MA5158 ENGINEERING MATHEMATICS I

UNIT I - MATRICES

Faculty
Department of Mathematics
Anna University, Chennai

Strictly for University Departments only 1

Section 1. Eigen values and eigen vectors

Contents
Eigen values and eigen vectors
Characteristic equation and eigen values
I Synthetic division

Eigen value problem

I Linearly independent vectors

Examples
Practice Problems and MCQs

Strictly for University Departments only 2

Eigen values and eigen vectors
 
a11 a12 ... a1n
 a21 a22 ... a2n 
Let A = 
 ...
 be a real matrix of order m × n.
... ... ... 
a a ... amn
 m1  m2
x1
 x2  n
 ...  be a vector in R . Then
Let X =  

x
n    
a11 a12 ... a1n x1 y1
 a21 a22 ... a2n   x2   y2 
Y = AX =  ...
 = 
... ... ...   ...   ... 
am1 am2 ... amn xn ym

is a vector in Rm . Thus Am×n defines a mapping from Rn into Rm .

Strictly for University Departments only 3

If A is a square matrix of order n, then A defines a mapping from Rn into
Rn . We are interested in those vectors X ∈ Rn which are mapped by A
onto a vector λX where λ is a scalar. That is we are interested in those
vectors X which satisfy the equation AX = λX. The scalar λ is called an
eigen value of A and X is called an eigen vector corresponding to
the eigen value λ. The equation AX = λX can be written as

(A − λI)X = 0nx1 .
Here 0n×1 is the zero matrix of order n × 1. X = 0nx1 is a trivial solution
of this system. We are interested only in the non-zero solutions of this
system.

Strictly for University Departments only 4

Characteristic equation and eigen values
A non-zero solution of the system of equations (A − λI)X = 0nx1 exists if
the determinant of the coefficient matrix (A − λI) is zero. That is

a11 − λ a12 ... a1n

a21 a22 − λ ... a2n
|A − λI| = =0
... ... ... ...
an1 an2 ... ann − λ

When |A − λI| = 0 is expanded, we get a nth degree polynomial equation

in λ,
(−1)n λn + k1 λn−1 + k2 λn−2 + ....... + kn = 0
where the ki ’s are expressible in terms of aij ’s. This equation
|A − λI| = 0 or equivalently (−1)n λn + k1 λn−1 + k2 λn−2 + ....... + kn = 0
is called the characteristic equation of A. Note that an eigen value

of A is precisely a root of the characteristic equation.

Strictly for University Departments only 5
Preliminaries. 1. Finding the roots of a polynomial using
Synthetic division
Example: Find the roots of the cubic equation λ3 − 5λ2 + 8λ − 4 = 0.

Solution: We have to find the first root by trial and error. We know
that the constant term of a polynomial is the product of its roots. Here
the constant term is -4. Its factors are ±1, ±2, ±4. First we will try
whether −1 is a root.
We use the following steps:

1. First write coefficients of the polynomial λ3 − 5λ2 + 8λ − 4 in

the top row. Construct a horizontal line just below the coefficients of the

polynomial.
Strictly for University Departments only 6
 1 −5 8 −4
−1 −1 6 − 14
1 −6 14 − 18

2. Since we want to find whether -1 is a root, write -1 on the left

3. Drop the first coefficient 1 below the horizontal line.
4. Multiply that number you drop, namely 1, by the number −1 on the
left.
Place the product −1 above the horizontal line just below the second
coefficient −5.
5. Add the column of numbers −5 and −1, then put the sum −6 directly
below the horizontal line.

Strictly for University Departments only 7

6. Repeat the process until you come to the last column.

1 −5 8 −4
−1 −1 6 − 14
1 −6 14 − 18

The last number below the horizontal line is always the remainder! The
remainder in this case is −18. That means λ = −1 is not a root of the
cubic equation λ3 − 5λ2 + 8λ − 4 = 0.
Now let us try whether 1 is a root of λ3 − 5λ2 + 8λ − 4 = 0 using
synthetic division.

1 −5 8 −4
1 1 −4 4
1 −4 4 0

Strictly for University Departments only 8

The last number below the horizontal line is 0. Therefore the remainder
is zero. Therefore 1 is a root of the cubic equation
λ3 − 5λ2 + 8λ − 4 = 0.
The numbers 1, −4, 4 below the horizontal line correspond to the
quadratic equation λ2 − 4λ + 4 = 0. The roots of this quadratic
equation are λ = 2, 2. Therefore, the roots of the given cubic equation are
λ = 1, 2, 2.

Strictly for University Departments only 9

Therefore, to find the first root of a cubic equation, choose a suitable
factor of the constant term of the polynomial for which the remainder is
zero. Then you can find the remaining two roots either by proceeding with
synthetic division or from the quadratic equation obtained from the
coefficients below the horizontal line.
Thus synthetic division helps in finding the first root (and all the roots
of course) of a polynomial of degree more than 2.

Strictly for University Departments only 10

Preliminaries. 2. Finding the characteristic equation
without expanding the determinant
 
a11 a12
If A = is a 2 × 2 matrix, then the characteristic equation
a21 a22
|A − λI| = 0, without expanding the determinant, can be written as

λ2 − (the sum of the diagonal elements of A)λ + |A| = 0.

 
a11 a12 a13
For a 3 × 3 matrix A = a21 a22 a23 , the characteristic equation
a31 a32 a33
|A − λI| = 0, without expanding the determinant, can be written as

λ3 − (the sum of the diagonal elements of A)λ2

+(the sum of the minors of the diagonal elements of A)λ − |A| = 0.

Strictly for University Departments only 11

 Finding the characteristic equation and the eigen values of
a 2 × 2 matrix
 
1 4
Example 1. Find the eigen values of the matrix A = .
3 2

Solution: Step 1: Find the characteristic equation of A.

The characteristic equation of A is |A − λI| = 0.

1−λ 4
= 0.
3 2−λ

Without expanding the determinant, we know that the characteristic

equation of A is

λ2 − (the sum of the diagonal elements of A)λ + |A| = 0.

Strictly for University Departments only 12

Here, the sum of the diagonal elements of A = 1 + 2 = 3 and |A| = −10.
Hence the characteristic equation of A is λ2 − 3λ − 10 = 0.
Step 2: The eigen values of A are the roots of the characteristic equation
of A.
We can factorize the characteristic polynomial as

λ2 − 3λ − 10 = (λ − 5)(λ + 2).
Therefore, the roots of the characteristic equation are λ = 5, −2.
Therefore the eigen values of A are λ = 5, −2.

Strictly for University Departments only 13

 Finding the characteristic equation and the eigen values of
a 3 × 3 matrix

 
1 2 2
Example 2. Find the eigen values of the matrix A =  0 2 1  .
−1 2 2

Solution: Step 1: Find the characteristic equation of A.

The characteristic equation of A is |A − λI| = 0.

1−λ 2 2
0 2−λ 1 = 0.
−1 2 2−λ

Strictly for University Departments only 14

The characteristic equation A, without expanding the determinant, can be
obtained like this:
λ3 − (the sum of the diagonal elements of A)λ2
+(the sum of the minors of the diagonal elements of A)λ − |A| = 0.
Here the sum of the diagonal elements = 1 + 2 + 2 = 5.
The sum of the minors of the diagonal elements is
2 1 1 2 1 2
+ + = (4 − 2) + (2 + 2) + (2 − 0) = 8
2 2 −1 2 0 2
1 2 2
|A| = 0 2 1 = 4.
−1 2 2

Therefore, the characteristic equation of A is λ3 − 5λ2 + 8λ − 4 = 0.

Strictly for University Departments only 15

Step 2: Find the roots of |A − λI| = 0.
The characteristic equation is a cubic equation. Using synthetic
division we find the first root to be 1.

1 −5 8 −4
1 1 −4 4
1 −4 4 0

The other two roots are 2, 2. Therefore the eigen values of A are
λ = 1, 2, 2.

Strictly for University Departments only 16

Eigen values and eigen vectors, eigen value problem

 
x1
 x2 
A non-zero solution X =   ...  of the system (A − λI)X = 0nx1 is


xn
called eigen vector of A corresponding to an eigen-value λ of A.
The problem of finding the eigen values and eigen vectors of a matrix A is
called an eigen value problem.

Strictly for University Departments only 17

Remarks: 1. If X is a solution for AX = λX, corresponding to the eigen
value λ, then kX is also an eigen vector where k is an arbitrary
non-zero constant. Therefore, an eigen vector corresponding to an eigen
value is unique upto scalar multiplication.

2. If the n eigen values are distinct, then the corresponding n eigen

vectors are linearly independent. But when an eigen value λ is repeated
say m times, we may not be able to find m linearly independent eigen
vectors corresponding to λ.

Let us see what do we mean by linearly independent vectors.

Strictly for University Departments only 18

Linearly independent vectors
Geometric meaning:
   
x1 y1
 x2   y2  n
Two vectors X = 
 ...  and Y =  ...  in R are said to be
  

xn yn
linearly independent if they do not lie on the same line passing through
the origin or equivalently one is not a constant multiple of the other. If
they are not linearly independent, they are said to be
linearly dependent.
   
2 −4
Example: Consider the vectors X = Y = and
3 11
 
−2/3
Z= in R2 .
−1
The vectors X and Y are linearly independent since they are not a
constant multiple of each other.
Strictly for University Departments only 19
 The vectors Y and Z are linearly independent.
1
The vectors X and Z are linearly dependent since Z = − X or
3
X = −3Z.
Geometrically, the vectors X and Z lie on the same line passing
through the origin in R2 but Y lies on a different line passing through
the origin.
Similarly, three vectors X1 , X2 , X3 in Rn are said to be linearly

independent if they do not lie on the same line or on the same plane
passing through the origin. i.e. they should lie in a three dimensional
space. In particular we cannot write X1 = a1 X2 + a2 X3 or
X2 = b1 X1 + b2 X3 or X3 = c1 X1 + c2 X2 for some constants
a1 , a2 , b1 , b2 , c1 , c2 , with atleast one of the ai ’s, bi ’s and ci ’s non-zero.

Strictly for University Departments only 20

Algebraic interpretation:
Definition. Let X1 , X2 , . . . , Xn be vectors in Rn and c1 , c2 , . . . , cn be
scalars. An expression of the form c1 X1 + c2 X2 + · · · + cn Xn is called a
linear combination of the vectors X1 , X2 , . . . , Xn .
In terms of linear combinations of vectors we can define the linear
independence and linear dependence of vectors as follows:

A set of vectors X1 , X2 ,. . . ,Xn in Rn are said to be

linearly independent if we have c1 X1 + c2 X2 + · · · + cn Xn = 0,
then c1 = c2 = · · · = cn = 0.
A set of vectors X1 , X2 ,. . . ,Xn in Rn are said to be linearly
dependent if we can find constants c1 , c2 , . . . , cn not all zero such
that c1 X1 + c2 X2 + · · · + cn Xn = 0.

Strictly for University Departments only 21

An easy test for linear independence
An easy test for the linear dependence or linear independence of a set
vectors X1 , X2 , . . . , Xm in Rn , (m ≤ n):
Form a matrix A whose row (column) vectors are X1 , X2 , . . . , Xm .
Row reduce (column reduce) the matrix A to its echelon form. The
number of non-zero rows (the number of non-zero columns) in the
row echelon form ( column echelon form) of A is called the row rank
(column rank) of A.
If the row rank (column rank) of A is m, then the vectors
X1 , X2 , . . . , Xm are linearly independent. If the row rank (column
rank) of A is less than m, then the vectors X1 , X2 , . . . , Xm are
linearly dependent.

Strictly for University Departments only 22

Suppose m = n. Then A is a square matrix of size n. Then the
row rank of A = the column rank of A = the rank of A is n if and
only if |A| is non- zero.
I If |A| is non-zero, then the vectors X1 , X2 , . . . , Xn are linearly
independent.
I If |A| is zero then the vectors X1 , X2 , . . . , Xn are linearly dependent.

Strictly for University Departments only 23

In theprevious
 example
 that
 we saw  we considered
 the vectors
2 −4 −2/3
X= Y = and Z = in R2 .
3 11 −1
 

2 −4
If A = X Y = then |A| = 22 + 12 = 34 6= 0.
3 11
Therefore X and Y are
 linearly independent.


2 −2/3
If A = X Z = then |A| = −2 + 2 = 0.
3 −1
Therefore X and Z are linearly dependent.
Exercise:    
1 2 2
1. Are the vectors 2, 1 ,−2 linearly independent? (Ans:Yes)
2 −2 1

2. Test whether the following vectors are linearly independent or not:

   
1 0 1
2,0,2 (Ans: No)
2 1 0
Strictly for University Departments only 24
Finding the eigen values and eigen vectors of a matrix A.
The case A has distinct eigen values

Example
 3. Find the eigen
 values and eigen vectors of the matrix
3 −1 1
A =  −1 5 −1  .
1 −1 3

Solution: Step 1: The characteristic equation of A is |A − λI| = 0.

3 − λ −1 1
−1 5 − λ −1 =0
1 −1 3 − λ
i.e. λ3 − 11λ2 + 36λ − 36 = 0 is the characteristic equation of A..

Strictly for University Departments only 25

Step 2: The eigen values of A are roots of the characteristic equation.
Using synthetic division we find that 2 is a root of the characteristic
equation λ3 − 11λ2 + 36λ − 36 = 0.

