Anda di halaman 1dari 38

UNIT-I

THEORY OF MATRICES
Real Matrices:
REAL MATRIX: A Matrix A= ( aij ) is said to be a real matrix if every element
real number.

ij

of A is a

3
1 2

Ex: 8 0 1
5 6 9

SYMMETRIC MATRIX: A Matrix A= ( aij ) is said to be a symmetric matrix if

aij = aji for i and j .Thus A is symmetric matrix if A=AT.


1 2 3
Ex: 2 6 9
3 9 7

SKEW-SYMMETRIC MATRIX: A real Matrix A= ( aij ) is said to be a skew-symmetric


matrix if aij= -aji for every i and j. Thus A is skew-symmetric if
A=-AT.
Note: Every diagonal element of a skew-symmetric matrix is necessarily zero.

a b
0

c
Ex: a 0
b c 0
ORTHOGONAL MATRIX: A real Matrix A= ( aij ) is said to be a orthogonal matrix
If A-1=-AT.
1 2 2
1
2 1 2
Ex:

3
2 2 1

Complex Matrices:
Complex matrix: A matrix A is called a complex matrix if the elements of A are complex.
5
2 3i
Ex:

6 7i 5 i
Conjugate complex matrix: If the elements of the matrix of A are replaced by their conjugate
complexes then the resulting matrix is defined as conjugate complex matrix.It is denoted by A
bar
5
5
2 3i
2 3i
Ex: If A=
, then conjugate of A =

6 7i 5 i
6 7i 5 i
Hermition:: A complex square matrix A is said to be a Hermition if A= A = A T

A=

2 1+j 2-j
1-j 1 j
2+j -j 1

Skew-Hermition: A complex square matrix A is said to be skew


skew-Hermition
Hermition if A= - A T

Unitary: A complex square matrix A is said to Unitary if A-1=conjugate (AT )

Idempotent Matrix: A square matrix is said to be Idempotent if A2=A

Elementary row Transformation:


1) Interchange of two rows, if it h row and jt h row
w are interchanged it is denoted as RiRj
Ri

2) Multiplication of each element of a row with a non-zero scalar if ith row is multiplied with K
then it is denoted as RiKRi
3) Multiplying every element of a row with a non-zero scalar and adding to the corresponding
elements of another row. If all the elements of ith row are multiplied with k and added to the
corresponding elements of jth row then it is denoted by RjRj+kRi
Elementary column Transformation:
1) Interchange of two rows, if it h column and jt h column are interchanged it is denoted as CiCj
2) Multiplication of each element of a column with a non-zero scalar if ith column is multiplied with
K then it is denoted as CiKCi
3) Multiplying every element of a column with a non-zero scalar and adding to the corresponding
elements of another column. If all the elements of ith column are multiplied with k and added to the
corresponding elements of jth column then it is denoted by CjCj+kCi
Elementary Matrix:
A matrix which is obtained by applying elementary transformations in known as Elementary
matrix.
Rank of a Matrix:
A matrix is said to be of rank r if
(i) It has atleast one non-zero minor of order r and
(ii) Every minor of order higher than r vanishes.
And it is denoted by (A).
Properties:
1) The rank of a null matrix is zero.
2) For a non-zero matrix A,(A) 1
3) The rank of every non-singular matrix of order n is n. The rank of a singular matrix of order n is
< n.
4) The rank of a unit matrix of order n is n.
5) The rank of an mn matrix min(m,n).
6) The rank of a matrix every element of which is unity is unity.
Different methods to find the rank of a matrix:

Method 1:
Echelon form:: A matrix is said to be Echelon form if
1) Zero rows, if any, are below any non
non-zero row
2) The first non-zero entry in each
ach non
non-zero row is equal to one
3) The number of zeros before the first non
non-zero
zero elements in a row is less than the number of such
zeros in the next rows.
0
0
Ex: the rank of matrix which is in Echelon form
0

0
rows is

1
0
0
0

3
1
0
0

4 5
2 9
is 3 since the no. of non-zero
non
1 3

0 0

Note: Apply only row operations.


Method 2:
Normal Form: Every mn matrix of rank r can be reduced to the form

or I r
0 0

by a finite chain of elementary row or column operations.


2
4
Ex: the rank of matrix A=
1

1
1 0 0 0
0 1 0 0 I 3

form as
0 0 1 0 0

0 0 0 0

2
2
1
2

0
0
0
1

6
2
after applying elementary operations reduced to normal
3

0
0

the rank of A is 3.
Finding the Inverse of a Nonsingular Matrix using Row/Column Transformations(GaussTransformations(Gauss
Jordan Method):

EXAMPLE: Find the inverse of the matrix

using the Gauss-Jordan


Jordan method.

