MATHEMATICS-II

UNIT-1 : APPLICATIONS OF MATRICES

ii
Chapter 2
Eigenvalues, Eigenvectors

2.1 Tutorial : Eigenvalues and Eigenvectors

Definition

Let A be an n × n matrix. A scalar λ is called an eigenvalue of A if there exists a nonzero vector

x in Rn such that
Ax = λx.
This vector x is called an eigenvector of A corresponding to λ.

Computation of Eigenvalues

Let A be an n × n matrix. Then the eigenvalues of A can be obtained by solving the equation
det(A − λI) = 0.
This equation is called the characteristic equation of A. The expansion of det(A − λI) is a
polynomial p(λ) of degree n, called the characteristic polynomial of A. Thus every n × n matrix
has exactly n eigenvalues. The eigenvalues may be real or may be complex. In the present book,
we shall confine ourself to real eigenvalues.

Computation of Eigenvectors

Let A be an n × n matrix and λ be an eigenvalue of A. Then the eigenvectors of A corresponding

to λ are the nonzero vectors satisfying Ax = λx. Thus all such eigenvectors can be obtained by
finding the solution space of the system (A−λI)x = 0. This solution space is called the eigenspace
of A corresponding to λ.

Theorems on Eigenvalues and Eigenvectors

(1) If A is a 2 × 2 matrix, then the characteristic equation is given by

λ2 − tr(A)λ + det(A) = 0.

(2) If A is a 3 × 3 matrix, then the characteristic equation is given by

λ3 − tr(A)λ2 + (M11 + M22 + M33 )λ − det(A) = 0,
where M11 , M22 , M33 are the minors of the diagonal entries in A.

1
 2 Chapter 2 Eigenvalues, Eigenvectors

(3) If A is an n × n triangular matrix, then the eigenvalues of A are the diagonal entries of A.
(4) If λ is an eigenvalue of a matrix A and x is a corresponding eigenvector, then λk is an eigenvalue
of Ak and x is a corresponding eigenvector.
1
(5) If λ is an eigenvalue of a matrix A and x is a corresponding eigenvector, then λ
is an eigenvalue
of A−1 and x is a corresponding eigenvector.
(6) If λ is an eigenvalue of a matrix A and x is a corresponding eigenvector, then λ − k is an
eigenvalue of A − kI and x is a corresponding eigenvector.
(7) If λ is an eigenvalue of a matrix A and x is a corresponding eigenvector, then kλ is an
eigenvalue of kA and x is a corresponding eigenvector.
(8) A square matrix A is invertible if and only if λ = 0 is not an eigenvalue of A.
(9) The constant term of the characteristic polynomial of A is given by (−1)n det(A).
(10) If A is a square matrix, then A and AT have the same eigenvalues but may not have the same
eigenvectors.

Solved Examples

Example 2.1.1. Find the eigenvalues and eigenvectors of the matrix

 
4 −1
A=
2 1
Solution. The characteristic equation of A is
λ2 − tr(A)λ + det(A) = 0
⇒ λ2 − 5λ + 6 = 0
⇒ (λ − 2)(λ − 3) = 0
Thus the eigenvalues of A are λ = 2, 3. To find the eigenvectors x corresponding to λ, we have to
solve the system (A − λI)x = 0, i.e.,
    
4 − λ −1 x1 0
= (2.1.1)
2 1−λ x2 0
• For λ = 2 in (2.1.1), we get
    
2 −1 x1 0
=
2 −1 x2 0
The equations of this system are
2x1 − x2 = 0
2x1 − x2 = 0
Solving, we get
s
x1 = , x2 = s (s ∈ R).
2
Thus the eigenvectors corresponding to λ = 2 are
  " t # " 1 #
x1 2 2
x= = =s (s ∈ R).
x2 s 1
Tutorial 2.1 Eigenvalues and Eigenvectors 3