1 − 11 36 − 36
2 2 − 18 36
1 −9 18 0

The coefficients 1, − 9, 18 below the horizontal line correspond to the

quadratic equation λ2 − 9λ + 18 = 0. The roots of this quadratic equation
are 3, 6.
Therefore the eigen values of A are λ = 2, 3, 6.

Strictly for University Departments only 26

Step 3: To
 find
the eigen vectors:
x1
Let X =  x2  be an eigen vector corresponding to the eigen value λ.
x3
Then (A − λI)X = 0.

    
3 − λ −1 1 x1 0
 −1 5 − λ −1  x2  = 0 (1)
1 −1 3 − λ x3 0
When λ = 2 in (1), we have the following system of equations:

Strictly for University Departments only 27

 x1 − x2 + x3 = 0 (2)
−x1 + 3x2 − x3 = 0 (3)
x1 − x2 + x3 = 0 (4)

Equations (2) and (4) are the same. Therefore, from equations (2) and
(3) we have
x1 x2 x3
= =
1−3 −(−1 + 1) 3−1
or
x1 x2 x3
= =
−2 0 2
or
x1 x2 x3
= =
−1 0 1

Strictly for University Departments only 28

Therefore
  an eigen vector corresponding to the eigen value λ = 2 is
−1
 0 .
1
Any non-zero multiple of this vector is also an eigen vector corresponding
to
the eigen value λ = 2.
When λ = 3 in (1), we have the following system of equations:

−x2 + x3 = 0 (5)
−x1 + 2x2 − x3 = 0 (6)
x1 − x2 = 0 (7)

Strictly for University Departments only 29

Therefore, we get
x1 = x2 = x3 .
Take x1 = 1. Then
 x2 = x3 = 1 and an eigen vector corresponding to the
1
e-value λ = 3 is 1.

1

Strictly for University Departments only 30

When λ = 6 in (1), we have the following system of equations:

−3x1 − x2 + x3 = 0 (8)
x1 + x2 + x3 = 0 (9)
x1 − x2 − 3x3 = 0 (10)

As the determinant of the coefficient matrix is zero, at the most only two
of these equations are linearly independent. Therefore, from equations (8)
and (9) we have
x1 x2 x3
= =
(−1 − 1) −(−3 − 1) −3 + 1

Strictly for University Departments only 31

or
x1 x2 x3
= =
−2 4 −2
or
x1 x2 x3
= =
1 −2 1
Therefore
  an eigenvector corresponding to the eigenvalue λ
= 6is  
1 −1 1 1
−2. Therefore, 2, 3, 6 are the eigen values and  0 , 1, −2
1 1 1 1
are the corresponding eigen vectors of A.

Strictly for University Departments only 32

Repeated eigen values, the total number of linearly
independent eigen vectors equals the order of A

Example
 4. Find the 
eigen values and eigen vectors of the matrix
−2 2 −3
A= 2 1 −6  .
−1 −2 0

Solution: Step 1: The characteristic equation of A is |A − λI| = 0.

−2 − λ 2 −3
2 1 − λ −6 = 0
−1 −2 −λ
3 2
i.e. λ + λ − 21λ − 45 = 0 is the characteristic equation of A..

Strictly for University Departments only 33

Step 2: The eigen values of A are roots of the characteristic equation.
Using synthetic division we find that 5 is a root of the characteristic
equation λ3 + λ2 − 21λ − 45 = 0.

1 1 − 21 − 45
5 5 30 45
1 6 9 0

The coefficients 1, 6, 9 below the horizontal line correspond to the

quadratic equation λ2 + 6λ + 9 = 0. The roots of this quadratic equation
are −3, −3.
Therefore the eigen values of A are λ = −3, −3, 5.

Strictly for University Departments only 34

Step 3: To
 find
the eigen vectors:
x1
Let X =  x2  be an eigen vector corresponding to the eigen value λ.
x3
Then (A − λI)X = 0.
    
−2 − λ 2 −3 x1 0
 2 1 − λ −6  x2  = 0 (11)
−1 −2 −λ x3 0
When λ = 5 in equation (11), we have the following system of equations:

Strictly for University Departments only 35

 −7x1 + 2x2 − 3x3 = 0 (12)
2x1 − 4x2 − 6x3 = 0 (13)
−x1 − 2x2 − 5x3 = 0 (14)

As the determinant of the coefficient matrix is zero, at the most only two
of these equations are linearly independent. Therefore, from equations
(12) and (13) we have
x1 x2 x3
= =
−12 − 12 −(42 + 6) 28 − 4
or
x1 x2 x3
= =
−24 −48 24
or
x1 x2 x3
= =
1 2 −1

Strictly for University Departments only 36

Therefore
 an  eigen vector corresponding to the eigen value λ = 5 is
1
X1 =  2 
−1
Any non-zero multiple of this vector is also an eigen vector
corresponding to the eigen value λ = 5.

When λ = −3 in equation (11), we have the following system of equations:

x1 + 2x2 − 3x3 = 0 (15)

2x1 + 4x2 − 6x3 = 0 (16)
−x1 − 2x2 + 3x3 = 0 (17)

Strictly for University Departments only 37

We get only one independent equation,

x1 + 2x2 − 3x3 = 0. (18)

We may take one of the co-ordinates to be zero, say x2 = 0 in equation

(18). Then we get x1 − 3x3 = 0. Let x3 be an arbitrary non-zero real
number, say 1. ( We cannot take x3 = 0 as this will make x1 to be zero.
But an eigen vector is a non-zero vector.) Then x1 = 3. Therefore,
 one
3
eigen vector corresponding to the eigen value λ = −3 is X2 = 0 .

1
To choose another linearly independent eigen vector corresponding to
λ = −3, choose either x1 = 0 or x3 = 0 in equation (18). Suppose we
take x3 = 0.

Strictly for University Departments only 38

Then we get x1 + 2x2 = 0. Take x2 = 1. Then x1 = −2.
Therefore,
 the second independent eigen vector for λ = −3 is
−2
X3 =  1  .
0
(Question: Here one may ask what if you take x1 = 0 in equation (18)?
Will you get another linearly independent eigen vector corresponding to
λ = −3?
That is, are there three linearly independent eigen vectors corresponding to
λ = −3?. The answer is NO.

Suppose you find the third eigen vector corresponding to λ = −3 by taking

x1 = 0. Call that eigen vector as X4 . You can verify that X2 , X3 , X4 are
linearly dependent. In fact if an eigen value λ is repeated m times then
you can find definitely one and at the most only m number of linearly

independent eigen vectors corresponding to λ.)

Strictly for University Departments only 39
    
1 3
Therefore, 5, −3, −3 are the eigen values and X1 =  2 , X2 = 0,
 
−1 1
 
−2
X3 =  1  are the corresponding eigen vectors of A.
0

Note: In this case we are able to find two linearly independent eigen
vectors corresponding to the eigen value λ = −3, which is repeated two
times.

In the next example we will see that this is not the case always.

Strictly for University Departments only 40

Repeated eigen values, the total number of linearly
independent eigen vectors is less than the order of A

Example
 5. Find the eigen values and eigen vectors of the matrix
1 2 2
A =  0 2 1 .
−1 2 2
Solution: Step 1: The characteristic equation of A is |A − λI| = 0.
1−λ 2 2
0 2−λ 1 = 0 i.e. λ3 − 5λ2 + 8λ − 4 = 0 is the
−1 2 2−λ
characteristic equation of A..
Step 2: The eigen values of A (roots of this characteristic equation) are
λ = 1, 2, 2.

Strictly for University Departments only 41

Step 3: To find the eigen vectors of A
 
x1
Let X =  x2  be an eigen vector corresponding to the eigen value λ.
x3
Then (A − λI)X = 0.
    
1−λ 2 2 x1 0
 0 2−λ 1   x2 = 0
  (19)
−1 2 2−λ x3 0
When λ = 1 in equation (27), we have the following system of equations:

Strictly for University Departments only 42

 2x2 + 2x3 = 0 (20)
x2 + x3 = 0 (21)
−x1 + 2x2 + x3 = 0 (22)

From equations (20) and (21) we have

x2 = −x3 . Substituting this in (22), we get, x1 = −x3 . i.e.

x1 = x2 = −x3 .

Choose x3 = 1.
 
−1
Then an eigen vector corresponding to the eigen value λ = 1 is −1.

1

Strictly for University Departments only 43

When λ = 2 in equation (27), we have the following system of equations:

−x1 + 2x2 + 2x3 = 0 (23)

x3 = 0 (24)
−x1 + 2x2 = 0 (25)

We get x3 = 0 and x1 = 2x2 .

Let x2 = 1 in the above equation. Then we  get
 x1 = 2. An eigen vector
2
corresponding to the eigen value λ = 2 is 1.
0
This is the only linearly independent eigen vector (upto scalar multiples)
corresponding to the repeated eigen value λ = 2.

Strictly for University Departments only 44

    
−1 2
Therefore, 1, 2, 2 are the eigen values and −1, 1, are the
1 0
corresponding eigen vectors of A.

Note: In this case we are able to find only one linearly independent
eigen vector corresponding to the eigen value 2, which is repeated two
times.

Strictly for University Departments only 45

Practice Problems

Find theeigen values

 and eigen vectors of the following Matrices:
2 1 1
1) A = 1 2 1,
0 0 1
   
1 1 1
Ans: Eigen values: 3,1,1. Eigen vectors: 1 , 0 , −1.
    
0 −1 0
 
8 −6 2
2) A = −6 7 −4,
2 −4 3
   
1 2 2
Ans: λ = 0, 3, 15. Eigen vectors: 2 , 1 , −2.
    
2 −2 1

Strictly for University Departments only 46

  
3 10 5
3) A = −2 −3 −4,
3 5 7
 
1 5
Ans: λ = 3, 2, 2. Eigen vectors: 1 , 2 .
−2 −5

Strictly for University Departments only 47

Multiple Choice Questions
 
3 10 5
1. The characteristic equation of −2 −3 −4 is
3 5 7
3 2
(a) λ − 7λ − 12λ + 16 = 0
(b) λ3 − 7λ2 + 12λ − 16 = 0
(c) λ3 − 7λ2 + 16λ − 12 = 0
(d) λ3 − 7λ2 − 16λ + 12 = 0  
1 1 1
2. Pick out the correct answer: The eigen values of 1 1 1 are
1 1 1
(a) 0, 0, 0
(b) 0, 0, 1
(c) 0, 0, 3
(d) 1, 1, 1

Strictly for University Departments only 48

Answers to MCQs:

1. (c)
2. (c)

Strictly for University Departments only 49

Section 2. Properties of eigen values and eigen
vectors

Contents
Properties
Examples
Practice Problems and MCQs

Strictly for University Departments only 50

Transpose

1. Transpose: Any square matrix A and its transpose AT have the same
eigen values.

Proof: The characteristic equation of AT is

|AT − λI| = 0

. Also, |AT − λI| = |(A − λI)T | = |A − λI|. Therefore, the characteristic

equation of A is the same as the characteristic equation of AT . Therefore,
the eigen values of A and AT are the same.

Strictly for University Departments only 51

Triangular matrix

2. Triangular matrix: The eigen values of a triangular matrix are just the
diagonal elements of the matrix.
 
a11 a12 ... a1n
 0 a22 ... a2n 
Proof: Let A =  ... ... ...
 be an upper triangular matrix.
... 
0 0 ... ann
Then, expanding the determinant |A − λI| along the first column, we get
|A − λI| = (a11 − λ)(a22 − λ) · · · (ann − λ). Therefore, from
|A − λI| = 0, we see that the eigen values of A are λ = a11 , a22 , ..., ann .

Strictly for University Departments only 52

Idempotent matrix
3. Idempotent matrix: The eigen values of an idempotent matrix are
either 0 or 1.
Proof: Let A be an idempotent matrix. i.e. A2 = A. Let λ be an eigen
value
of A and X its corresponding eigen vector. Hence we have

AX = λX

=⇒ A2 X = λX =⇒ A(AX) = A(λX) = λX

=⇒ λ2 X = λX =⇒ (λ2 − λ)X = 0

Since X 6= 0, we have λ2 − λ = 0.
Therefore, either λ = 0 or λ = 1.
Strictly for University Departments only 53
Powers of A

4. Powers of A: If λ is an eigen value of A, then λm is an eigenvalue of

Am .
Proof: Proof by induction on m.
When m = 0, X = A0 X = λ0 X. When m = 1, AX = λX. Therefore the
result is true when m = 1.
Assume it is true for m = k. i.e.Ak X = λk X. To prove it is true for
m = k + 1:
Consider Ak+1 X = A(Ak X) = A(λk X) by induction.
=⇒ A(Ak X) = λk (AX) = λk λX = λk+1 X. Hence, Ak+1 X = λk+1 X.
This completes the induction. Therefore, λm is an eigen value of Am .

Strictly for University Departments only 54

Example 1
 
8 −4
Example 1. Given A = find the eigen values of AT , A4 .
2 2
Solution: The characteristic equation of A is |A − λI| = 0.
8 − λ −4
=0
2 2−λ
i.e. λ2 − 10λ + 24 = 0 is the characteristic equation of A.
The eigen values of A (roots of this characteristic equation) are λ = 4, 6.
By property 1, eigen values of A and the eigenvalues of AT are the
same. Therefore eigen values of AT are 4, 6.
By property 4, eigen values of A4 are of the form λ4 where λ is an
eigen value of A. Hence 44 , 64 are the eigen values of A4 .