Solution: Consider the matrix


Jordan method are:

1.

2.

3.

4.

5.

6.

7.

8. Thus, the inverse of the given matrix is

A sequence of steps in the GaussGauss

EXERCISE : Find the inverse of the following matrices using the Gauss
Gauss-Jordan
Jordan method.

Consistency of System of Linear equations (Homogeneous and NonHomogeneous) Using Rank


of the Matrix:
Nature of solution:
1. m n non-homogeneous
homogeneous with m equations and n unknowns
The system of equations AX=B is said to be
i)
ii)

consistent if rank of A = rank of [AB]


consistent and unique solution if rank of A = rank of [AB]=r=n
Where r is the rank of A and n is the no. of unknowns.

iii)

Consistent
tent and an infinite no. of solutions if rank of A < rank of [AB] i.e., r < n. In
this case we have to give arbitrary values to nn-rr variables and the remaining variables
can be expressed in terms of these arbitrary values.
Inconsistent if rank of A rank of [AB]

iv)

2. m = n non-homogeneous
homogeneous equations with n equations and n unknowns
If A be an n-rowed non-singular
singular matrix, the ranks of matrices A and [AB] are both n.
Therefore the system of equations AX=B is consistent i.e., possesses a solution.
3. Method of finding the rank of A and [AB]:

Reduce the augmented matrix [A:B] to Echelon form by elementary row transformations.

Ex: Discuss for what values of and the simultaneous equations


x+y+z = 6 , x+2y+3z = 10 , x+2y+
x+2y+z = have
(i)
(ii)
(iii)

no solution
A uniquee solution
An infinite no. of solutions

Sol: The given equations can be written as AX=B

1 1 1
i.e., 1 2 3
1 2

x 6
y = 10

z

1 1 1 . 6
and we have the augmented matrix [AB] = 1 2 3 . 10
1 2 .
1
.
6
1 1
2
.
4 R2R2-R1, R3R3-R1
0 1
0 1 1 . 6

1
.
6
1 1

2
.
4 R3R3-R2
0 1
0 0 3 . 10
Case-I: When 3, then rank of A = 3 = rank of [AB]. So that the system of equations is
consistent. And r=3=n, so the system has unique solution.
Case-II: When =3 and 10, then rank of A = 2. And rank of [AB] = 3.
Therefore rank of A rank of [AB]. So the system is inconsistent.
Case-III: When =3 and =10, then rank of A = rank of [AB] = 2.
Therefore the system is consistent and has an infinite no. of solutions.
Since the no. of unknowns = n = 3 > rank of A = 2.
Homogeneous linear equations
Consider the system of m homogeneous equations in n unknowns

a x a x .......... a x 0
a x a x .......... a x 0
11

12

21

22

1n

2n

(1)

............................................................

a x a x
m1

n2

.......... a mn x n 0

(1) can be written as AX=O

a11 a12
a 21 a 22

Where A is the coefficient matrix formed by A =

am1 am 2

a1n
a 2 n

amn

x1
0
x2
0


.
.
X= and B=
.
.
xn
0
Consistency: The matrix A and [AB] are same. So rank of A = rank of [AB]
Therefore the system (1) is always consistent.
Nature of solution:
Trivial solution: Obviously x1=x2=x3= -------- =xn=0 is always a solution of the given system
and this solution is called trivial solution.
Therefore trivial solution or zero solution always exists.
Non-Trivial solution: Let r be the rank of the matrix A and n be the no. of unknowns.
Case-I: If r=n, the equations AX=O will have n-n i.e., no linearly independent solutions. In
this case, the zero solution will be the only solution.
Case-II: If r<n, we shall have n-r linearly independent solutions. Any linear combination of
these n-r solutions will also be a solution of AX=O.
Case-III: If m<n then rm<n. Thus in this case n-r > 0.

Therefore when the no. of equations < No. of unknowns, the equations will have an infinite
no. of solutions.
Ex: Show that the only real number for which the system
X+2y+3z = x, 3x+y+2z = y, 2x+3y+z = z has non-zero solution is 6 and solve
them when =6.

Sol: Given system of equations can be expressed as AX=O

2
3
1
x
0

1
2 ; X= y and O= 0
Where A= 3
2
z
0
3
1
Here the no. of variables = n = 3.
The given system of equations possesses a non-zero (non-zero) solution, if rank of A <
number of unknowns i.e., rank of A < 3.
For this we must have detA = 0

1
2 =0
3
1

3
2

6 6 6
i.e.,

1
3

3
2
1

2 =0
1

R1R1+R2+R3

2 =0
i.e., (6-) 3 1
2
3
1
1
i.e., (6-)

3 2
2
1

0
1 =0
1

C2C2-C1, C3C3-C1

i.e., (6-) (2+3+3) = 0


i.e., = 6 is the only real value and other values are complex.

