix 𝑃 which can diagonalize the square matrix 𝐴 =
1 0 −1
[1 2 1 ].
2 2 3
Exercise 2.12 Find a non-singular matrix 𝑃 which can diagonalize the square matrix 𝐴 =
8 −6 2
[−6 7 −4].
2 −4 3

Theorem 2.1 (Cayley-Hamilton) Every square matrix satisfies of its characteristic equation.
2 4
Example 2.12 Find the matrix 𝐴3 and 𝐴−1 corresponding to the square matrix 𝐴 = [ ],
1 1
using the statement of Cayley-Hamilton theorem.
2 4
Solution: The characteristic equation of the given matrix 𝐴 = [ ] can be written as
1 1
 𝑑𝑒𝑡 (𝐴 − 𝜆𝐼) = 0
2−𝜆 4
⟹| |=0
1 1−𝜆
⟹ 𝜆2 − 3𝜆 − 2 = 0 (2.23)
2 4
This is the characteristic equation of the given matrix 𝐴 = [ ]. Therefore, according the
1 1
Cayley-Hamilton theorem,
𝐴2 − 3𝐴 − 2𝐼 = 𝑂 (2.24)
1 0 0 0
In Equation (2.24), 𝐼 = [ ], the identity matrix and 𝑂 = [ ] is the zero matrix. From
0 1 0 0
Equation (2.24) we get
𝐴2 = 3𝐴 + 2𝐼 (2.25)
Multiplying both side by the matrix 𝐴, we get
𝐴3 = 3𝐴2 + 2𝐴
2 4 2 4 2 4
= 3[ ][ ]+2[ ]
1 1 1 1 1 1
8 12 4 8
= 3[ ]+[ ]
3 5 2 2
24 36 4 8
=[ ]+[ ]
9 15 2 2
28 44
=[ ]
11 17
Again, from Equation (2.24), we can write
1 1
(𝐴 − 3𝐼)𝐴 = 𝐴 (𝐴 − 3𝐼) = 𝐼
2 2
1
Therefore, 𝐴−1 = 2 (𝐴 − 3𝐼)

1 2 4 1 0
= ([ ]− 3[ ])
2 1 1 0 1
1 −1 4
= [ ]
2 1 −2

1 2
Exercise 2.13 Verify the state of the Cayley-Hamilton theorem for the matrix 𝐴 = [ ].
2 −1
3 1 1
−1
Example 2.13 Find the matrix 𝐴 corresponding to the square matrix 𝐴 = [−1 5 −1],
1 −1 3
using the statement of Cayley-Hamilton theorem.
 3 1 1
Solution: The characteristic equation of the given matrix 𝐴 = [−1 5 −1] can be written
1 −1 3
as
𝑑𝑒𝑡 (𝐴 − 𝜆𝐼) = 0
3−𝜆 1 1
⟹ | −1 5−𝜆 −1 | = 0
1 −1 3−𝜆
⟹ 𝜆3 − 11𝜆2 + 38𝜆 − 40 = 0 (2.26)
Therefore, the characteristic equation of the given matrix is 𝜆3 − 11𝜆2 + 38𝜆 − 40 = 0.
According to the Cayley-Hamilton theorem,
𝐴3 − 11𝐴2 + 38𝐴 − 40𝐼 = 𝑂 (2.27)

1 0 0 0 0 0
In Equation (2.27), 𝐼 = [0 1 0], the identity matrix and 𝑂 = [0 0 0] is the zero matrix.
0 0 1 0 0 0
From Equation (2.27) we get,
1 2 1 2
𝐴 (𝐴 − 11𝐴 + 38𝐼) = (𝐴 − 11𝐴 + 38𝐼)𝐴 = 𝐼
40 40
Hence,
1 2
𝐴−1 = (𝐴 − 11𝐴 + 38𝐼)
40
1 3 1 1 3 1 1 3 1 1 1 0 0
= ([−1 5 −1] [−1 5 −1] − 11 [−1 5 −1] + 38 [0 1 0])
40
1 −1 3 1 −1 3 1 −1 3 0 0 1
1 9 7 5 33 11 11 38 0 0
= ([−9 25 −9] − [−11 55 −11] + [ 0 38 0 ])
40
7 −7 11 11 −11 33 0 0 38
1 14 −4 −6
= [2 8 2]
40
−4 4 16
1 1 2
Example 2.14 Find the matrix 𝐴3 corresponding to the square matrix 𝐴 = [3 1 1], using
2 3 1
the statement of Cayley-Hamilton theorem.
1 0 3
−1
Example 2.15 Find the matrix 𝐴 corresponding to the square matrix 𝐴 = [2 1 −1],
1 −1 1
after verifying the statement of Cayley-Hamilton theorem.
Example 2.16 Find the matrices 𝐴−1 and 𝐴4 corresponding to the square matrix 𝐴 =
1 2 3
[2 4 5], using the statement of Cayley-Hamilton theorem.
3 5 6
Example 2.17 Verify the statement of Cayley-Hamilton theorem for the square matrix 𝐴 =
2 2 1
[1 3 1].
1 2 2