• For λ = 3 in (2.1.1), we get

    
1 −1 x1 0
=
2 −2 x2 0
The equations of this system are
x1 − x2 = 0
2x1 − 2x2 = 0
Solving, we get
x1 = t, x2 = t (t ∈ R).
Thus the eigenvectors corresponding to λ = 3 are
     
x1 t 1
x= = =t (t ∈ R). 
x2 t 1

Example 2.1.2. Find the eigenvalues and eigenvectors of the matrix

 
0 0 −2
A= 1 2 1 
1 0 3
Solution. The characteristic equation of A is
λ3 − tr(A)λ2 + (M11 + M22 + M33 )λ − det(A) = 0
⇒ λ3 − 5λ2 + (6 + 2 + 0)λ − 4 = 0
⇒ λ3 − 5λ2 + 8λ − 4 = 0
⇒ (λ − 1)(λ2 − 4λ + 4) = 0
⇒ (λ − 1)(λ − 2)2 = 0
Thus the eigenvalues of A are λ = 1, 2, 2. To find the eigenvectors corresponding to λ, we have to
solve the system (A − λI)x = 0, i.e.,
    
−λ 0 −2 x1 0
 1 2−λ 1   x2 = 0 
  (2.1.2)
1 0 3−λ x3 0
• For λ = 1 in (2.1.2), we obtain
    
−1 0 −2 x1 0
 1 1 1   x2  =  0 
1 0 2 x3 0
The augmented matrix of the system is
 
−1 0 −2 0
 1 1 1 0 
1 0 2 0
Applying R1 → (−1)R1 , we obtain  
1 0 2 0
 1 1 1 0 
1 0 2 0
4 Chapter 2 Eigenvalues, Eigenvectors

Applying R2 → R2 − R1 and R3 → R3 − R1 , we obtain

 
1 0 2 0
 0 1 −1 0 
0 0 0 0
which is in row echelon form. The system corresponding to the last matrix is

x1 + 2x3 = 0
x2 − x3 = 0

Solving, we get
x1 = −2s, x2 = s, x3 = s.
Thus the eigenvectors corresponding to λ = 1 are
     
x1 −2s −2
x =  x2  =  s  = s 1  (s ∈ R).
x3 s 1

• For λ = 2 in (2.1.2), we obtain

    
−2 0 −2 x1 0
 1 0 1   x2  =  0 
1 0 1 x3 0
The augmented matrix of the system is
 
−2 0 −2 0
 1 0 1 0 
1 0 1 0

Applying R1 → − 21 R1 , we obtain

 
1 0 1 0
 1 0 1 0 
1 0 1 0
Applying R2 → R2 − R1 and R3 → R3 − R1 , we obtain
 
1 0 1 0
 0 0 0 0 
0 0 0 0
which is in row echelon form. The system corresponding to the last matrix is

x1 + x3 = 0

Solving, we get
x1 = −t2 , x2 = t1 , x3 = t2 .
Thus the eigenvectors corresponding to λ = 2 are
       
x1 −t2 0 −1
x =  x2  =  t1  = t1  1  + t2  0  (t1 , t2 ∈ R). 
x3 t2 0 1
Tutorial 2.1 Eigenvalues and Eigenvectors 5

Example 2.1.3. Find the eigenvalues and bases for the eigenspaces of the matrix
 
4 0 1
A =  −2 1 0 
−2 0 1

Solution. The characteristic equation of A is

λ3 − tr(A)λ2 + (M11 + M22 + M33 )λ − det(A) = 0

⇒ λ3 − 6λ2 + (1 + 6 + 4)λ − 6 = 0
⇒ λ3 − 6λ2 + 11λ − 6 = 0
⇒ (λ − 1)(λ2 − 5λ + 6) = 0
⇒ (λ − 1)(λ − 2)(λ − 3) = 0

Thus the eigenvalues of A are λ = 1, 2, 3. To find the eigenvectors corresponding to λ, we have to

solve the system (A − λI)x = 0, i.e.,
    
4−λ 0 1 x1 0
 −2 1 − λ 0   x2 = 0 
  (2.1.3)
−2 0 1−λ x3 0

• For λ = 1 in (2.1.3), we obtain

    
3 0 1 x1 0
 −2 0 0   x2  =  0 
−2 0 0 x3 0

The augmented matrix of the system is

 
3 0 1 0
 −2 0 0 0 
−2 0 0 0

Applying R1 ↔ R2 , we obtain  
−2 0 0 0
 3 0 1 0 
−2 0 0 0
1


Applying R1 → − 2
R1 , we obtain
 
1 0 0 0
 3 0 1 0 
−2 0 0 0

Applying R2 → R2 − 3R1 and R3 → R3 + 2R1 , we obtain

 
1 0 0 0
 0 0 1 0 
0 0 0 0

which is in row echelon form. The system corresponding to the last matrix is

x1 = 0
6 Chapter 2 Eigenvalues, Eigenvectors

x3 = 0
From this, we get
x1 = 0, x2 = s, x3 = 0.
Thus the eigenvectors corresponding to λ = 1 are
     
x1 0 0
x = x2 = s = s 1  (s ∈ R).
    