When = 6, the given system becomes

3
5 2
3 5 2

2
3 5

x 0
y = 0

z 0

3
5 2

0 19 19
0
19 19

x 0
y = 0

z 0

R25R2+3R1, R35R3+2R1

2
3 x 0
5

0 19 19 y = 0
0
0
0 z 0

R3R3+R2

-5x+2y+3z = 0 and -19y+19z = 0


y=z
Since rank of A < No. of unknowns i.e., r < n (2 < 3)
Therefore, the given system has infinite no. of non-trivial solutions.
Let z=k y=k and -5x+2k+3k = 0 x=k

x=k, y=k and z=k is the solutions


Solving mxn and nxm Linear System of Equations by Gauss Elimination:
GAUSSIAN ELIMINATION METHOD: Consider the system of linear
a1 x1 b1 x2 c1 x3 d 1
equation s

a x b x c x d
a x b x c x d
2

The augmented matrix is

[A:B]=

a1 b1 c1 d1
a 2 b 2 c 2 d 2

a3 b3 c3 d 3

After performing row operations or column operations we get

a1 b1 c1 d1
[A:B] = 0 b2 c 2 d 2
0 0 c3 d 3
Ex: solve the equations
2x1 + x2 + x3 = 10
3x1 + 2x2 +3x3 = 18
X1 + 4x2 + 9x3 = 16 using gauss elimination method

2 1 1 10
Sol: The augmented matrix of the given matrix is [A B] = 3 2 3 18
1 4 9 16
1
10
2 1

[A B] 0 1 / 2 3 / 2 3 R22R3-R1
0 7 / 2 17 / 4 11
1
10
2 1

3 R3R3-7R2
[A B] 0 1 / 2 3 / 2
0 0
2 10
This augmented matrix corresponds to the following upper triangular system. By backward
substitution we get
2x1 + x2 + x3 = 10;

1
3
x2 + x3 = 3; -2x3 = -10
2
2

x3 = 5 ; x2 = -9 ; x1 = 7

CAYLEY-HAMILTON THEOREM:

Every square matrix satisfies its own characteristics equation


Verification: Example 1: Cayley-Hamilton theorem
Consider the matrix
11
A=

21

Its characteristic polynomial is


1- 1
2 1-

p( ) = det (A - I) =

= ( 1 - )2 - 2 = 2 - 2 - 1

p(A) = A2 - 2 A - I
=

32
43

11

11

21

21

22
42

10

01

-2

00
00

11
21

10
01

, which holds for every matrix.

Every matrix satisfies its own characteristic equation.

Finding Inverse of a Matrix and powers of a Matrix by CAYLEY-HAMILTON THEOREM


Linear Dependence and Independence of Vectors:
LINEAR TRANSFORMATION:
Consider a set of n linear equations
a11 x1 a12 x2 .......... a1n xn y

a x a x
21

21

............ a 2 n xn

.....................................................

a x a x
n1

n2

................ ann xn

It can represented as Y=AX


If the above equation is additive and homogenous
The transformation is called linear transformation
Orthogonal Transformation:
Suppose that A is a real square matrix of order n. Then the linear transformation Y=AX is
called an orthogonal linear transformation or orthogonal transformation if A is an orthogonal
matrix.

EIGEN VALUES AND EIGEN VECTORS of a Matrix:


Consider the following n homogeneous equations in n unknowns as given below
(a11 ) x1 + a12x2 + - - - - - - a1nxn = 0
a21x1 + (a22 ) x2 + - - - - - -a2nxn = 0
. . . . . . . . . . . . . . . . . ..
an1x1 + an2x2 + - - - - - - (ann )xn = 0
The above system of equations in matrix notation can be written as (A I) X = 0
Where is a parameter.
The matrix (A I) is called Characteristic Matrix and |A I| = 0 is called Characteristic
Equation. i.e.,
|A I| = (-1)n n + k1 n-1 + k2 n-2 + - - - - - - + kn = 0
Where k1,k2,- - - - - ,kn are expressible in terms of the elements aij
Eigen Value: The roots of characteristic equation are called the characteristic roots or latent
roots or eigen values.
Eigen Vector: If is a characteristic root of a matrix then a non-zero vector X such that AX =
X is called a characteristic vector or Eigen vector of A corresponding to the characteristic root
.
Note: Eigen vector must be a non-zero vector
PROPERTIES OF THE EIGEN VALUES:
1) The sum of the Eigen values of the square matrix is equal to its trace and product of the Eigen
values is equal to its determinant.