Strictly for University Departments only 55

Trace and determinant
5. trace(A) and Det(A):(i) Trace(A)=the sum of the principal diagonal
elements of A= the sum of the eigen values.
(ii) det(A)= the product of the eigen values of A.
 
a11 a12 a13
Proof: Take n = 3. Let A = a21 a22 a23 . The characteristic
a31 a32 a33
a11 − λ a12 a13
polynomial is |A − λI| = a21 a22 − λ a23
a31 a32 a33 − λ
3 2
=(-λ) + (a11 + a22 + a33 )λ − ..... + |A|.
On the other hand, if λ1 , λ2 , λ3 are the eigen values of A, then the
characteristic polynomial of A is
(λ1 − λ)(λ2 − λ)(λ3 − λ) = −[(λ − λ1 )(λ − λ2 )(λ − λ3 )]

Strictly for University Departments only 56

= −[(λ)3 − (λ1 + λ2 + λ3 )λ2 + (λ1 λ2 + λ2 λ3 + λ3 λ1 )λ − λ1 λ2 λ3 ].

Hence we have two expressions for the characteristic polynomial of A.

We compare the coefficient of λ2 in both the above expressions. We get

Trace(A) = a11 +a22 +a33 = λ1 +λ2 +λ3 = sum of the eigenvalues of A .

Comparing the constant terms in both the expressions we get,

det(A) = λ1 λ2 λ3 = product of the eigen values of A.

For n = 3 the result is proved.
For general n also one can give a similar proof.

Strictly for University Departments only 57

Example 2

 
2 3 −2
Example 2. Given A = −2 1 1  find the sum and product of the
1 0 2
eigen values of A.

Solution: We know that

The sum of the eigen values of A = trace(A) = sum of the

diagonal elements of A = 2 + 1 + 2 = 5

The product of the eigen values = the determinant of A = 21.

Strictly for University Departments only 58

Inverse
1
6. Inverse: If λ is an eigen value of A, then is an eigen value of A−1 ,
λ
provided A−1 exists.
Proof: Let λ be an eigen value of A, with eigen vector X. Then, we have

AX = λX.

Suppose A−1 exists. Then, pre-multiplying both the sides of the above
equation by A−1 , we get

A−1 AX = A−1 λX = λA−1 X.

Or
1
A−1 X = X.
λ
1
Therefore, is an eigen value of A−1 with eigen vector X, if A−1 exists.
λ
Strictly for University Departments only 59
Orthogonal matrix

7. Orthogonal matrix: If λ is an eigen value of an orthogonal matrix A,

1
then is also an eigen value of the matrix A.
λ

Proof: Let A be an orthogonal matrix. Then AAT = I =⇒ A−1 = AT .

We know that the eigen values of A and AT are the same.
Also, we know that when λ is an eigen value of matrix A, λ−1 is an eigen
value of matrix A−1 .Here AT = A−1 .
Therefore, whenever λ is an eigen value of an orthogonal matrix A, λ−1 is
also an eigen value of the matrix A.

Strictly for University Departments only 60

Similar matrices

Definition: Two matrices A and B are said to be similar, if there exists a

non-singular matrix P such that B = P −1 AP .

8. Similar matrices: Similar matrices have the same set of eigen values.
Proof: Let A and B be similar matrices. Then B = P −1 AP for a
non-singular matrix P.
The characteristic polynomial of B is |B − λI| = |P −1 AP − λI|
= |P −1 AP − P −1 (λI)P | = |P −1 (A − λI)P | = |P −1 ||A − λI||P |
= |A − λI|
which is the characteristic polynomial of A. Therefore, A and B have the
same eigen values.

Strictly for University Departments only 61

Example 3

 
8 −4
Example 3: Given A = find the eigen values of A−1 .
2 2

Solution: In Example 1, we computed the eigen values of A as 4 and 6.

It is clear that A−1 exists as |A| = 24.
By property 6, eigen values of A−1 are 1/4, 1/6.

Strictly for University Departments only 62

Example 4

 
8 −2 1
Example 4. If A = 0 0 11, find the eigen values of (i) A3 , (ii)
0 0 9
A−1 , if A−1 exists.
Solution: A is a triangular matrix. Therefore the eigen values of A are
the principal diagonal elements, namely 8, 0, 9.
The eigen values of A3 are 83 , 0, 93 .

Determinant of A is the product of the eigen values of A. Therefore

|A| = 0. Hence A−1 does not exist. Therefore we cannot find the
eigen values of A−1 .

Strictly for University Departments only 63

Example 5
Example 5. If λ is an eigen value of a non-singular matrix A, show that
|A|
is an eigen value of the matrix adj A.
λ
Solution: We have AX = λX where X is an eigen vector corresponding
to the eigen value λ. Since A is non-singular A−1 exists. Therefore we
have
A−1 AX = λA−1 X
1
A−1 X = X
λ
adj A 1 |A|
X = X (or) adj AX = X.
|A| λ λ

|A|
This proves that is an eigen value of adj A.
λ

Strictly for University Departments only 64

Orthogonal vectors
   
x1 y1
 x2   y2  n
Definition: Two vectors X =   ..  and Y =  ..  in R are said to
  

xn yn
be orthogonal, if their dot product is zero.
i.e. if X · Y = x1 y1 + x2 y2+ ...+ xn yn = 0 or
y1
 y2 
X T Y = x1 x2 .. xn 


 ..  = 0.


yn
   
2 1
Example: X = −1 and Y = 4 are orthogonal in R3 as
2 1
X · Y = 2 · 1 + (−1) · 4 + 2 · 1 = 0.

Strictly for University Departments only 65

Symmetric matrices
9. Symmetric matrices: Let λ1 and λ2 be two distinct eigen values of a
symmetric matrix A with corresponding eigen vectors X1 and X2
respectively.
Then the eigen vectors X1 and X2 are orthogonal.
Proof: We have
AX1 = λ1 X1
. On taking transpose on both sides we get

(AX1 )T = (λ1 X1 )T =⇒ X1T AT = (λ1 X1 )T ,

=⇒ X1T A = λ1 X1T since AT = A.

Post-multiplying both sides by X2 we get

=⇒ X1T AX2 = λ1 X1T X2

Strictly for University Departments only 66

 =⇒ λ2 X1T X2 = λ1 X1T X2 , (since AX2 = λ2 X2 .)

=⇒ (λ2 − λ1 )X1T X2 = 0

Since λ1 6= λ2 , we have X1T X2 = 0.

Therefore X1 and X2 are orthogonal.
Therefore, the eigen vectors corresponding to distinct eigen values of a
real symmetric matrix are orthogonal.

Strictly for University Departments only 67

Spectral shift and scalar multiples
10.Spectral shift: If A has an eigen value λ, then A ± kI has eigen
values
λ ± k.

Proof: The characteristic equation of A is |A − λI| = 0 or

|(A ± kI) − (λ ± k)I| = 0. Therefore, eigen values of A ± kI are
λ ± k.

11. Scalar Multiples: If λ is an eigen value of A, then kλ is an eigen

value of
kA.

Proof: If (A − λI)X = 0, then k(A − λI)X = (kA − kλI)X = 0.

Therefore, if |A − λI| = 0, then |kA − kλI| = 0. This means kλ is an
eigen value of kA.
Strictly for University Departments only 68
Spectral mapping theorem

12. Spectral mapping theorem: If λ is an eigen value of A, then

P (λ) = km λm + km−1 λm−1 + .... + k1 λ + k0 is an eigen value of the
polynomial P (A) = km Am + km−1 Am−1 + .... + k1 A + k0 I.
Proof: Let λ be an eigen value of A with eigen vector X.
Then P (A)X = (km Am + km−1 Am−1 + .... + k1 A + k0 I)X
= km Am X + km−1 Am−1 X + .... + (k1 A + k0 I)X
= [km λm + km−1 λm−1 + .... + (k1 λ + k0 )]X = P (λ)X.
Therefore, P (λ) = km λm + km−1 λm−1 + .... + k1 λ + k0
is an eigen value of the polynomial P (A).

Strictly for University Departments only 69

Practice Problems

 
2 1
1. If A = find the eigen values of A8 , 15A, A−1 .
1 2
 
3 8 −2
2. If A = 0 −1 11  find the sum of the squares of the eigenvalues
0 0 6
of A.

3. If 
the product of the
 two eigen values of the matrix
6 −2 2
A = −2 3
 −1 is 16, find the third eigen value.
2 −1 3

Strictly for University Departments only 70

Multiple Choice Questions
 
2 −3
1. The sum and product of the eigen values of the matrix are
4 −2
(a) 2,-2
(b) 0,8
(c) 4,-3
(d) 0,12  
3 1 4
2. If A = 0 2 6 , then eigen value of A−1 are
0 0 5
−1 −1 −1
(a) 3 , 2 , 5
(b) −3, −2, −5
(c) 31 , 12 , 15
(d) 9, 4, 25

Strictly for University Departments only 71

3. I. The sum of the eigen values of A is equal to the trace of A.
II. The vectors [1, 1, −1, 1], [1, −1, 2, −1], [4, 0, 2, 0] are linearly
independent.
(a) Both I and II are true
(b) Only I is true
(c) Only II is true
(d) Both I and II are false

Strictly for University Departments only 72

Answers to MCQs

1. (b)
2. (c)
3. (b)

Strictly for University Departments only 73

Section 3. Cayley-Hamilton theorem

Contents
Cayley-Hamilton Theorem
I Verification of Cayley-Hamilton theorem
Applications
I Finding the inverse of a matrix
I Finding the polynomial expressions in A
Examples
Practice Problems and MCQs

Strictly for University Departments only 74

Cayley-Hamilton Theorem

Cayley-Hamilton Theorem: Every square matrix satisfies its

own characteristic equation.
(i.e.) if the characteristic equation for the nth order square matrix A is
|A − λI| = (−1)n λn + k1 λn−1 + ..... + kn−1 λ + kn = 0 then
(−1)n An + k1 An−1 + ..... + kn−1 A + kn In = 0n×n .
Applications: 1. This method is used for the computation of the
inverse of large matrices.
Note that kn = product of the eigen values of A = |A|. Hence A−1 exists
iff kn is not equal to zero. Suppose |A| = kn 6= 0.

Strictly for University Departments only 75

By pre-multiplying the above equation by A−1 and re-arranging the
equation we get
1
A−1 = − ((−1)n An−1 + k1 An−2 + ..... + kn−1 In ).
kn
This method will be useful for finding the inverses of very large matrices
using a computer.
2. Given a polynomial P (A) in A of degree greater than or equal to
n, using division algorithm we can write
P (A) = Quotient((−1)n An + k1 An−1 + ..... + kn−1 A + kn In ) + Remainder.

By Cayley-Hamilton theorem
(−1)n An + k1 An−1 + ..... + kn−1 A + kn In = 0. Therefore
P (A) = Remainder, which is a polynomial in A of degree atmost (n − 1).
This makes the computation of P (A) very simple.
Strictly for University Departments only 76
Examples
Example
 1: Verify
 Cayley Hamilton theorem for the matrix
2 5
A= and hence find A−1 .
1 −3
Solution:
The characteristic equation of A is |A − λI| = 0

2−λ 5
=0
1 −3 − λ

λ2 − (sum of the diagonal elements)λ + |A| = 0

λ2 − (2 − 3)λ + (−11) = 0
λ2 + λ − 11 = 0
    
2 2 5 2 5 9 −5
Now A = =
1 −3 1 −3 −1 14

Strictly for University Departments only 77

        
9 −5 2 5 11 0 0 0
A2 + A − 11I = + − = .
−1 14 1 −3 0 11 0 0

Therefore the matrix A satisfies its own characteristic equation.

Hence Cayley Hamilton Theorem is verified for the matrix A.
To find A−1 ,
we have A2 + A − 11I = 0
Pre-multiplying both sides by A−1 we get,
A + I − 11A−1 = 0
   
1 1 2 5 1 0
A−1 = (A + I) = +
11 11 1 −3 0 1
 
1 3 5
A−1 =
11 1 −2

Strictly for University Departments only 78

  
1 0 3
Example 2: Verify Cayley-Hamilton theorem for A =  2 1 −1 
1 −1 1
and hence find A−1 and A4 .
Solution:
The characteristic equation of A is |A − λI| = 0
1−λ 0 3
2 1 − λ −1 =0
1 −1 1 − λ
λ3 − (sum of the diagonal elements of A)λ2 +
(sum of the minors of the diagonal elements)λ − |A| = 0
λ3 − (1 + 1 + 1)λ2 + (0 − 2 + 1)λ − |A| = 0.
λ3 − 3λ2 − λ + 9 = 0 as |A| = −9.
Now,    
1 0 3 1 0 3 4 −3 6
A2 = 2 1 −1   2 1 −1  =  3 2 4 
1 −1 1 1 −1 1 0 −2 5
Strictly for University Departments only 79
     
4 −3 6 1 0 3 4 −9 21
A3 =A2 A= 3 2 4   2 1 −1  =  11 −2 11 
0 −2 5 1 −1 1 1 −7 7
   
4 −9 21 12 −9 18
A3 − 3A2 − A + 9I= 11 −2 11  −  9 6 12 
1 −7 7 0 −6 15
   
1 0 3 9 0 0
- 2 1 −1
  +  0 9 0 
1 −1 1 0 0 9
 
0 0 0
= 0 0 0 
0 0 0

Therefore the matrix A satisfies its own characteristic

equation.