x3 0 0
 
0
Hence, the vector 1  forms a basis for the eigenspace of A corresponding to λ = 1.

0
• For λ = 2 in (2.1.3), we obtain
    
2 0 1 x1 0
 −2 −1 0   x2 = 0 
 
−2 0 −1 x3 0
The augmented matrix of the system is
 
2 0 1 0
 −2 −1 0 0 
−2 0 −1 0
Applying R2 → R2 + R1 and R3 → R3 + R1 , we obtain
 
2 0 1 0
 0 −1 1 0 
0 0 0 0
Applying R1 → 21 R1 and R2 → (−1)R2 , we obtain

1
 
1 0 2
0
 0 1 −1 0 
0 0 0 0
which is in row echelon form. The system corresponding to the last matrix is
1
x1 + x3 = 0
2
x2 − x3 = 0
Solving, we get
1
x1 = − t, x2 = t, x3 = t.
2
Thus the eigenvectors corresponding to λ = 2 are
  1   1 
−2t −2

x1
x = x2 =  t  = t  1  (t ∈ R).
     
x3 t 1
 1 
−2
Hence, the vector  1  forms a basis for the eigenspace of A corresponding to λ = 2.
 
1
Tutorial 2.1 Eigenvalues and Eigenvectors 7

• For λ = 3 in (2.1.3), we obtain

    
1 0 1 x1 0
 −2 −2 0   x2 = 0 
 
−2 0 −2 x3 0
The augmented matrix of the system is
 
1 0 1 0
 −2 −2 0 0 
−2 0 −2 0
Applying R2 → R2 + 2R1 , we obtain
 
1 0 1 0
 0 −2 2 0 
0 0 0 0

Applying R1 → − 12 R1 , we obtain

 
1 0 1 0
 0 1 −1 0 
0 0 0 0
which is in row echelon form. The system corresponding to the last matrix is
x1 + x3 = 0
x2 − x3 = 0.
Solving, we get
x1 = −r, x2 = r, x3 = r.
Thus the eigenvectors corresponding to λ = 1 are
     
x1 −r −1
x =  x2  =  r  = r  1  (r ∈ R).
x3 r 1
 
−1
Hence, the vector  1  forms a basis for the eigenspace of A corresponding to λ = 3. 
1
 
1 2 −3
Example 2.1.4. If A =  0 2 3 , then find the eigenvalues of AT and 5A.
0 0 2
Solution. Here the matrix A is an upper triangular matrix. So the eigenvalues of A are the
diagonal entries, namely,
λ = 1, 2, 2.
Since A and AT have the same eigenvalues, the eigenvalues of AT are also
λ = 1, 2, 2.
It is known that, if λ is an eigenvalue of a matrix A, then kλ is an eigenvalue of kA. Thus the
eigenvalues of 5A are
λ = 5(1), 5(2), 5(2) ⇒ λ = 5, 10, 10. 
8 Chapter 2 Eigenvalues, Eigenvectors
 
1 3 7 11
9
 0 21 3 8 
Example 2.1.5. Find the eigenvalues of A for A = 
 

 0 0 0 4 
0 0 0 2

Solution. Here the matrix A is an upper triangular matrix. So the eigenvalues of A are the
diagonal entries, namely,
1
λ = 1, , 0, 2.
2
It is known that, if λ is an eigenvalue of a matrix A, then λk is an eigenvalue of Ak . Thus the
eigenvalues of A9 are
 9
19 1
λ=1 , , 09 , 29 ⇒ λ = 1, , 0, 512. 
2 512

Exercises



3 1 4
Exercise 2.1.1. Find the eigenvalues of the matrix A =  0 2 6 
0 0 5

Exercise 2.1.2. Find the eigenvalues and bases for the eigenspaces of the matrix
 
1 2 1
A =  1 −1 −1 
1 0 1
 
0 0 2
Exercise 2.1.3. Find bases for the eigenspace of A =  1 2 1 
1 0 3

Exercise 2.1.4. Find the eigenvalues of the matrix AT . Is A invertible ?