Proof: consider A =

a11 a12 a13


a 21 a 22 a 23

a31 a 32 a 33

The characteristic equation is | A I | = 0

a11

a12

a13

a 21

a 22

a 23

a31

a32

a 33

=0

- 3 + 2 (a11+a22+a33) + - - - - - - - = 0
| A I | = - 3 + 2 (a11+a22+a33) + - - - - - - - (1)
Also if 1, 2, 3 be the Eigen values, then
|A I| = (-1)3(-1) (-2)( -3)
| A I | = -3 + 2 (1+2+3) + - - - - - - -

(2)

From (1) and (2), we can write


1+2+3 = a11+a22+a33
Hence the sum of the Eigen values of a matrix is the trace of the matrix.
2) If is an eigen values of A corresponding to the eigen vector X then n is the eigen value of the
matrix An corresponding to the eigen vector Xn
3) If 1 2 3 ......... n are the latent roots of A then A3 has the latent roots as
4)A square matrix A and its transpose AT have the same eigen values

......
3

5) If A and B are n rowed square matrix and if A is invertible then A-1B and BA-1 have the same
eigen values
6) If 1 2 3 ......... n are the eigen values of a matrix A then product of each 1 2 3 ......... n
with K are the eigen values of the matrix KA where K is anon-zero scalar
7) If is an Eigen value of the matrix A then K+ is an Eigen value of the matrix A+KI
8) If is the Eigen value of A then -K are the eigen values of the matrix A-KI.
9) If is the Eigen value of the matrix of A, then (-K) 2 is the eigern value of the matrix
2

(A-KI)

10) If is an Eigen value of a non-singular matrix A then A/ is an Eigen value of the matrix Adj
(A)
Eigen values and Eigen vectors of complex matrix and their properties:
1) The Eigen value of a hermitian matrix is real
2) The eigen values of askew-hermitioan are purely imaginary or zero
3) The eigenvalues of unitary matrix have absolute value one
4) Transpose of a unitary matrix is unitary

5) Inverse of unitary matrix is unitary


6) If A is hermitian (skew-hermitian) then iA is skew-hermitian (hermitian)
7) The square matrix A can be written as sum of Hermitian and Skew-hermitian

Method of finding Eigen values and Eigen vectors:


Eigen values: Consider the matrix A of order n and its characteristic equation is |A-I| = 0. Expand
the determinant to obtain Eigen values =1, 2, 3, ----, n.
Eigen vectors:
Case-I: Eigen values are distinct 123

Method1: Corresponding to the eigen value 1, the eigen vector X =

x1
x2
can be obtained from the
x3

matrix equation (A-I)X=O and by expanding it we get three homogeneous linearly independent

x1
x2
equations are obtained and solving any two equations for x1,x2,x3 the eigen vector X1= can be
x3
obtained. Similarly, the remaining Eigen vectors X2, X3 can be obtained corresponding to the Eigen
values 2 and 3.
Method2: (Rank method) in the matrix equation (A-I) X=O, reduce the coefficient matrix to
Echelon form, the rank of the coefficient matrix is less than the number of unknowns. So give
arbitrary constants to n-r variables and solve as in case of homogeneous equations.
Case-II: Finding Linearly Independent Eigen vectors of a matrix whenthe Eigen values of the matrix
are repeated (1=2)
Method1: The matrix equation (A- I) X=O gives three equations which represent a single
independent equation of the form.
a1x1+a2x2+a3x3 = 0
Put x3=0, a1x1+a2x2 = 0 give values to x1 and x2 to satisfy the equation.

Let x1=k1, x2=k2 and x3=0 then X1 =

k1
k 2

0

Put x2=0, a1x1+a3x3 = 0 give values to x1 and x3 to satisfy the equation.

Let x1=c1 and x3=c3 be the values, then X2 =

c1
0

c3

So we can get two linearly independent Eigen vectors X1 and X2


Method2: (Rank method)
Ex: Find the Eigen values and corresponding Eigen vectors of the matrix

6 2 2
A = 2 3 1
2 1 3
Sol: The characteristic equation of the matrix A is |A- I| = 0

2
2

3 1 = 0
1 3

3 102 + 20 32 = 0

(-2)(-2)(-8) = 0

= 2, 2, 8

The Eigen values of A are 2, 2 and 8.