Strictly for University Departments only 80

To find A−1 . We have

A3 − 3A2 − A + 9I = 0 (26)

Pre-multiplying both sides by A−1 we get,

A2 − 3A − I + 9A−1 = 0
1
A−1 = (−A2 + 3A + I)
9
     
−4 3 −6 3 0 9 1 0 0
1
=  −3 −2 −4  +  6 3 −3  +  0 1 0 
9
0 2 −5 3 −3 3 0 0 1
 
0 3 3
−1 1 
Therefore A = 3 2 −7  .
9
3 −1 −1

Strictly for University Departments only 81

To find A4 .
Rearranging equation (1) we get,

A3 = 3A2 + A − 9I

A4 = 3A3 + A2 − 9A
     
4 −9 21 4 −3 6 9 0 27
A4 = 3 11 −2 11  +  3 2 4  −  18 9 −9 
1 −7 7 0 −2 5 9 −9 9
 
12 + 4 − 9 −27 − 3 63 + 6 − 27
= 33 + 3 − 18 −6 + 2 − 9 33 + 4 + 9 
3−9 −21 − 2 + 9 21 + 5 − 9
 
7 −30 42
=⇒ A4 = 18 −13 46 
−6 −14 17
Strictly for University Departments only 82
Example
 3: Using Cayley-Hamilton
 theorem determine A−1 , A−2 , A−3 if
4 6 6
A= 1 3 2 .
−1 −4 −3
Solution:
The characteristic equation of A is |A − λI| = 0
4−λ 6 6
1 3−λ 2 =0
−1 −4 −3 − λ
λ3 − (sum of the diagonal elements of A)λ2 +
(sum of the minors of the diagonal elements)λ − |A| = 0
λ3 − (4 + 3 − 3)λ2 + (−1 − 6 + 6)λ − (−4) = 0
or λ3 − 4λ2 − λ + 4 = 0
By Cayley Hamilton theorem A satisfies A3 − 4A2 − A + 4I = 0

Strictly for University Departments only 83

 1
=⇒ A−1 = −A2 + 4A + I

4
Now,    
4 6 6 4 6 6 16 18 18
2
A =  1 3 2   1 3 2  =  5 7 6 
−1 −4 −3 −1 −4 −3 −5 −6 −5
      
16 18 18 4 6 6 1 0 0
1
A−1 = −  5 7 6  + 4 1 3 2 + 0 1 0 
4
−5 −6 −5 −1 −4 −3 0 0 1
 
1 6 6
1
= −1 6 2 
4
1 −10 −6

Pre-multiplying by A−1 on both sides, 

−2 −1 −1 −1 1 2
A =A A =A (−A + 4A + I)
4
1
−A + 4I + A−1

=
4

Strictly for University Departments only 84

 1 −9 −9
 
 4 2 2 
 
 
 −5 5 −3 
1 
=  4 2 2 
4



 5 3 11 
 
4 2 2
 

Similarly, pre-multiplying by A−1 on both sides,


−3 −1 −2 −1 1 −1
A =A A =A (−A + 4I + A )
4
1
−I + 4A−1 + A−2

=
4
 
1 78 78
1 
= −21 90 26  .
64
21 −154 −90

Strictly for University Departments only 85

Example
 4: Verify
 Cayley Hamilton theorem for the matrix
1 4
A= and hence find A−1 . Also find the matrix represented by
3 2
A5 − 3A4 − 7A3 − 9A2 − 20A − 10I.
Solution:
The characteristic equation of A is |A − λI| = 0

1−λ 4
=0
3 2−λ

(1 − λ)(2 − λ) − 12 = 0

λ2 − 3λ − 10 = 0

Now,    
2 1 4 1 4 13 12
A = =
3 2 3 2 9 16

Strictly for University Departments only 86

    
1 4 3 12
3A=3 =
3 2 9 6
     
13 12 3 12 10 0
A2 − 3A − 10I= − −
9 16 9 6 0 10
 
0 0
=
0 0
Therefore the matrix A satisfies its own characteristic equation.
Hence Cayley Hamilton Theorem is verified for the matrix A.
To find A−1 ,we have A2 − 3A − 10I = 0
Pre-multiplying both sides by A−1 we get,
A − 3I − 10A−1 = 0
   
1 1 1 4 1 0
A−1 = (A − 3I) = −3
10 10 3 2 0 1
 
1 −2 4
A−1 =
10 3 −1

Strictly for University Departments only 87

Now we divide the polynomial A5 − 3A4 − 7A3 − 9A2 − 20A − 10I by
A2 − 3A − 10I

Rough work: Instead of A write x and we divide

x5 − 3x4 − 7x3 − 9x2 − 20x − 10 by x2 − 3x − 10.
x3 + 3x
2 5 4 3 2


x − 3x − 10 x − 3x − 7x − 9x − 20x − 10
− x5 + 3x4 + 10x3
3x3 − 9x2 − 20x
− 3x3 + 9x2 + 30x
10x
Therefore we have,
x5 − 3x4 − 7x3 − 9x2 − 20x − 10 = Quotient(x2 − 3x − 10) + Remainder.

x5 − 3x4 − 7x3 − 9x2 − 20x − 10 = (x3 + 3x)(x2 − 3x − 10) + 10x − 10.

Strictly for University Departments only 88

Replace x by the matrix A. Note that the constant term −10 in the
polynomial should be replaced by the matrix −10I. Now we go back
to the solution of the problem.

We obtain

A5 − 3A4 − 7A3 − 9A2 − 20A − 10I = (A3 + 3A)(A2 − 3A − 10I)+

(10A − 10I)

= 10A − 10I (as A2 − 3A − 10I = 0.)

   
1 4 1 0
= 10 − 10
3 2 0 1
 
0 40
= .
30 10

Strictly for University Departments only 89

Example 5: By using Cayley Hamilton Theorem, express
B = A8 − 11A7 − 4A6 + A5 + A4 − 11A3 −3A2 + 2A + I as a quadratic
1 2 3
polynomial in A and also find B where A =  2 4 5  .
3 5 6
Solution:
The characteristic equation of A is |A − λI| = 0
1−λ 2 3
2 4−λ 5 =0
3 5 6−λ
(1 − λ)[(4 − λ)(6 − λ) − 25] − 2[2(6 − λ) − 15]
+3[10 − 3(4 − λ)] = 0
=⇒ λ3 − 11λ2 − 4λ + 1 = 0

Strictly for University Departments only 90

We know that by Cayley Hamilton theorem every square matrix satisfies its

own characteristic equation.

Hence A3 − 11A2 − 4A + I = 0
To find B: Divide B by A3 − 11A2 − 4A + I. Then write B as
B = Quotient(A3 − 11A2 − 4A + I) + Remainder
Therefore, B = A8 − 11A7 − 4A6 + A5 + A4 − 11A3 − 3A2 + 2A + I
= (A5 + A)(A3 − 11A2 − 4A + I) + A2 + A + I
= (A5 + A)(0) + A2 + A + I
= A2 + A + I
      
1 2 3 1 2 3 1 2 3 1 0 0
B =  2 4 5  2 4 5  +  2 4 5  +  0 1 0 
3 5 6 3 5 6 3 5 6 0 0 1

Strictly for University Departments only 91

      
14 25 31 1 2 3 1 0 0
B =  25 45 56  +  2 4 5  +  0 1 0 
31 56 70 3 5 6 0 0 1
 
16 27 34
=  27 50 61  .
34 61 77
Remark: Here A is a square matrix of order 3. Note that using division
algorithm, any polynomial B in A can be reduced to a polynomial of
degree less than or equal to 2.
In general, if A is a square matrix of order n, then any polynomial in A, in
particular higher powers of A, are in fact equal to a polynomial in A of
degree less than or equal to (n − 1).
Hence using Cayley Hamilton theorem computations are made simpler.

Strictly for University Departments only 92

Practice Problems

 
1 1 3
1. Find the characteristic equation of the matrix A =  1 3 −3 
−2 −4 −4
and hence find its inverse.
 
2 1 1
2. Find the characteristic equation of the matrix A =  0 1 0  and
1 1 2
−1
hence compute A . Also find the matrix represented by
A8 − 5A7 + 7A6 − 3A5 + A4 − 5A3 + 8A2 − 2A + I.

Strictly for University Departments only 93

Multiple Choice Questions

1. I. Any set of vectors which includes the zero vector is linearly

independent.
II. Every square matrix does not satisfy its own characteristic equation.
(a) Both I and II are true
(b) Only I true
(c) Only II true
(d) Both I and II are false  
4 3 2 2 3
2. The value of A − 4A − 5A − A + 2I when A = is
3 2
   
0 1 1 0
(a) −3 (b)−3
1 0 0 1
   
0 −1 −1 0
(c) −3 (d) −3
1 0 0 −1

Strictly for University Departments only 94

Answers to MCQs

1. (d)
2. (c)

Strictly for University Departments only 95

 Section 4. Diagonalization using similarity
transformation

Contents
Similarity transformation
Diagonalization process
Application: Finding powers of a matrix using
diagonalization
Examples
Practice Problems and MCQ’s

Strictly for University Departments only 96

Similarity transformation

Let us recall the definition of similar matrices. Two square matrices A, B

of size n are said to be similar if there exists a non-singular matrix P such
that P −1 AP = B.
This transformation of A into B is known as similarity transformation.
Also recall from the properties of eigen values and eigen vectors that
similar matrices have the same eigen values .
We will use this similarity transformation to diagonalize a square matrix.

Strictly for University Departments only 97

Diagonalization

Theorem: An n × n square matrix A with n linearly independent eigen

vectors is similar to a diagonal matrix D whose diagonal elements
are the eigen values of A.
Proof: We prove the result for n = 3.
Let A be asquare
 matrixof order
 3. Let
 λ1, λ2 , λ3 be ites eigen values
x1 x2 x3
and X1 =  y1 , X2 =  y2 , X3 =  y3  be the corresponding eigen
z1 z2 z3
vectors.
 
x1 x2 x3
Let P = [X1 X2 X3 ] =  y1 y2 y3  .
z1 z2 z3

Strictly for University Departments only 98

P is a square matrix whose column vectors are the eigen vectors of A.
Since the eigen vectors X1 , X2 , X3 are linearly independent, P −1 exists.
Consider AP = A[X1 X2 X3 ] = [AX1 AX2 AX3 ]
= [λ1 X1 λ2 X2 λ3 X3 ]
 
λ1 0 0
= [X1 X2 X3 ]  0 λ2 0 
0 0 λ3
 
λ1 0 0
= P D where D =  0 λ2 0  a diagonal matrix whose
0 0 λ3
diagonal elements are the eigen values of A.

Strictly for University Departments only 99

Pre-multiply bothsides by P −1 . We get P −1 AP = D. Thus we can
transform A into a diagonal matrix D using a similarity transformation.
The matrix P which diagonalizes the matrix A is called the modal matrix
of A.
Remark 1. If a matrix A of order n has n linearly independent

eigen vectors and the sequence of the eigen vectors are fixed, then

the modal matrix P is unique upto permutation of the eigen vectors

For example, suppose λ1 , λ2 , λ3 are the three eigen values of a 3 × 3
matrix A with X1 , X2 , X3 as the corresponding eigen vectors. If we
take the modal matrix P as P = [X1 X2 X3 ] , then the diagonal matrix
 
λ1 0 0
P −1 AP = D will be D =  0 λ2 0  .
0 0 λ3

Strictly for University Departments only 100

Suppose we take P = [X2 , X1 , X3 ] , then the diagonal matrix
 
λ2 0 0
P −1 AP = D will be D =  0 λ1 0.
0 0 λ3
Both are correct. Recall that even the eigen vectors are unique upto scalar
multiples.

Remark 2. Note that not all square matrices are diagonalizable.

Suppose square matrix A of size n has less than n linearly independent

eigen vectors, then A is not diagonalizable.
For example, let us refer to Example 4 in section 1 of this unit.
 
1 2 2
In this example, the matrix A =  0 2 1  is of size 3. The eigen
−1 2 2
values of A are λ = 1, 2, 2.
Strictly for University Departments only 101
  
−1
An eigen vector corresponding to the eigen value λ = 1 is −1.
1
 
2
An eigen vector corresponding to the eigen value λ = 2 is 1.
0

This is the only linearly independent eigen vector (upto scalar multiples)
corresponding to the repeated eigen value λ = 2.
Therefore, in this case we are able to find only two linearly independent
eigen vectors in total.
Since we do not have three linearly independent eigen vectors for A,
A is not diagonalizable .

Strictly for University Departments only 102

Application: Finding powers of a matrix

Let A be a square matrix which is diagonalizable. Let P be a non-singular

matrix such that

P −1 AP = D

Then

(P −1 AP )2 = D2
P −1 A(P P −1 )AP = D2
P −1 A2 P = D2 .

Similarly, P −1 A3 P = D3
In general, P −1 Am P = Dm

Strictly for University Departments only 103

 λm
 
1 0 0 ... 0
 0 λm 0 ... 0 
Therefore, Am = P Dm P −1 where Dm 2
 ... ... ... ... ...  .
= 

0 0 0 ... λm n

It is easy to find Dm . Then we can find Am which is equal to P Dm P −1 .