 
1 3 4
A= 0 0
 5 
0 0 9
 
1 0
Exercise 2.1.5. If A = , then find the eigenvalues of A2 and A−1 .
0 2

Exercise 2.1.6. Find the eigenvalues and bases for the eigenspaces of A25 and A + 2I, where
 
3 0
A=
8 −1
 
cos θ − sin θ
Exercise 2.1.7. Show that if 0 < θ < π, then A = has no real eigenvalues and
sin θ cos θ
consequently no eigenvector.
Tutorial 2.2 Cayley-Hamilton Theorem 9

Answers
     
−3 1 0
2.1.1 λ = 3, 2, 5 2.1.2 λ = −2 :  4  ; λ = 2 :  0  ; λ = 1 :  −1 
1 1 2
 3 √17     3 √17 
√ −2 + 2 0 √ −2 − 2
17 
3
2.1.3 λ = 2 + 2 : 1  ; λ = 2 : 1  ; 23 − 217 : 
 1 
1 0 1
 1   
1 25 0
2.1.4 1, 0, 9; no 2.1.5 1, 4; 1, 2 2.1.6 λ = 3 : 2 ; λ = −1 :
1 1

2.2 Tutorial : Cayley-Hamilton Theorem

Cayley-Hamilton Theorem

Every square matrix A satisfies its characteristic equation.

Solved Examples

Example 2.2.1. Verify Cayley-Hamilton theorem for the matrix A and hence find A4 , where
 
1 0 2
A= 0 0 1 
1 3 3

Solution. The characteristic equation of A is

λ3 − tr(A)λ2 + (M11 + M22 + M33 )λ − det(A) = 0

⇒ λ3 − 4λ2 + (−3 + 1 + 0)λ + 3 = 0
⇒ λ3 − 4λ2 − 2λ + 3 = 0.

We have to verify that

A3 − 4A2 − 2A + 3I = 0. (2.2.1)
Now
    
1 0 2 1 0 2 3 6 8
A2 =  0 0 1   0 0 1 = 1 3
  3 
1 3 3 1 3 3 4 9 14
    
3 6 8 1 0 2 11 24 36
A3 =  1 3 3   0 0 1 =
  4 9 14 
4 9 14 1 3 3 18 42 59

Thus

A3 − 4A2 − 2A + 3I
10 Chapter 2 Eigenvalues, Eigenvectors
       
11 24 36 12 24 32 2 0 4 3 0 0
=  4 9 14  −  4 12 12  −  0 0 2  +  0 3 0 
18 42 59 16 36 56 2 6 6 0 0 3
 
0 0 0
=  0 0 0 
0 0 0

Hence, Cayley-Hamilton theorem is verified. Also, from (2.2.1), we have

A3 = 4A2 + 2A − 3I
⇒ A4 = 4A 3 2
 + 2A − 3A     
44 96 144 6 12 16 3 0 6
⇒ A4 =  16 36 56  +  2 6 6  −  0 0 3 
72 168 236 8 18 28 3 9 9
 
47 108 154
⇒ A4 =  18 42 59  
77 177 253
 
1 3 7
Example 2.2.2. Use Cayley-Hamilton theorem to find A−1 for A =  4 2 3 
1 2 1
Solution. The characteristic equation of A is

λ3 − tr(A)λ2 + (M11 + M22 + M33 )λ − det(A) = 0

⇒ λ3 − 4λ2 + (−4 − 6 − 10)λ − 35 = 0
⇒ λ3 − 4λ2 − 20λ − 35 = 0.