Let the Eigen vector X =

x1
x2

x3

of A corresponding to the Eigen value is given by the non-zero

solution of the equation (A-I) X = O

2 x1
6 2
2 3 1

x2 =
2
1 3 x3

0
0

If =8, then the Eigen vector X1 is given by (A-8I) X = O

2 x1
6 8 2
2 3 8 1

x2 =
2
1 3 8 x3

2 2 2 x1
2 5 1

x2 =
2 1 5 x3

0
0

0
0

-2x1-2x2+2x3 = 0
-2x1-5x2-x3 = 0
2x1-x2-5x3 = 0

Solving any two of the equations, we get

x1 x 2 x 3
=
=
=k (say)
2 1 1

x1=2k, x2= -k, x3=k (k is arbitrary)


DIAGONALIZATION OF THE MATRIX:
If a square matrix A of order n has linearly independent Eigen vectors
Corresponding to the n Eigen values
such that
Let

AP=D (DIAGONAL MATRIX)

X X X
1

.........

........ X n be the Eigen vector

X X X
1

........ X n

respectively then a matrix P can be found

.........
1

Define P=(

X X X
1

be eigen values

........ X n )

then

AP=PD

p
p

AP=

PD

AP=D

MODAL AND SPECTRAL MATRIX:


The matrix P in the above result which diagonals the square matrix A is called the modal matrix of A
and the resulting diagonal matrix D is know as spectral matrix.
Calculation of powers of Matrix:
Let A be the square matrix and
Let P a non-singular matrix such that D=P-1AP
D2=P-1A2P
Similarly D3=P-1A3P Like for n tern we have Dn=P-1AnP

Quadratic form(Upto Three Variables):


A homogeneous polynomial of second degree in n variables (x1,x2,-------,xn) is called a quadratic
form. The most general quadratic form is
a11 x1 a12 x2 .......... a1n xn
Q=

a x a x
21

21

............ a 2 n x n

.....................................................

a x a x
n1

n2

................ a nn xn

Where aij are elements of a field F and we can write Q=XTAX =

a x x
i 1 j 1

The symmetric matrix is called the matrix of quadratic form


Canonical form:

ij

let Q=XTAX be a quadratic form in n variables then there exists a real non-singular linear
transformation X=PY which transforms Q=XTAX to the form

y y

.......

2
n

this form is

called canonical form or normal form


Rank of the quadratic form : let XTAX be a quadratic form over R . The rank r of A is called
the rank of the quadratic form if (r<n) where n is order of A or det(A)=0 or A is singular . The
quadratic form is called singular or non-singular
Index: The index of the quadratic form is the number of positive square terms in the canonical
form
Signature: signature of the quadratic form is the difference between the number of positive and
negative square terms of be the canonical form
Definiteness: A real non-singular quadratic form Q= XTAX (with det(A)0) is said to be
positive definite if rank and index are equal that is r=n and s=n. and the quadratic form is called
negative definite if index equal to zero that is r=n and s=0 and the quadratic form is called
positive semi-definite if the rank and index are equal but less than n that is s=r<n and it is
negative semi- definite if index is zero

Reduction of quadratic form to canonical form:


Method1: Orthogonalization
Procedure:
Step1: Write the matrix A of the given quadratic form.
Step2: Find the Eigen values 1, 2, 3and corresponding Eigen vectors X1 ,X2 and
X3 in the normalized form.
Step3: Write the modal matrix P = [X1 X2 X3] formed by normalized vectors.
Step4: P being an orthogonal matrix P-1=P so that PTAP = D where D is the diagonal matrix
formed by Eigen values.
Step5: The canonical form is YT (PTAP) Y = 1y12+2y22+3y32
Step6: The orthogonal transformation is X=PY

Note: The matrix A of the quadratic form is symmetric matrix and so diagonalization is by
orthogonal transformation.
Nature of a quadratic form:
When the quadratic form Q=XTAX is reduced to a canonical form, it will contain r terms if the
rank of A is r.
1. Rank of A = Rank of quadratic form = r = No. of terms in the canonical form.
2. The Index = No. of positive terms in the canonical form and is denoted by p
3. Signature = the difference of the no. of positive terms and the no. of negative terms i.e., 2p-r.
4. The quadratic form is
(i) Positive definite, if all the Eigen values of A are positive i.e., r=n and p=n.
(ii) Positive semi definite, if at least one of the Eigen values of A is zero and
other
positive i.e., r<n and p=0.
(iii) Negative definite, if all the Eigen values are negative i.e., r=n and p=0
(iv) Negative semi definite, if at least one of the Eigen values of A is zero and others are
negative i.e., r<n and p=0.
(v) Indefinite, if the Eigen values of A are positive and negative.
Ex: Reduce the quadratic form 6x12+3x22+3x32-4x1x2+4x1x3-2x2x3 to canonical form by an
orthogonalization. Find rank, Index and signature of quadratic form.
Sol: The matrix A of the given quadratic form is
6 2 2
A = 2 3 1
2 1 3
The characteristic equation of the matrix A is |A- I| = 0
6 2
2

2
2

3 1 = 0
1 3

3 102 + 20 32 = 0
(-2)(-2)( -8) = 0
= 2, 2, 8
The Eigen values of A are 2,2 and 8.