This involves less computation than computing Am directly.

Strictly for University Departments only 104

Examples
Example 1: (A has distinct eigen values)  
4 1
Find a matrix P which diagonalizes the matrix A = . Verify
2 3
P −1 AP = D where D is a diagonal matrix. Hence find A6 .
Solution:
Step 1: The characteristic equation of A is |A − λI| = 0

4−λ 1
=0
2 3−λ

(4 − λ)(3 − λ) − 2 = 0

i.e., (λ − 2)(λ − 5) = 0

Strictly for University Departments only 105

Step 2: We get λ1 = 2, λ2 = 5 are the eigen values of A.

Step 3: An eigen vector X = (x1 x2 )T corresponding to an eigen value λ

is given by (A − λI)X = 0
When λ = 2, the corresponding eigen vector is given by
    
2 1 x1 0
(A − λI)X = =
2 1 x2 0

2x1 + x2 = 0
x2 = −2x1
Take x1 = 1, then x2 = −2.
 
1
Hence X1 = is an eigen vector of A corresponding to
−2
the eigen value λ = 2.

Strictly for University Departments only 106

When λ = 5, the corresponding eigen vector is given by
    
−1 1 x1 0
(A − λI)X = =
2 −2 x2 0

−x1 + x2 = 0
x2 = x1
We take x1 = 1 so that x2 = 1.
 
1
Then X2 = is an eigen vector of A corresponding to the
1
eigen value λ = 5.
Hence the modal matrix is
 
1 1
P = [X1 X2 ] = .
−2 1

Strictly for University Departments only 107

Step 4: Diagonalization
 
1 −1
−1 1
P =
2 1 3
   
1 1 −1 4 1 1 1
P −1 AP =
3 2 1 2 3 −2 1
  
1 2 −2 1 1
=
3 10 5 −2 1
   
−1 1 6 0 2 0
P AP = = =D
3 0 15 0 5
D is a diagonal matrix containing the eigen values 2, 5 of A as
the diagonal elements.

Strictly for University Departments only 108

Step 5: To find A6 .

26 0
    
6 6 −1 1 1 1 1 −1
A = PD P =
0 56
−2 13 2 1
  
6 1 64 15625 1 −1
A =
3 −128 15625 2 1
 
1 31314 15561
=
3 31122 15753
 
6 10438 5187
∴A =
10374 5251

Strictly for University Departments only 109

Example 2: (A has distinct eigen values)  
1 6 1
Find the matrix P which transforms the matrix A =  1 2 0  to the
0 0 3
8
diagonal form. Hence calculate A .
Solution:
Step 1: The characteristic equation of A is |A − λI| = 0
1−λ 6 1
1 2−λ 0 =0
0 0 3−λ

(1 − λ)[(2 − λ)(3 − λ)] − 6(3 − λ) = 0

i.e., (λ + 1)(λ − 3)(λ − 4) = 0

Step 2: We get λ1 = −1, λ2 = 3, λ3 = 4 are the eigen values of A.
Step 3: An eigen vector X = (x1 x2 x3 )T corresponding to an eigen
value λ is
Strictly for University Departments only 110
given by (A − λI)X = 0.
When λ = −1, the corresponding eigen vector is given by
    
2 6 1 x1 0
(A − λI)X =  1 3 0   x2 = 0 
0 0 4 x3 0

2x1 + 6x2 + x3 = 0
x1 + 3x2 + 0 = 0
4x3 = 0
=⇒ x3 = 0 and x1 = −3x2
Take x2 = 1. Then x1 = −3.
 
−3
The vector X1 =  1  is an eigen vector corresponding to the
0
eigen-value λ = −1 of A.

Strictly for University Departments only 111

When λ = 3, the corresponding eigen vector is given by
    
−2 6 1 x1 0
(A − λI)X =  1 −1 0   x2 = 0 
 
0 0 0 x3 0

−2x1 + 6x2 + x3 = 0
x1 − x2 = 0
=⇒ x1 = x2 ;
and x3 = −4x2
Take x1 = 1. Then x2 = 1 and x3 = −4
 
1
The vector X2 =  1  is an eigen vector corresponding to the
−4
eigen-value λ = 3.

Strictly for University Departments only 112

When λ = 4, the corresponding eigen vector is given by
    
−3 6 1 x1 0
(A − λI)X =  1 −2 0   x2 = 0 
0 0 −1 x3 0

−3x1 + 6x2 + x3 = 0
x1 − 2x2 = 0
−x3 = 0
=⇒ x3 = 0
and x1 = 2x2
Take x2 = 1 so that x1 = 2. We have x3 = 0.
 
2
The vector X3 =  1  is an eigen vector corresponding to the eigen
0
value λ = 4.

Strictly for University Departments only 113

Let P be a matrix whose column vectors are the eigen vectors X1 , X2 , X3
of A. Thus
   
−3 1 2 p11 p12 p13
P = [X1 X2 X3 ] =  1 1 1  =  p21 p22 p23 
0 −4 0 p31 p32 p33
P is the required modal matrix.
We have to find P −1 .
1
P −1 = (adjP )
|P |
Then the co-factors are given by
P11 = p22 p33 − p23 p32 = 4; P12 = −(p21 p33 − p23 p31 ) = 0;
P13 = p21 p32 − p22 p31 = −4; P21 = −(p12 p33 − p13 p32 ) = −8;
P22 = p11 p33 − p13 p31 = 0; P23 = −(p11 p32 − p31 p12 ) = −12;
P31 = p12 p23 − p22 p13 = −1; P32 = −(p11 p23 − p21 p13 ) = 5;
P33 = p11 p22 − p21 p12 = −4;

Strictly for University Departments only 114

  T  
4 0 −4 4 −8 −1
∴ adj P =  −8 0 −12  =  0 0 5 
−1 5 −4 −4 −12 −4

|P | = −3(0 + 4) − 1(0) + 2(−4) = −20

 
4 −8 −1
1 
P −1 = 0 0 5 
−20
−4 −12 −4
Step 4: Diagonalization
Consider P −1
 the matrix D =  AP .  
4 −8 −1 1 6 1 −3 1 2
1 
D= 0 0 5  1 2 0  1 1 1 
−20
−4 −12 −4 0 0 3 0 −4 0

Strictly for University Departments only 115

   
−4 8 1 −3 1 2
1 
D= 0 0 15   1 1 1 
−20
−16 −48 −16 0 −4 0
   
20 0 0 −1 0 0
1 
= 0 −60 0 = 0 3 0 
−20
0 0 −80 0 0 4
D is a diagonal matrix which contain the eigen values -1, 3 and 4 of A as
its diagonal elements. Thus we have diagonalized A.

Step 5: To find A8 .

Since P −1 AP = D, we have A = P DP −1 .
Therefore, A8 = P D8 P −1 .

Strictly for University Departments only 116

 (−1)8 0 0
   
−3 1 2 4 −8 −1
1 
A8 = 1 1 1  0 38 0   0 0 5 
−20 8
0 −4 0 0 0 4 −4 −12 −4
   
−3 1 2 1 0 0 −4 8 1
1 
= 1 1 1   0 6561 0  0 0 −5 
20
0 −4 0 0 0 65536 4 12 4
  
−3 6561 131072 −4 8 1
1 
= 1 6561 65536   0 0 −5 
20
0 −26244 0 4 12 4
 
524300 1572840 491480
1 
= 262140 786440 229340 
20
0 0 131220
 
26215 78642 24574
∴ A8 =  13107 39322 11467 
0 0 6561
Strictly for University Departments only 117
 Example 3. (A has repeated eigen values)
 
2 2 1
Diagonalize the matrix A =  1 3 1  .
1 2 2
Solution:
Step 1: The characteristic equation of A is |A − λI| = 0.
2−λ 2 1
1 3−λ 1 = 0 i.e. λ3 − 7λ2 + 11λ − 5 = 0 is the
1 2 2−λ
characteristic equation of A..

Step 2: The eigen values of A (roots of this characteristic equation) are

λ = 1, 1, 5.

Strictly for University Departments only 118

Step 3: To
 find
the eigen vectors:
x1
Let X =  x2  be an eigen vector corresponding to the eigen value λ.
x3
Then (A − λI)X = 0.
    
2−λ 2 1 x1 0
 1 3−λ 1  x2  = 0 (27)
1 2 2−λ x3 0
When λ = 5 in equation (1), we have the following system of equations:

Strictly for University Departments only 119

 −3x1 + 2x2 + x3 = 0 (28)
x1 − 2x2 + x3 = 0 (29)
x1 + 2x2 − 3x3 = 0 (30)

Since the determinant of the coefficient matrix is zero, at the most only
two of these equations are linearly independent. Therefore, from equations
(2) and (3) we have
x1 x2 x3
= =
2+2 −(−3 − 1) 6−2
or
x1 x2 x3
= =
4 4 4
or
x1 x2 x3
= =
1 1 1

Strictly for University Departments only 120

Therefore
 an eigen vector corresponding to the eigen value λ = 5 is
1
X1 = 1
1
Any non-zero multiple of this vector is also an eigen vector corresponding
to
the eigen value λ = 5.
When λ = 1 in equation (1), we have the following system of equations:

x1 + 2x2 + x3 = 0 (31)
x1 + 2x2 + x3 = 0 (32)
x1 + 2x2 + x3 = 0 (33)

Strictly for University Departments only 121

Thus we get only one independent equation,

x1 + 2x2 + x3 = 0 (34)

We may take one of the co-ordinates to be zero, say x2 = 0 in equation

(8). Then we get x1 + x3 = 0 or x1 = −x3 . Let x3 be an arbitrary
non-zero real number, say 1. ( We cannot take x3 = 0 as this will make x1

to be zero. But an eigen vector is a non-zero vector.) Then x1 = −1.

Therefore, one 
independent
 eigen vector corresponding to the eigen value
−1
λ = 1 is X2 =  0  .
1

Strictly for University Departments only 122

To choose another linearly independent eigen vector corresponding to
λ = 1,
choose either x1 = 0 or x3 = 0 in equation (8). Suppose we take x3 = 0.
Then we get x1 + 2x2 = 0. Take x2 = 1. Then x1 = −2.
 
−2
Therefore, the second independent eigen vector for λ = 1 is X3 =  1 .
0
   
1 −1
Therefore, 5, 1, 1 are the eigen values and X1 = 1, X2 =  0 ,
1 1
 
−2
X3 =  1  are the corresponding eigen vectors of A.
0

Strictly for University Departments only 123

Step 4: Formation of the modal matrix and diagonalization
 
1 −1 −2
Now we form the modal matrix P = [X1 X2 X3 ] =  1 0 1 .
1 1 0
 
1/4 1/2 1/4
P −1 = −1/4 −1/2 3/4 
−1/4 1/2 −1/4

Consider    
1/4 1/2 1/4 2 2 1 1 −1 −2
P −1 AP = −1/4 −1/2 3/4   1 3 1   1 0 1 
−1/4 1/2 −1/4 1 2 2 1 1 0
 
5 0 0
=  0 1 0  = D.
0 0 1

Thus we have diagonalized A.

Strictly for University Departments only 124

Practice Problems

1. Diagonalize the following matrices:

 
−19 7
(i) A =
−42 16
 
−1 2 −2
(ii) A =  1 2 1 
−1 −1 0

Strictly for University Departments only 125

Multiple Choice questions

1. Two square matrices A and B are similar if

(a) A = B
(b) B = P −1 AP
(c) AT = B T
(d) A−1 = B −1
   
4 1 2 0
2. The matrices A = and B = are similar.
3 2 0 −5
(a) True
(b) False

Strictly for University Departments only 126

Answers to MCQs

1. (b)
2. (b)

Strictly for University Departments only 127

Section 5. Diagonalization using an orthogonal
transformation

Contents
Preliminaries. Orthogonal vectors
Orthogonal reduction / diagonalization using an
orthogonal matrix
Examples
Practice Problems and MCQ’s

Strictly for University Departments only 128

Preliminaries: Orthogonal vectors
  
x1 y1
 x2   y2 
    n
Definition: Two vectors X =  .  and Y = 
 
 .  in R are said

 .   . 
xn yn
to be orthogonal if their dot product X · Y = x1 y1 + x2 y2 + · · · + xn yn is
zero.
Geometrically, the vector X and the vector Y are orthogonal if they are
perpendicular in Rn .
  
1 −5
Example: The vectors X =  −2  and Y =  11  are orthogonal
9 3
in R3 as their dot product
X · Y = 1.(−5) + (−2).11 + 9.3 = −5 − 22 + 27 = 0.

Strictly for University Departments only 129

Orthogonal vectors vs linearly independent vectors

Note: Orthogonal vectors are linearly independent. For example if

X1 , X2 ∈ Rn are orthogonal, they are perpendicular to each other. So
they do not lie on the same line. Therefore they are independent. But
the converse need not be true.
   
1 0
Example: The vectors X = and Y = are linearly
1 1
independent in R2 but not orthogonal in R2 as their dot product
X · Y = 1.1 + 1.0 = 1 6= 0.
Remark: Geometrically, if two vectors in Rn are linearly independent,
then they do not lie on the same line. But the angle between the vectors
need not be 90◦ . Therefore they need not be perpendicular.