By the Cayley-Hamilton theorem,

A3 − 4A2 − 35I = 0

Multiplying both sides by A−1 , we get

A−1 (A3 − 4A2 − 20A − 35I) = A−1 (0)

⇒ A2 − 4A − 20I − 35A−1 = 0
1
⇒ A−1 = (A2 − 4A − 20I) (2.2.2)
35
Now     
1 3 7 1 3 7 20 23 23
A2 =  4 2 3   4 2 3  =  15 22 37 
1 2 1 1 2 1 10 9 14
Hence by (2.2.2), we obtain
     
20 23 23 4 12 28 20 0 0
−1 1 
A = 15 22 37  −  16 8 12  −  0 20 0 
35
10 9 14 4 8 4 0 0 20
 
−4 11 −5
1 
= −1 −6 25  
35
6 1 −10
Tutorial 2.2 Cayley-Hamilton Theorem 11

Exercises

Exercise 2.2.1. Verify Cayley-Hamilton theorem for the matrix A and hence find A4 , where
 
0 1 0
A= 0  0 1 
1 −3 3

Exercise 2.2.2. Verify Cayley-Hamilton theorem for the matrix

 
2 −1 1
A =  −1 2 −1 
1 −1 2

and hence find A−1 .

 
2 1 1
Exercise 2.2.3. If A =  0 1 0 , then prove that
1 1 2

8 5 5
A8 − 5A7 + 7A6 − 3A5 + A4 − 5A3 + 8A2 − 2A + I =  0 3 0 
5 5 8

Answers

3 1
− 41
   
3 −8 6 4 4
1 3 1
2.2.1  6 −15 10  2.2.2 A−1 = 
 
4 4 4 
10 −24 15 1
−4 1
4
3
4

zzzzzzz
12 Chapter 2 Eigenvalues, Eigenvectors

2.3 Tutorial : Diagonalization

Definition

A square matrix A is called diagonalizable if there exists an invertible matrix P such that P −1 AP
is a diagonal matrix. The matrix P is said to diagonalize A.

Theorems on Diagonalization

(1) Let A be an n × n matrix. Then A is diagonalizable if and only if A has n basis vectors for
the eigenspace of A.

(2) An n × n matrix A having n distinct eigenvalues is diagonalizable.

Method for Diagonalization

Let A be an n × n matrix. Then following is the procedure for diagonalizing A.

• Find the eigenvalues of A.

• Find a basis for the eigenspace of A corresponding to each eigenvalue.

• If there are total n basis vectors, then the matrix is diagonalizable, otherwise not.

• Form the matrix P having the n basis vectors as its columns.

• Find the inverse of P .

• Find the matrix P −1 AP which will be a diagonal matrix having the eigenvalues of A as its
diagonal entries.

Computation of Powers of a matrix

Let A be an n × n diagonalizable matrix. Then there exists an invertible matrix P such that

P −1 AP = D, (2.3.1)

where D is a diagonal matrix. Now,

(P −1 AP )2 = P −1 AP P −1 AP = P −1 AIAP = P −1 A2 P.

In general,
(P −1 AP )k = P −1 Ak P.
Using (2.3.1), we get
P −1 Ak P = Dk ⇒ Ak = P Dk P −1 .
The last equation expresses the k th power of A in terms of the k th power of the diagonal matrix
D and Dk can be easily obtained by taking the k th power of the diagonal entries of D.
Tutorial 2.3 Diagonalization 13

Solved Examples

Example 2.3.1. Find a matrix P that diagonalizes A and determine P −1 AP , where

 
1 0 0
A= 0 1
 1 
0 1 1
Solution. The characteristic equation of A is
λ3 − tr(A)λ2 + (M11 + M22 + M33 )λ − det(A) = 0
⇒ λ3 − 3λ2 + (0 + 1 + 1)λ − 0 = 0
⇒ λ3 − 3λ2 + 2λ = 0.
⇒ λ(λ2 − 3λ + 2) = 0
⇒ λ(λ − 1)(λ − 2) = 0
Thus the eigenvalues of A are λ = 0, 1, 2. To find the eigenvectors corresponding to λ, we have to
solve the system (A − λI)x = 0, i.e.,
    
1−λ 0 0 x1 0
 0 1−λ 1   x2 = 0 
  (2.3.2)
0 1 1−λ x3 0
• For λ = 0 in (2.3.2), we obtain
    
1 0 0 x1 0
 0 1 1   x2  =  0 
0 1 1 x3 0
The augmented matrix of the system is
 
1 0 0 0
 0 1 1 0 
0 1 1 0
Applying R3 ↔ R3 − R2 , we obtain
 
1 0 0 0
 0 1 1 0 
0 0 0 0
which is in row echelon form. The system corresponding to the last matrix is
x1 = 0
x2 + x3 = 0
From this, we get
x1 = 0, x2 = −t, x3 = t.
Thus the eigenvectors corresponding to λ = 0 are
     
x1 0 0
x =  x2  =  −t  = t  −1  (t ∈ R).
x3 t 1
 
0
Hence, the vector  −1  forms a basis for the eigenspace of A corresponding to λ = 0.
1
14 Chapter 2 Eigenvalues, Eigenvectors