Let the Eigen vector X =

x1
x2
of A corresponding to the Eigen value is given by the non-zero
x3

solution of the equation (A-I)X = O

2 x1
6 2
2 3 1

x2 =
2
1 3 x3

0
0

If =8, then the Eigen vector X1 is given by (A-8I)X = O

2 x1
6 8 2
2 3 8 1

x2 =
2
1 3 8 x3
2 2 2 x1
2 5 1

x2 =
2 1 5 x3

0
0

0
0

-2x1-2x2+2x3 = 0
-2x1-5x2-x3 = 0
2x1-x2-5x3 = 0
Solving any two of the equations, we get

x1 x 2 x 3
=
=
=k (say)
2 1 1
x1=2k , x2=-k and x3=k

2k
For the eigen value =8, the eigen vector X3 = k
k
2
In particular k=1, X3 = 1
1
If =2, the Eigen vector X is given by (A-2I) X=O

4 2 2 x1
2 1 1

x2 =
2 1 1 x3

0
0

4x1-2x2+2x3 = 0
-2x1+x2-x3 = 0
2x1-x2+x3 = 0
All the above three equations represent only one independent equation
2x1-x2+x3 = 0
Let x1 = 0, then -x2+x3 = 0 x2 = x3 =1
Let x2 = 0, then 2x1+ x3 = 0 x1 = 1, x3 = -2
So corresponding Eigen value =2, then linearly independent Eigen vectors are given by

0
1


X1 = 1 and X2 = 0
1
2

The normalized Eigen vector of

0
1/ 2

X1 =
1/ 2

The normalized Eigen vector of X2 =

The normalized Eigen vector of

1/ 5

2/ 5

2/ 6

1
/
6

X3 =
1/ 6

= e1

= e2

= e3

The modal matrix in normalized form P = [e1 e2 e3]

PTAP =

2 0 0
0 2 0

0 0 8

0
1/ 5
2/ 6

1
/
2
0

1
/
6

1/ 2 2 / 5 1/ 6

= diag [2 2 8]

And the quadratic form will be reduced to the normal form 2y12+2y22+8 y32 by the orthogonal
transformation X = PY. i.e.,
y2 2 y3
y1 y3
y1 2 y 2 y3
x1 =
+
; x2 =
; x3 =
+
5
6
6
6
2
5
6
Method2: Diagonalization
Procedure:
Step1: Write the matrix A of the given quadratic form, A is a symmetric matrix
Step2: Write A=IAI
Step3: Reduce the matrix A to diagonal form by applying row and column transformations.

Step4: Apply same row operations on prefactor I on R.H.S and identical column operations on
postfactor I on R.H.S
Step5: A=IAI will be transformed to the form D = PTAP
The canonical form is given by

YT D Y = [y1 y2 y3]

1 0 0
0 2 0

0 0 3

YT D Y = 1y12+2y22+3y32

y1
y 2

y3

Ex: Reduce the quadratic form 6x12+3x22+3x32-4x1x2+4x1x3-2x2x3 to canonical form by


diagonalization. Find rank, Index and signature of quadratic form.
Sol: The matrix A of the given quadratic form is
6 2 2
A = 2 3 1
2 1 3
The characteristic equation of the matrix A is |A- I| = 0
6 2
2

2
2

3 1 = 0
1 3

3 102 + 20 32 = 0
(-2)( -2)( -8) = 0
= 2,2,8
The Eigen values of A are 2,2 and 8.

Let the Eigen vector X =

x1
x2
of A corresponding to the Eigen value is given by the non-zero
x3

solution of the equation (A-I)X = O

2 x1
6 2
2 3 1

x2 =
2
1 3 x3

0
0

If =8, then the Eigen vector X1 is given by (A-8I)X = O

2 x1
6 8 2
2 3 8 1

x2 =
2
1 3 8 x3

0
0

2 2 2 x1
2 5 1

x2 =
2 1 5 x3

0
0

-2x1-2x2+2x3 = 0
-2x1-5x2-x3 = 0
2x1-x2-5x3 = 0
Solving any two of the equations ,we get

x1 x 2 x 3
=
=
=k (say)
2 1 1
x1=2k , x2=-k and x3=k

2k
For the eigen value =8, the eigen vector X3 = k
k
2
Inparticular k=1 , X3 = 1
1
If =2, the Eigen vector X is given by (A-2I)X=O