Strictly for University Departments only 130

Orthonormal system
Definition: A set of vectors X1 , X2 , X3 , . . . , Xn in Rn is said to be
orthonormal if
(i) Xi · Xj = 0 for i 6= j (i.e. Xi ’s are pairwise orthogonal),

(ii) kXi k = 1, i = 1, 2, . . . , n (i.e. each Xi has norm 1).

   
1/3 2/3
Example: The vectors X1 =  2/3 , X2 =  1/3 ,
2/3 −2/3
 
2/3
X3 =  −2/3  form an orthonormal system in R3 .
1/3
Check: X1 · X2 = X2 · X3 = X1 · X3 = 0 and kX1 k = kX2 k = kX3 k = 1.

Strictly for University Departments only 131

Diagonalization of real symmetric matrices using an
orthogonal matrix/orthogonal transformation

We know how to diagonalize a square matrix A of order n, when it has n

linearly independent eigen-vectors.

When the matrix A is a real symmetric matrix of order n with n linearly

independent eigen-vectors, we can diagonalize it using an orthogonal
matrix.

This is due to the following property of the eigen vectors of a symmetric

matrix which we have already studied.

Property 9: The eigen-vectors corresponding to distinct eigen-values of

a symmetric matrix A are orthogonal.

Strictly for University Departments only 132

Orthogonal reduction

Theorem: Let A be a real symmetric matrix of size n with n orthonormal

eigen vectors. Then we can diagonalize A using an orthogonal matrix.
Proof:Take n= 3. Letλ1 , λ2 ,
λ3 be theeigen-values
 of A. Let
x11 x12 x13
X1 =  x21 , X2 =  x22 , X3 =  x23  be the corresponding
x31 x32 x33

eigen-vectors of A which form an orthonormal system.

Strictly for University Departments only 133

  
x11 x12 x13
Let P = [X1 X2 X3 ] =  x21 x22 x23  be the matrix whose column
x31 x32 x33
vectors are the eigen-vectors of A which are pairwise orthogonal and
each eigen vector has norm 1.
  
x11 x21 x31 x11 x12 x13
T
Consider P P =  x12 x22 x32   x21 x22 x23 
x13 x23 x33 x31 x32 x33
kX1 k2
   
X1 · X2 X1 · X3 1 0 0
=  X2 · X1 kX2 k2 X2 · X3  =  0 1 0 ,
X3 · X1 X3 · X2 kX3 k2 0 0 1
the identity matrix. Similarly one can check P P T = I. Therefore P is an
orthogonal matrix. P is called the normalized modal matrix.

Strictly for University Departments only 134

Use this P for diagonalizing A.
Note that since P P T = I, we have P −1 = P T .
 
λ1 0 0
You will get P T AP = D =  0 λ2 0  which is a diagonal matrix
0 0 λ3
whose diagonal elements are the eigen values of A.

This method of diagonalizing a real symmetric matrix using an orthogonal

matrix is called orthogonal reduction of A.

Strictly for University Departments only 135

Example 1 - A has distinct eigen values
Example 1. (A has distinct eigen values)
 
1 1 3
Diagonalize the matrix A =  1 5 1  using an orthogonal
3 1 1
transformation.
Solution:
Step 1: The characteristic equation of A is |A − λI| = 0.
1−λ 1 3
1 5−λ 1 = 0 i.e. λ3 − 7λ2 + 36 = 0 is the characteristic
3 1 1−λ
equation of A.

Step 2:The eigen values of A (roots of this characteristic equation) are

λ = −2, 3, 6.

Strictly for University Departments only 136

Step 3: To
 find
the eigen vectors:
x1
Let X =  x2  be an eigen vector corresponding to the eigen value λ.
x3
Then (A − λI)X = 0.
    
1−λ 1 3 x1 0
 1 5−λ 1  x2  = 0 (35)
3 1 1−λ x3 0
When λ = −2 in equation 76, we have the following system of equations:

3x1 + x2 + 3x3 = 0 (36)

x1 + 7x2 + x3 = 0 (37)
3x1 + x2 + 3x3 = 0 (38)

Strictly for University Departments only 137

Equations (2) and (4) are the same. Therefore, from equations (2) and (3)
we have
x1 x2 x3
= =
1 − 21 −(3 − 3) 21 − 1
or
x1 x2 x3
= =
−20 0 20
or
x1 x2 x3
= =
−1 0 1
 an eigen vector corresponding to the eigen value λ = −2 is
Therefore

−1
 0 .
1
Any non-zero multiple of this vector is also an eigen vector corresponding
to the eigen value λ = −2.

Strictly for University Departments only 138

When λ = 3 in equation (76), we have the following system of equations:

−2x1 + x2 + 3x3 = 0 (39)

x1 + 2x2 + x3 = 0 (40)
3x1 + x2 − 2x3 = 0 (41)

As the determinant of the coefficient matrix is zero, at the most only two
of these equations are linearly independent. Therefore, from equations
(6) and (7) we have
x1 x2 x3
= =
(−4 − 1) −(−2 − 3) 1−6
or
x1 x2 x3
= =
−5 5 −5
or
x1 x2 x3
= =
1 −1 1

Strictly for University Departments only 139

Therefore
  an eigen vector corresponding to the eigen value λ = 3 is
1
−1.
1
When λ = 6 in equation (76), we have the following system of equations:

−5x1 + x2 + 3x3 = 0 (42)

x1 − x2 + x3 = 0 (43)
3x1 + x2 − 5x3 = 0 (44)

As the determinant of the coefficient matrix is zero, at the most only two
of these equations are linearly independent. Therefore, from equations
(9) and (10) we have
x1 x2 x3
= =
(5 − 1) −(−5 − 3) 1+3

Strictly for University Departments only 140

or
x1 x2 x3
= =
4 8 4
or
x1 x2 x3
= =
1 2 1
 
1
Therefore an eigen vector corresponding to the eigen value λ = 6 is 2

1
Therefore,
  −2, 3, 6 are the
  eigen values and
 
−1 1 1
0
X1 =   0   0
0 , X2 = −1 , X3 = 2 are the corresponding

1 1 1
eigen vectors of A.
Step 4: Form the normalized modal matrix
Note that these eigen vectors X10 , X20 , X30 are pairwise orthogonal as they
correspond to distinct eigen values −2, 3, 6 of the symmetric matrix A.

Strictly for University Departments only 141

Let us normalize these vectors.
 √   √ 
0 −1/ 2 0 1/ √3
X1  0  , Let X2 = X2 = −1/ 3, Let
Let X1 = = √ √
kX10 k kX20 k
1/ 2 1/ 3
 √ 
1/√6
X30
X3 = = 2/√6 .
kX30 k
1/ 6
X1 , X2 , X3 are orthonormal.
 √ √ √ 
−1/ 2 1/ √3 1/√6
Take P = [X1 X2 X3 ] =  0√ −1/√ 3 2/√6 .
1/ 2 1/ 3 1/ 6

Then P is an orthogonal matrix. (Therefore, P −1 = P T ).

P is the normalized modal matrix.

Strictly for University Departments only 142

Step 5: Diagonalization by an orthogonal matrix
Now
 consider
√ P −1 AP =P T
√ AP
  √ √ √ 
−1/√ 2 0√ 1/√2 1 1 3 −1/ 2 1/ √3 1/√6
= 1/√3 −1/√ 3
 1/√31 5 1  0√ −1/√ 3 2/√6
1/ 6 2/ 6 1/ 6 3 1 1 1/ 2 1/ 3 1/ 6
 
−2 0 0
= 0 3 0=D
0 0 6

which is a diagonal matrix whose diagonal entries are the eigen values of
A. Thus we have diagonalized A using an orthogonal matrix P .

Strictly for University Departments only 143

Example 2 - A has repeated eigen values
 
1 −2 2
Example 2: Diagonalize the matrix A =  −2 1 −2  using an
2 −2 1
orthogonal transformation.
Solution:
Step 1: The characteristic equation of A is |A − λI| = 0.
1 − λ −2 2
−2 1 − λ −2 = 0 i.e. λ3 − 3λ2 − 9λ − 5 = 0 is the
2 −2 1 − λ
characteristic equation of A.

Step 2:The eigen values of A (roots of this characteristic equation) are

λ = −1, −1, 5. Note that the eigen value −1 is repeated two times.

Strictly for University Departments only 144

Step 3: To
 find
the eigen vectors:
x1
Let X =  x2  be an eigen vector corresponding to the eigen value λ.
x3
Then AX = λX or (A − λI)X = 0.
    
1 − λ −2 2 x1 0
 −2 1 − λ −2  x2  = 0 (45)
2 −2 1 − λ x3 0
When λ = 5 in equation (45), we have the following system of equations:

−4x1 − 2x2 + 2x3 = 0 (46)

−2x1 − 4x2 − 2x3 = 0 (47)
2x1 − 2x2 − 4x3 = 0 (48)

Strictly for University Departments only 145

From equations (46) and (47) we have
x1 x2 x3
= =
4+8 −(8 + 4) 21 − 1
or
x1 x2 x3
= =
1 −1 1
Therefore
 an  eigen vector corresponding to the eigen value λ = 5 is
1
X10 = −1
1
Any non-zero multiple of this vector is also an eigen vector corresponding
to the eigen value λ = 5.

Strictly for University Departments only 146

When λ = −1 in equation (45), we have the following system of equations:

2x1 − 2x2 + 2x3 = 0 (49)

−2x1 + 2x2 − 2x3 = 0 (50)
2x1 − 2x2 + 2x3 = 0 (51)

All these equations are the same. They are equal to

x1 − x2 + x3 = 0. (52)

In equation (52), take x3 = 0. We get x1 = x2 . if we take x2 = 1 then

x1 = 1.  
1
Let X20 = 1. This is one linearly independent eigen vector
0
corresponding to the eigen value λ = −1.

Strictly for University Departments only 147

  
x
We have to find another linearly independent eigen vector X30 = y 
z
corresponding to λ = −1 such that X20 · X30 = 0.
i.e X30 should satisfy
 equation
 (52) and
x
X20 · X30 = (1 1 0) y  = 1.x + 1.y + 0.z = 0.
z
Therefore, we have to solve the equations x − y + z = 0 and x + y = 0
simultaneously. From these two equations we have,
x y z
= =
−1 1 2

Strictly for University Departments only 148

Therefore the second
 eigen
 vector corresponding to the eigen value
−1
λ = −1 is X30 =  1 .
2
These eigen vectors satisfy X10 · X20 = 0, X20 · X30 = 0, X30 · X10 = 0.

They are pairwise orthogonal.

Step 4: Form the normalized modal matrix

Let us normalize these eigen vectors.

Strictly for University Departments only 149

  √   √ 
1/ √3 1/√2
X10 X20
Let X1 = = −1/√ 3 , X2 =
  = 1/ 2,

kX10 k kX20 k
1/ 3 0
 √ 
−1/√ 6
X30
X3 = =  1/ 6  .
√
kX30 k
2/ 6
X1 , X2 , X3 are orthonormal.
√ √ √ 
1/ √3 1/√2 −1/√ 6
Take P = [X1 X2 X3 ] = −1/√ 3 1/ 2 1/√6  .
1/ 3 0 2/ 6

Then P is an orthogonal matrix. (Therefore, P −1 = P T ).

P is the normalized modal matrix.

Strictly for University Departments only 150

Step 5: Diagonalization by an orthogonal matrix
Now consider P −1 AP = P T AP
 √ √ √  
1/√3 −1/√ 3 1/ 3 1 −2 2
= 1/ √2 1/√2 0√ −2 1 −2
−1/ 6 1/ 6 2/ 6 2 −2 1
 √ √ √   
1/ √3 1/√2 −1/√ 6 5 0 0
−1/ 3 1/ 2 1/ 6  =0
√ √ −1 0 =D
1/ 3 0 2/ 6 0 0 −1

which is a diagonal matrix whose diagonal entries are the eigen values of
A. Thus we have diagonalized A using an orthogonal matrix P .

Remark. This procedure has immediate application in the next section.

Strictly for University Departments only 151

Practice Problems

Diagonalize the following symmetric matrices using orthogonal

transformations:
 
1 1 3
(i) A =  1 5 1 
3 1 1
 
6 −2 2
(ii) A =  −2 3 −1 
2 −1 3

Strictly for University Departments only 152

Multiple Choice Questions

1. A set of linearly independent vectors are pairwise orthogonal.

(a) True
(b) False

2. If P is orthogonal then P −1 is
(a) P
(b) P T
(c) P −1 does not exist
(d) None of the above

Strictly for University Departments only 153

Answers to MCQ’s

1. (b)
2. (b)

Strictly for University Departments only 154

Section 6. Quadratic forms

Contents
Quadratic forms and symmetric matrices
Orthogonal reduction of a quadratic form to its
canonical form
Rank, index, signature and nature of a quadratic form
Examples
Practice Problems and MCQs

Strictly for University Departments only 155

Quadratic forms

Definition: A homogeneous polynomial of second degree in any

number of variables is called a quadratic form.
Examples:
1. ax2 + by 2 + cz 2 + 2hxy + 2gyz + 2f zx is a quadratic form in 3
variables x, y, z.
2. ax2 + by 2 + 2hxy is a quadratic form in 2 variables x and y.
Definition: The general quadratic form in n variables x1 , x2 , x3 , ...xn is
of the form
n X
X n
bij xi xj .
j=1 i=1

Strictly for University Departments only 156

A quadratic form and its associated real symmetric matrix
n X
X n
Let bij xi xj be a quadratic form in x1 , x2 , . . . , xn .
j=1 i=1

Define aij = 12 (bij + bji ). Then we have aij = aji .