• For λ = 1 in (2.3.2), we obtain

    
0 0 0 x1 0
 0 0 1   x2  =  0 
0 1 0 x3 0

The augmented matrix of the system is

 
0 0 0 0
 0 0 1 0 
0 1 0 0

Applying R1 ↔ R3 , we obtain  
0 1 0 0
 0 0 1 0 
0 0 0 0
which is in row echelon form. The system corresponding to the last matrix is

x2 = 0
x3 = 0

So, the general solution of the system is given by

x1 = s, x2 = 0, x3 = 0.

Thus the eigenvectors corresponding to λ = 1 are

     
x1 s 1
x =  x2  =  0  = s  0  (s ∈ R).
x3 0 0
 
1
Hence, the vector  0  forms a basis for the eigenspace of A corresponding to λ = 1.
0

• For λ = 2 in (2.3.2), we obtain

    
−1 0 0 x1 0
 0 −1 1   x2  =  0 
0 1 −1 x3 0

The augmented matrix of the system is

 
−1 0 0 0
 0 −1 1 0 
0 1 −1 0

Applying R1 → (−1)R1 and R2 → (−1)R2 , we obtain

 
1 0 0 0
 0 1 −1 0 
0 1 −1 0
Tutorial 2.3 Diagonalization 15

Applying R3 → R3 − R2 , we obtain
 
1 0 0 0
 0 1 −1 0 
0 0 0 0

which is in row echelon form. The system corresponding to the last matrix is

x1 = 0
x2 − x3 = 0

Solving, we get
x1 = 0, x2 = r, x3 = r.
Thus the eigenvectors corresponding to λ = 1 are
     
x1 0 0
x = x2 = r = r 1 
     (r ∈ R).
x3 r 1
 
0
Hence, the vector  1  forms a basis for the eigenspace of A corresponding to λ = 2.
1
Since there are total three basis vectors for the eigenspace, the matrix A is diagonalizable and
 
0 1 0
P =  −1 0 1 
1 0 1

To find P −1 , form the block matrix

 
  0 1 0 1 0 0
P | I =  −1 0 1 0 1 0 
1 0 1 0 0 1

Applying R1 ↔ R3 , we obtain
 
  1 0 1 0 0 1
P | I =  −1 0 1 0 1 0 
0 1 0 1 0 0

Applying R2 → R2 + R1 , we obtain
 
  1 0 1 0 0 1
P |I = 0 0 2 0 1 1 
0 1 0 1 0 0

Applying R2 → R3 , we obtain
 
  1 0 1 0 0 1
P |I = 0 1 0 1 0 0 
0 0 2 0 1 1
16 Chapter 2 Eigenvalues, Eigenvectors

1


Applying R3 → 2
R3 , we obtain
 
  1 0 1 0 0 1
P |I = 0 1 0 1 0 0 
0 0 1 0 12 12

Applying R1 → R1 − R3 , we obtain

1 0 0 0 − 12 12
 
  
P |I = 0 1 0 1 0 0  
1 1
0 0 1 0 2 2

Thus
0 − 21 12
 

P −1 = 1 0 0 


1 1
0 2 2

Also,
0 − 12 12  1 0 0   0 1 0   0 0 0 
 

P −1 AP =  1 0 0    0 1 1   −1 0 1  =  0 1 0  

1 1 0 1 1 1 0 1 0 0 2
0 2 2

 
3 0 0
Example 2.3.2. Show that the matrix A =  0 2 0  is not diagonalizable.
0 1 2
Solution. Observe that the matrix A is a lower triangular matrix. So, the eigenvalues of A are
the diagonal entries, namely
λ = 3, 2, 2.
For λ = 3, there is one basis vector. Now we find eigenvectors for λ = 2.
Consider the system (A − λI)x = 0, i.e.,
    
3−λ 0 0 x1 0
 0 2−λ 0   x2  =  0 
0 1 2−λ x3 0

For λ = 2, we obtain     
1 0 0 x1 0
 0 0 0   x2  =  0 
0 1 0 x3 0
The augmented matrix of the system is
 
1 0 0 0
 0 0 0 0 
0 1 0 0

Applying R2 ↔ R3 , we obtain  
1 0 0 0
 0 1 0 0 
0 0 0 0
Tutorial 2.3 Diagonalization 17

which is in row echelon form. The system corresponds to the last matrix is

x1 = 0
x2 = 0

So, the general solution of the system is given by

x1 = 0, x2 = 0, x3 = t.