4 2 2 x1
2 1 1

x2 =
2 1 1 x3

0
0

4x1-2x2+2x3 = 0
-2x1+x2-x3 = 0
2x1-x2+x3 = 0
All the above three equations represent only one independent equation
2x1-x2+x3 = 0
Let x1 = 0, then -x2+x3 = 0 x2 = x3 =1
Let x2 = 0, then 2x1+ x3 = 0 x1 = 1, x3 = -2
So corresponding Eigen value =2, then linearly independent Eigen vectors are given by

0
1


X1 = 1 and X2 = 0
1
2

The normalized Eigen vector of

0
1/ 2

X1 =
1/ 2

= e1

The normalized Eigen vector of X2 =

The normalized Eigen vector of

1/ 5

2/ 5

2/ 6

1
/
6

X3 =
1/ 6

= e2

= e3

The modal matrix in normalized form P = [e1 e2 e3]

=
The matrix P is orthogonal so P-1 = PT
Diagonalization PTAP = D

0
1/ 5
2/ 6

1
/
2
0

1
/
6

1/ 2 2 / 5 1/ 6

0
1/ 2
1 / 2 6 2 2 0
1/ 5
2/ 6

2
3

1
1
/
5
0

2
/
5
1
/
2
0

1
/
6

2 / 6 1/ 6 1/ 6 2 1 3 1/ 2 2 / 5 1/ 6

2 0 0
0 2 0

0 0 8

Canonical form YT (PTAP) Y = YT D Y

y1
2 0 0
0 2 0 y2
= [y1 y2 y3]

0 0 8 y3

= 2y12+2y22+8 y32
The rank of quadratic form = no. of non-zero terms in the canonical form = 3
The Index P = no. of positive terms in the canonical form = 3
Signature = 2P rank of quadratic form = 3.

Method3: Lagranges Reduction


Procedure:
Step1: Take the common terms from product terms of given quadratic form.
Step2: Make perfect squares suitably by regrouping the terms.
Step3: The resulting relation gives the canonical form.
Ex: Reduce the following quadratic form x12+2x22-7x32-4x1x2+8x1x3 to canonical form by
lagranges reduction.
Sol: Let the quadratic form be x12+2x22-7x32-4x1x2+8x1x3
= x12+2x22-7x32-4 x1(x2-2x3)
= [x12-4x1(x2-2x3)+22(x2-2x3)2]-22(x2-2x3)2+2x22-7x32
= [x1-2(x2-2x3)]2-4[x22-4x2x3+4x32]+2x22-7x32
= [x1-2(x2-2x3)]2-4x22+16x2x3-16x32+2x22-7x32
= [x1-2(x2-2x3)]2-2x22+16x2x3-23x32
= [x1-2(x2-2x3)]2-2[x22-8x2x3]-23x32
= [x1-2(x2-2x3)]2-2[x22-8x2x3+42x32]+32x32-23x32
= [x1-2(x2-2x3)]2-2[x2-4x3]2+9x32
= y12-2y22+9y32
Where y1= x1-2(x2-2x3) ; y2= x2-4x3 ; y3=x3
Let the quadratic form is XTAX, using orthogonal transformation X=PY we get

PY APY
T

T AP Y
p

where p AP D

y y

.......

DY

2
n

Where P is modal matrix


D is called spectral matrix
Practice Problems:
1. Reduce the quadratic form x12+3x22+3x32-2x2x3 to canonical form by orthogonal
transformation.
2. Reduce the quadratic form 6x2+3y2+3z2-4yz-4zx-2xy to normal form by diagonolization.
3. Find the nature of the quadratic form x2+3y2+4z2+2yz+4xz+2xy
4. Reduce the quadratic form 2x2+7y2+5z2-8xy-10yz+4zx to canonical form by lagranges
reduction.
5. (i) Show that if A is a Hermitian then iA is a Skew-Hermitian
(ii) Show that if A is Skew-Hermitian the iA is Hermitian

MATRICES 1
1.

Find the rank of following matrices by way of definition as well as by reducing it to


the normal form.
Also Verify Cayley Hamilton Theorem and find inverse for the followings.

1.

2 1 1
A 1 2 1
1 1 2

3 1 1
1
Ans.: A1 1 3 1
4
1 1 3

2.

7 2 2
A 6 1 2
6 2 1

3 2 2
1
Ans.: A 6 5 2
3
6 2 5

3.

3 1 1
A 1 5 1
1 1 3

7 2 3
1
1 4 1
Ans.: A

20
2 2 8

2. Using Cayley Hamilton Theorem, find the inverse of the following Matrices:

5 3
3 2

1.

2 3
3 5

Ans.:

2.

7 1 3
6 1 4

2 4 8

8 20 7
1
40 50 10
Ans.:

50
22 30 13

3.

1 0 3
2 1 1

1 1 1

4.

1 1 3
1 3 3

2 4 4

5.