We can write (bij xi xj + bji xj xi ) = (aij + aji )xi xj .
Therefore we can re-write the quadratic form as

X n
n X n X
X n
bij xi xj = aij xi xj .
j=1 i=1 j=1 i=1

Let A = (aij ), 1 ≤ i ≤ n; 1 ≤ j ≤ n. Then A is a real symmetric matrix.

 
x1
 x2 
Let X = 
 .. 


xn

Strictly for University Departments only 157

⇒XT = (x1 x2 . . . xn )
   
a11 a12 ... a1n x1
 a21 a22 ... a2n   x2 
X T AX = (x1 x2 . . . xn )  .
   
.. .. ..   .. 
 .. . . .   . 
an1 an2 · · · ann xn
n X
X n Xn Xn
= aij xi xj = bij xi xj
j=1 i=1 j=1 i=1

which is the given quadratic form.

n X
X n
Therefore we can write the given quadratic form bij xi xj
j=1 i=1

as X T AX , where aij = 12 (bij + bji ) and A=(aij ) is a real symmetric

matrix.
Thus we can associate a real symmetric matrix A with a quadratic form.

Strictly for University Departments only 158

Remark:
In general, given a quadratic form in n variables x1 , x2 , . . . , xn , the
associated real symmetric matrix A = (aij ) is constructed like this:
First write down the diagonal elements:

aii = coefficient of x2i .

The non-diagonal elements are then written as follows:

1
For i 6= j, aij = aji = coefficient of xi xj .
2

Strictly for University Departments only 159

Examples

Example 1. Write down the matrices of the following quadratic forms:

(i) ax21 + bx22 + 2hx1x2 
x1
Solution: Let X = . The given quadratic form is in two variables
x2
x1 , x2 . Therefore the associated real symmetric matrix is a 2 × 2 matrix.
 
a11 a12
Let A = . We take
a21 a22

a11 = coefficient of x21 , a22 = coefficient of x22 ,

1
a12 = a21 = coefficient of x1 x2 .
2
 
a h
Hence A = is the associated real symmetric matrix.
h b

Strictly for University Departments only 160

(ii) 2x2 + 3y 2 + 6xy
Solution:
 
2 3
The associated real symmetric matrix is A= .
3 3

(iii) 2x2 + 5y 2 − 6z 2 − 2xy − yz + 8zx.

Solution:
2 −1 4
 
−1 
−1 5

The associated real symmetric matrix is A =  2 .

−1

4 −6
2

Strictly for University Departments only 161

(iv) x21 + 2x22 + 3x23 + 4x24 + 2x1 x2 + 4x1 x3 − 6x1 x4 − 4x2 x3 − 8x2 x4
+12x3 x4 .
Solution:  
1 1 2 −3
1 2 −2 −4
The associated real symmetric matrix is A = 
 2 −2 3
.
6
−3 −4 6 4

Strictly for University Departments only 162

From a real symmetric matrix to a quadratic form
Suppose we are given a real symmetric matrix A = (aij ) of size n.
We will write a quadratic form in n variables say x1 , x2 , . . . , xn as
follows:
 
x1
 x2 
 .. . Then
Let X =  

xn
   
a11 a12 ... a1n x1
 a21 a22 ... a2n   x2 
X T AX = (x1 x2 . . . xn )  .
   
.. .. ..   .. 
 .. . . .   . 
an1 an2 · · · ann xn
n X
X n Xn Xn
= aij xi xj = bij xi xj
j=1 i=1 j=1 i=1

is a quadratic form.
Strictly for University Departments only 163
Here the coefficient bii of x2i is aii , for i = 1, 2, ..., n.

And for i 6= j, the coefficient bij of xi xj is 2 aij .

Thus the quadratic form given by the real symmetric matrix A is

a11 x21 +a22 x22 +· · ·+ann x2n +2a12 x1 x2 +2a13 x1 x3 +· · ·+2a(n−1)n xn−1 xn .

One can check the following:

There is a one-one correspondence between the set of all real symmetric
matrices of size n and the set of all quadratic forms in n variables.

Strictly for University Departments only 164

Examples
Example 2:
Write down the quadratic form corresponding to the following
symmetric matrices.
 
2 4 5
(i) A= 4 3 1
5 1 1

 A is a 3 × 3 matrix. We need three variables say x1 , x2 , x3 . Let

Solution:
x1
X = x2 . The required quadratic form is
x3
  
2 4 5 x1
X T AX = (x1 x2 x3 )4 3 1x2 
5 1 1 x3

= 2x21 + 3x22 + x23 + 8x1 x2 + 10x1 x3 + 2x2 x3 .

Strictly for University Departments only 165
  
1 1 −2 0
 1 −4 0 6
(ii) A=
−2 0

6 −3
0 6 −3 2

Solution: It is a quadratic form in 4 variables,say x1 , x2 , x3 , x4 .

X T AX = x21 − 4x22 + 6x23 + 2x24 + 2x1 x2 − 4x1 x3 + 12x2 x4 − 6x3 x4 .

This is the required quadratic form.

Strictly for University Departments only 166

Reduction of a quadratic form to its canonical form or sum
of squares
 form:

x1
 x2 
Let X = 
 .. . Let A = (aij )n×n be a symmetric matrix.


xn
Consider the quadratic form
n X
X n
T
X AX = aij xi xj .
j=1 i=1

Let λ1 , λ2 , ..., λn be the eigen values of A and X1 , X2 , ..., Xn be

the corresponding eigen vectors.
Assume that X1 , ...Xn are pair wise orthogonal and kXi k = 1, i=1,2...,n.

Since A is a real symmetric matrix we know how to diagonalize A

using an orthogonal transformation.
Strictly for University Departments only 167
If we take P = [X1 X2 ...X
 n ], then P is an orthogonal
 matrix such that
λ1 0 0 ... 0
 0 λ2 0 ... 0 
P T AP = D where D =   ... .. ...
 a diagonal matrix whose
.. ... 
0 0 0 ... λn
diagonal elements are the eigen values of A.
P T AP = D =⇒ A=P DP T .
Note as P is orthogonal, we have P −1 = P T .
Now we can write the given quadratic form as
X T AX = X T P DP T X = (P T X)T D(P T X)

Let Y = P T X. Then we have,

Strictly for University Departments only 168

X T AX = Y T DY  
λ1 y1
 λ2   y2 
=(y1 y2 . . . yn ) 
   
..   .. 
 .  .
λn yn

=λ1 y12 + λ2 y22 + . . . + λn yn2

which is the canonical form or sum of squares form of the given

quadratic form.

Strictly for University Departments only 169

Procedure for orthogonal reduction of a quadratic form to
its canonical form

Quadratic form
(in x1 , x2 , . . . , xn ) −→ real symmetric matrix A

↓ orthogonal reduction

canonical form ←− Diagonal matrix D

(quadratic form
in y1 , y2 , . . . , yn ) (a real symmetric matrix)

Here Y = P T X where P is an orthogonal matrix with P T AP = D.

Strictly for University Departments only 170

Distinct eigen values
Example 3: Reduce the quadratic form
8x2 + 7y 2 + 3z 2 − 12xy + 4xz − 8yz into its canonical form using an
orthogonal transformation.
Solution: Step
 1: The real symmetric
 matrix associated to the quadratic
8 −6 2
form is A =  −6 7 −4 
2 −4 3

Step 2: The characteristic equation of A is |A − λI| = 0.

8 − λ −6 2
−6 7 − λ −4 = 0.
2 −4 3 − λ
i.e. λ3 − 18λ2 + 45λ = 0 is the characteristic equation of A.
The eigen values of A are λ = 0, 3, 15.

Strictly for University Departments only 171

Note : All the eigen values of A are distinct.
Since all the eigen values of the symmetric matrix A are
distinct,the eigen vectors of A are pairwise orthogonal.
Step 3: Finding
 the eigen vectors of A.
x1
Let X =  x2  be an eigen vector corresponding to the eigen value λ.
x3
Then X satisfies the equation AX = λX or (A − λI)X = 0.
    
8 − λ −6 2 x1 0
 −6 7 − λ −4  x2  = 0 (53)
2 −4 3 − λ x3 0

(8 − λ)x1 − 6x2 + 2x3 = 0 (54)

−6x1 + (7 − λ)x2 − 4x3 = 0 (55)
2x1 − 4x2 + (3 − λ)x3 = 0 (56)

Strictly for University Departments only 172

Take λ = 0 in (1) we get

8x1 − 6x2 + 2x3 = 0 (57)

−6x1 + 7x2 − 4x3 = 0 (58)
2x1 − 4x2 + 3x3 = 0 (59)
As the determinant of the coefficient matrix is zero, at the most only two
of these equations are linearly independent. From equations (6) and (7)
we have
x1 x2 x3
= =
(21 − 16) −(−18 + 8) 24 − 14
x1 x2 x3
= =
5 10 10
x1 x2 x3
= =
1 2 2
 
1
Let X10 = 2
2
Strictly for University Departments only 173
 √
kX10 k = 1+4+4=3
 
1/3
X10
Let X1 = = 2/3

kX10 k
2/3
Now take λ = 3 in (1) we get
5x1 − 6x2 + 2x3 = 0 (60)
−6x1 + 4x2 − 4x3 = 0 (61)
2x1 − 4x2 + 0x3 = 0 (62)
From equations (9) and (10) we get
x1 x2 x3
= =
(−16) −(8) 16
x1 x2 x3
= =
2 1 −2
 
2
Let X20 =  1 
−2
Strictly for University Departments only 174
 √
kX20 k = 4+1+4=3
 
2/3
X20
X2 = =  1/3 
kX20 k
−2/3
Now take λ = 15 in (1) we get

−7x1 − 6x2 + 2x3 = 0 (63)

−6x1 − 8x2 − 4x3 = 0 (64)
2x1 − 4x2 − 12x3 = 0 (65)

from equations (12) and (13) we get

x1 x2 x3
= =
(96 − 16) −(72 + 8) 24 + 16
x1 x2 x3
= =
80 −80 40
x1 x2 x3
= =
2 −2 1
Strictly for University Departments only 175
  
2
Let X30 = −2
1
0
√
kX3 k = 4 + 1 + 4 = 3
 
2/3
X30
X3 = = −2/3
kX30 k
1/3

Step 4: Diagonalization of A by an orthogonal matrix

Note that X1 , X2 , X3 form an orthonormal set. i.e. X1 , X2 , X3 are
pairwise orthogonal and kX1 k = kX2 k = kX3 k = 1.
 
1/3 2/3 2/3
Take P = [X1 X2 X3 ] = 2/3 1/3 −2/3 .
2/3 −2/3 1/3

Then P is an orthogonal matrix. P P T implies P −1 = P T .

Strictly for University Departments only 176

Now consider P −1 AP = P T AP

   
1/3 2/3 2/3 8 −6 2 1/3 2/3 2/3
= 2/3
 1/3 −2/3−6 7 −4 2/3 1/3 −2/3
2/3 −2/3 1/3 2 −4 3 2/3 −2/3 1/3
 
0 0 0
=0 3 0 =D
0 0 15

which is a diagonal matrix whose diagonal entries are the eigen

values of A. Thus we have diagonalized A using an orthogonal
matrix P .

Strictly for University Departments only 177

Step 5: Canonical form

The given quadratic form is

X T AX = X T P DP T X = (P T X)T DP T X
 
y1
Let Y =  y2  = P T X
y
  3  
y1 1/3 2/3 2/3 x1
Y = y2  = 2/3 1/3 −2/3 x2 
y3 2/3 −2/3 1/3 x3

Then   
0 0 0 y1
X T AX = Y T DY = (y1 y2 y3 ) 0 3 0  y2  = 0y12 + 3y22 + 15y32
0 0 15 y3

Strictly for University Departments only 178

Therefore we have
X T AX = 0y12 + 3y22 + 15y32 which has only square terms.

This is the canonical form of the given quadratic form. Here

y1 = x1 /3 + 2x2 /3 + 2x3 /3
y2 = 2x1 /3 + x2 /3 − 2x3 /3
y3 = 2x1 /3 − 2x2 /3 + x3 /3.

Strictly for University Departments only 179

Repeated eigen values
Example 4: Reduce the quadratic form 2x21 + 6x22 + 2x23 + 8x1 x3 to its
canonical form using an orthogonal transformation .

Solution:
 Step 1:
 The real symmetric matrix of the quadratic form is
2 0 4
A= 0 6 0 
4 0 2
Step 2: The characteristic equation of A is |A − λI| = 0.

2−λ 0 4
0 6−λ 0 =0
4 0 2−λ
i.e. λ3 − 10λ2 + 12λ + 72 = 0

The eigen values of A are λ = −2, 6, 6.

Strictly for University Departments only 180
Note :The eigen value λ = 6 is repeated 2 times.

Step 3: We find the eigen vectors of A.

 
x1
Let X =  x2  be an eigen vector of A corresponding to the eigen
x3
value λ of A. Then X satisfies the equation
AX=λX or (A − λI)X = 0.
    