Thus the eigenvectors of A corresponding to λ = 2 are

     
x1 0 0
x = x2 = 0 = t 0  (t ∈ R).
    
x3 t 1
 
0
So, the vector  0  forms a basis for the eigenspace of A corresponding to λ = 2.
1
Thus there are only two basis vectors for the eigenspace. Hence the matrix A is not diagonal-
izable. 
 
4 0 0
Example 2.3.3. Determine whether the matrix A =  2 2 0  is diagonalizable.
2 1 1

Solution. Observe that the matrix A is a lower triangular matrix. So, the eigenvalues of A are
the diagonal entries, namely
λ = 4, 2, 1.
Since the A has three distinct eigenvalues, it is diagonalizable. 
 
1 0
Example 2.3.4. Find a matrix P that diagonalizes the matrix A = and hence find
−1 2
A13 .

Solution. Observe that the matrix A is a lower triangular matrix. So, the eigenvalues of A are
the diagonal entries, namely
λ = 1, 2.
We find the eigenvectors x corresponding to λ by solving the system (A − λI)x = 0, i.e.,
    
1−λ 0 x1 0
= (2.3.3)
−1 2 − λ x2 0

• For λ = 1 in (2.3.3), we get

    
0 0 x1 0
=
−1 1 x2 0

The equations of this system are

0x1 + 0x2 = 0
−x1 + x2 = 0
18 Chapter 2 Eigenvalues, Eigenvectors

Solving, we get
x1 = t, x2 = t (t ∈ R).
Thus the eigenvectors corresponding to λ = 1 are
     
x1 t 1
x= = =t (t ∈ R).
x2 t 1
 
1
So, the vector forms a basis for the eigenspace of A corresponding to λ = 1.
1

• For λ = 2 in (2.3.3), we get

    
−1 0 x1 0
=
−1 0 x2 0

The equations of this system are

−x1 = 0
−x1 = 0

So, the general solution of the system is given by

x1 = 0, x2 = s (s ∈ R).

Thus the eigenvectors corresponding to λ = 2 are

     
x1 0 0
x= = =s (s ∈ R).
x2 s 1
 
0
So, the vector forms a basis for the eigenspace of A corresponding to λ = 2.
1

Let  
1 0
P =
1 1
Then  
−1 1 0
P =
−1 1
So, we obtain  
−1 1 0
P AP = D =
0 2
Now,      
13 13 −1 1 0 113 0 1 0 1 0
A = PD P = = 
1 1 0 213 −1 1 −8191 8192

Exercises
 
1 0
Exercise 2.3.1. Find a matrix P that diagonalizes A = . Also determine P −1 AP .
6 −1
Tutorial 2.3 Diagonalization 19
 
−1 4 −2
Exercise 2.3.2. Find a matrix P that diagonalizes A =  −3 4 0 , and determine P −1 AP .
−3 1 3
 
1 0 0
Exercise 2.3.3. Show that the matrix A =  1 2 0  is not diagonalizable.
−3 5 2
 
5 1 1
Exercise 2.3.4. Determine whether the matrix A =  0 8 0  is diagonalizable.
0 0 4
 
0 0 0
Exercise 2.3.5. Find a matrix P that diagonalizes the matrix A =  1 0 −2  and hence find
0 1 −3
A11 .

Answers
   
 1
   1 2 1 1 0 0
0 −1 1 0
2.3.1 P = 3 ; P AP = 2.3.2 P =  1 3 3 ; P −1 AP =  0 2 0 
1 1 0 −1
1 3 4 0 0 3
   
0 0 2 0 0 0
2.3.4 A is diagonalizable 2.3.5 P =  1 2 3 ; A11 =  −1022 2046 −4094 
1 1 1 −1023 2047 −4095

zzzzzzz