2 1 3
1 0 2

4 2 1

6.

5 1 5
0 2
0

5 3 15

0 3 3
1
3 2 7
Ans.:

9
3 1 1

Ans.:

12 4 6
1
5 1 3

4
1 1 1

4 5 2
1
Ans.: 7 10 1
5
2 0
1

AError! Not a valid embedded object.ns.

3 0 1
1
0 5 0
10
1 1 1

7.

1 1 3
1 3 3

2 4 4

24 8 12
1
10 2 6
Ans.:
8
2 2 2

8.

1 2 4
1 0 3

3 1 2

6
3 8
1
7 14 7
Ans.:

7
2
1 5

3. Find the Characteristic equation of the following Matrix and Show that the equation is
satisfied by the matrix itself and hence obtain the inverse of the given matrix:

1 3 7
4 2 3

A=
1 2 1

4 11 5
1
Ans.: 4 20 35 0 , A 1 6 25
35
6 1 10
3

2 1 1
4. Find the Characteristic equation of the matrix A 0 1 0 and hence, find the
1 1 2
matrix represented by A8 5 A7 7 A6 3 A5 A4 5 A3 8 A2 2 A I
8 5 5
Ans.: 0 3 0
5 5 8
1 2 3
5. Find the Characteristic equation of the matrix A 2 1 4 . Show that the
3 1 1
equation is satisfied by A.
Ans.: 3 2 18 40 0
1

3
3 3 and hence find its
2 4 4

6. Find the Characteristic equation of the matrix A 1

inverse.

5
1
Ans.: A
4
1

1
1
4
1
4

3
2

3
4
1
4

1 2 2
7. Find the Characteristic equation of the matrix A 1 1 1 and hence find its
1 3 1
4 4 4
1
3
2
1
inverse.
Ans.: 4 4 0, A 2 1 3
4
2 1 1

2 1 1
8. Find the Characteristic equation of the symmetric matrix A 1 2 1 and
1 1 2
1
verify that it is satisfied by A and hence obtain A .
3 1 1
1
Ans.: 6 9 4 0, A 1 3 1
4
1 1 3
3

4 3 1
9. Find the Characteristic equation of the matrix A 2 1 2 .Hence find A1 .
1 2 1
7 2 10
1
1
Ans.:. A 2 2 1
9
1 1 4
1 2 2
10. Show that the matrix A 1 2 3 satisfies its Characteristic equation. Hence
0 1 2
7 2 10
1
1
1
find A .
Ans.: A 2 2 1
9
1 1 4

1 2
11. Using Cayley Hamilton Theorem find A8 , if A
.
2 1
Ans.: 625 I

MATRICES2

1)

5 4
Find the eigenvalues and eigenvectors of the matrix

1 2

Ans:

2)

3)

1, 6;

1
K1 ,
1

4
K2
1

Find the eigenvalues and eigenvectors of the matrices

1)

8 6 2
6 7 4

2 4 3

1
2
2

Ans.: 0, 3, 15 K1 2 , K 2 1 , K 3 2
2
2
1

2)

6 2 2
2 3 1

2 1 3

Ans.: 2, 2, 8

1
1
2

K 1 0 , K 2 2 , K 3 1
2
0
1

3)

2 0 1
0 2 0

1 0 2

Ans.: 1, 2, 3

1
0
1
K1 0 , K 2 1 , K 3 0
1
0
1

4)

2 1 1
1 2 1

0 0 1

Ans.: 1, 1, 3

1
1
1

K 1 2 , K 2 1 , K 3 1
1
0
0

5)

3 1 4
0 2 0

0 0 5

Ans.: 2, 3, 5

1
1
2

K1 1 , K 2 0 , K 3 0
0
0
1

3 1 1
Find the Characteristic Polynomial of the Matrix A 1 5 1
1 1 3

Verify Cayley Hamilton Theorem for this matrix. Hence find A 1


.

7 2 3
1
1
4
1
Ans.: A

20
2 2
8
1

4)

5)

Using Cayley Hamilton Theorem, find A 1 , given that

1)

2 1 3
A 1 0 2
4 2 1

5 2
4
1
Ans.: A 7 10 1
5
2 0
1

2)

5 1 5
A 0
2
0
5 3 15

3 0 1
1
Ans.: A 0 5 0
10
1 1 1

1 3 7
Find the characteristic Equation of the matrix A 4 2 3 and Show that
1 2 1
the equation is also satisfied by A .
Ans.: 3 4 2 20 35 0

6)

1 2 2
Show that the matrix A 1 2 3 satisfies its characteristic equation.
0 1 2
Hence, find A 1

7 2 10
1
Ans.: A 2 2 1
9
1 1
4
1