2−λ 0 4 x1 0
 0 6−λ 0   x2 = 0
  (66)
4 0 2−λ x3 0

(2 − λ)x1 + 0x2 + 4x3 = 0 (67)

0x1 + (6 − λ)x2 + 0x3 = 0 (68)
4x1 + 0x2 + (2 − λ)x3 = 0 (69)

Strictly for University Departments only 181

Take λ = −2 in (14) we get

4x1 + 0x2 + 4x3 = 0 (70)

0x1 + 8x2 + 0x3 = 0 (71)
4x1 + 0x2 + 4x3 = 0 (72)

from equations (19) and (20) we get

x1 x2 x3
= =
(32 − 0) (0 − 0) 0 − 32
x1 x2 x3
= =
32 0 −32
x1 x2 x3
= =
1 0 −1
 
1
Let X1 =  0 
−1

Strictly for University Departments only 182

Take λ = 6 in (14) we get

−4x1 + 0x2 + 4x3 = 0 (73)

0x1 + 0x2 + 0x3 = 0 (74)
4x1 + 0x2 − 4x3 = 0 (75)

Equations (21) and (23) reduce to the same equation x1 − x3 = 0.

Thus we get x1 = x3 . Here x2 is arbitrary.
Let x3 = 1. This gives x1 = 1. Choose x2 to be zero.
 
1
Let X2 = 0
1
This is one linearly independent eigen vector corresponding to λ = 6.

Strictly for University Departments only 183

  
x
We have to find another linearly independent eigen vector X3 = y 

z
such that X2 and X3 are orthogonal. That is X3 should satisfy
the equations (21) (22), (23) and X3 .X2 should be zero.
(Note: We can find another linearly independent vector
corresponding
  to λ = 6 by taking y = 1 and x = 1, z = 1. But the vector
1
X3 = 1 is not orthogonal to X2 as
1
X3 · X2 = X3T X2 = 1.1 + 1.0 + 1.1 = 2 6= 0. Therefore we have to
impose the condition X2 · X3 = 0)
If X3 satisfies equations (21), (22), (23), then x − z = 0 and y arbitrary.
 
1
If X3T X2 = 0, then X3T X2 = (x y z) 0 = x + z = 0.
1
Strictly for University Departments only 184
To find X3 we solve the system of equations x − z = 0
and x + z = 0 simultaneously.
We get x = z = 0. Recall y is arbitrary. Let y = 1.
 
0
Let X3 = 1 .
0

Note that X1 · X3 will be zero since X1 and X3 correspond to distinct

eigen values −2 and 6 of the symmetric matrix A. For the same
reason X1 · X2 = 0. Now we have X1 · X2 = 0 , X2 · X3 = 0 and
X1 · X3 = 0

Therefore X1 , X2 , X3 are pairwise orthogonal.

step 4: Orthogonal reduction of A. 
X1 X2 X3
Let P =
kX1 k kX2 k kX3 k

Strictly for University Departments only 185

 √ √
kX1 k = √1 + 0 + 1 = √2
kX2 k = √1 + 0 + 1 = √2
kX3 k = 0 + 1 + 0 = 1
 1 
√ √1 0
2 2
P = 0 0 1.
 
−1 1
√
2
√
2
0

Then P is orthogonal. Therefore P P T = I. ⇒ P −1 = P T

Consider 
−1
   √1 
√1 0 √ √1 0

2 0 4
 √12 2
√1  0
 2 2
P T AP = 0 6 0  0 0 1
 
 2 2
−1 √1
0 1 0 4 0 2 √
2 2
0

Strictly for University Departments only 186

  
−2 0 0
=  0 6 0 = D.
0 0 6
Thus we get P T AP = D or A = P DP T .
Consider the given quadratic form X T AX.
Now X T AX = X T (P DP T )X = X T (P T )T DP T X = (P T X)T DP T X.
Let Y = P T X. Then
 1  
√ √1 0
 
y1 2 2 x1
T
Y =  y2  = P X =  0

0 1 x2 

−1 √1
y3 √
2 2
0 x3
We can write the given quadratic
 form as   
-2 0 0 y1
X T AX = Y T DY = (y1 y2 y3 )  0 6 0  y2  = −2y12 + 6y22 + 6y32 .
0 0 6 y3

Strictly for University Departments only 187

Thus X T AX = −2y12 + 6y22 + 6y32 .
This is the canonical form of the given quadratic form.
Here the original variables x1 , x2 , x3 and the new variables y1 , y2 , y3
are related by the equation Y = P T X or X = P Y.
Thus
 we have
 1  
x1 √ √1 0 y1
2 2
x2  =  0

0 1 y2
 
−1 √1
x3 √
2 2
0 y3
x1 = √1 (y1 + y2 )
2

x2 = y3

x3 = √1 (−y1 + y2 ).
2

Strictly for University Departments only 188

Rank, Index, Signature and Nature
Definitions:
Let X T AX be a quadratic form .
1. Rank of a quadratic form
= number of non-zero eigen values of A
= number of non-zero terms in the canonical form

2. Index of a quadratic form

= number of positive terms in the canonical form
= number of positive eigen values of A

3. Signature of a quadratic form

= (number of positive terms) − (number of negative terms) in the
canonical form
= (number of positive eigen values of A) − (number of negative eigen
values of A)

Strictly for University Departments only 189

4. Nature of the quadratic form:
A quadratic form X T AX is said to be
a) Positive definite if all the eigen values of A are positive.

b) Positive semi-definite if all the eigen values of A are ≥ 0 and atleast

one of them is zero.

c) Negative definite if all the eigen values of A are negative.

d) Negative semi-definite if all the eigen values of A are ≤ 0 and atleast

one of them is zero.

e) Indefinte if some eigen values of A are positive and some negative.

Strictly for University Departments only 190

Repeated eigen values
Example 5. Reduce the quadratic form
x21 + x22 + x23 − 4x1 x2 + 4x1 x3 − 4x2 x3 to its canonical form using an
orthogonal transformation. Also find the rank, index, signature and nature
of the quadratic form.
Solution:
 Step 1: Thesymmetric matrix of the quadratic form is
1 −2 2
A =  −2 1 −2 
2 −2 1
Step 2: The characteristic equation of A is |A − λI| = 0.

1 − λ −2 2
−2 1 − λ −2 =0
2 −2 1 − λ
The characteristic equation of A is λ3 − 3λ2 − 9λ − 5 = 0
The eigen values of A are λ = −1, −1, 5.

Strictly for University Departments only 191

Note :The eigen value λ = −1 is repeated 2 times.
Step 3: we will find the eigen vectors of A.
 
x1
Let X =  x2  be an eigen vector of A corresponding to the eigen
x3
value λ of A.
X satisfies the equation AX=λ X or (A − λI)X = 0.
    
1 − λ −2 2 x1 0
 −2 1 − λ −2  x2  = 0 (76)
2 −2 1 − λ x3 0

(1 − λ)x1 − 2x2 + 2x3 = 0 (77)

−2x1 + (1 − λ)x2 − 2x3 = 0 (78)
2x1 − 2x2 + (1 − λ)x3 = 0 (79)

Strictly for University Departments only 192

Take λ = 5 in (24) we get

−4x1 − 2x2 + 2x3 = 0 (80)

−2x1 − 4x2 − 2x3 = 0 (81)
2x1 − 2x2 − 4x3 = 0 (82)

From equations (29) and (30) we get

x1 x2 x3
= =
(4 + 8) −(8 + 4) 16 − 4
x1 x2 x3
= =
12 −12 12
x1 x2 x3
= =
1 −1 1
 
1
Let X1 = −1
1

Strictly for University Departments only 193

Take λ = −1 in (24) we get

2x1 − 2x2 + 2x3 = 0 (83)

−2x1 + 2x2 − 2x3 = 0 (84)
2x1 − 2x2 + 2x3 = 0 (85)

All the above equations are equal to the same equation

x1 − x2 + x3 = 0

Take x3 = 0. Then we get x1 = x2

Let x2 = 1 ⇒ x1 = 1.
 
1
Let X2 = 1
0

Strictly for University Departments only 194

  −1.
This is one linearly independent eigen vector corresponding to λ = 
x
We have to find another linearly independent eigen vector X3 = y 
z
such that X3 satisfies equations (31), (32), (33) and X3 · X2 = 0.
(Note: We can find another linearly independent vector
correspondingto λ= −1 by taking y = 0 and x = 1, z = −1. But the
1
vector X3 =  0  is not orthogonal to X2 as
−1
X3 · X2 = X3T X2 = 1.1 + 1.0 + 0.(−1) = 1 6= 0. Therefore we have to
impose the condition X2 · X3 = 0

Strictly for University Departments only 195

If X3 satisfies equations (31), (32), (33) then we have x − y + z = 0.

From X3 · X2 = X3T X2 = 0 we have,

 

1
x y z  1 =0
0
=⇒ x + y = 0

We solve the equations x − y + z = 0 and x + y = 0 simultaneously.

x y z
= =
−1 1 2

Strictly for University Departments only 196

  
−1
Let X3 =  1
2

Note that X1 · X3 is zero since X1 and X3 correspond to distinct eigen

values 5 and −1 of the symmetric matrix A. For the same reason
X1 · X2 = 0. Now we have X1 · X2 = 0 , X2 · X3 = 0 and X1 · X3 = 0.
Therefore X1 , X2 , X3 are pairwise orthogonal.

step 4: Diagonalization of A by an orthogonal matrix.

 
X1 X2 X3
Let P =
kX1 k kX2 k kX3 k

Strictly for University Departments only 197

 √ √
kX1 k = √1 + 1 + 1√= 3
kX2 k = √1 + 1 = 2 √
kX3 k = 1 + 1 + 4 = 6

 √ √ √ 
1/ √3 1/√2 −1/√ 6
P =−1/√ 3 1/ 2 1/√ 6 
1/ 3 0 2/ 6
P is orthogonal. P P T = I implies P −1 = P T

Consider 
√1 −1 √1   √1 √1 −1
 
√ √
3 3 3 1 −2 2 3 2 6
 √1 √1  −1 √1 √1 
P T AP =  2 2
0 −2 1 −2  √3 2 6
−1 √1 √2 2 −2 1 √1 1
√
6 6 6 3
0 √
6

Strictly for University Departments only 198

  
5 0 0
= 0 −1 0  = D
0 0 −1
Thus we have P T AP = D or A = P DP T .
Step 5: Canonical form
The given quadratic form is X T AX.
We can write X T AX = X T P DP T X = (P T X)T DP T X.
 
y1
Let Y =  y2  = P T X.
y3
 1 −1 1   
√ √ √
3 3 3 x1
T  √1 √1 0
Y =P X= 2 x2
 
2 
−1
√ √1 √2 x3
6 6 6

Strictly for University Departments only 199

Now we can write the given quadratic
 form as   
5 0 0 y1
T T
X AX = Y DY = (y1 y2 y3 ) 0 −1 0   y2  = 5y12 − y22 − y32 .
0 0 −1 y3
X T AX = 5y12 − y22 − y32 is the canonical form of the given quadratic form.
Here the original variables x1 , x2 , x3 and the new variables y1 , y2 , y3
are related by the equation Y = P T X. Thus we have
y1 = √1 (x1 − x2 + x3 )
3
y2 = √1 (x1 + x2 )
2
y3 = √1 (−x1 + x2 + 2x3 ).
6

Rank = 3; Index = 1; Signature = 1 − 2 = −1; Nature : Indefinite

(Note that to find the rank, index, signature, nature of the quadratic form
it is enough to know the eigen values of the real symmetric matrix
associated with the quadratic form.)
Strictly for University Departments only 200
Practice Problems

1. Reduce the quadratic form 6x21 + 3x22 + 3x23 − 4x1 x2 − 2x2 x3 + 4x3 x1
into its canonical form .

2. Find the rank, index, signature and nature of the following

quadratic forms:
a) 2x21 + 2x22 + 3x23 + 2x1 x2 − 4x1 x3 − 4x2 x3
Ans:
√ √
1, 3+2 2, 3 − 2 2
rank = index = signature = 3, nature : positive definite

Strictly for University Departments only 201

b) 2x1 x2 + 2x1 x3 + 2x2 x3
Ans:
λ = 2, −1, −1
rank = 3, index = 1, signature = −1, nature : Indefinite
3. Reduce the quadratic form 3x2 + 5y 2 + 3z 2 − 2yz + 2zx − 2xy into its
canonical form. Find the rank,index,signature and nature of the quadratic
form.

Strictly for University Departments only 202

Multiple Choice Questions
 
1 2
1.The quadratic form corresponding to the symmetric matrix is
2 −4
(a) x2 + 4xy − 4y 2
(b) x2 − 4xy + 4y 2
(c) 4x2 − 4xy + y 2
(d) 4x2 + 4xy − y 2
2. The matrix of the quadratic form q = 4x2 − 2y 2 + z 2 − 2xy + 6zx is
   
4 −1 3 4 1 3
(a)−1 −2 0 (b) 1 −2 0
3 0 1 3 0 1
   
4 1 3 4 1 3
(c) −1 2 0 (d) 1 2 0
3 0 1 3 0 1

Strictly for University Departments only 203

3. The nature of the quadratic form
x21 + 4x22 + x23 − 4x1 x2 + 2x1 x3 − 4x2 x3 is
(a) positive definite
(b) negative definite
(c) indefinite
(d) positive semi-definite

Strictly for University Departments only 204

Answer to MCQs

1. (a)
2. (a)
3. (d)

Strictly for University Departments only 205