Matrices Eigenvalues and Inverse

www.Vidyarthiplus.
com
St. Joseph’s College of Engineering, Chennai 600 119.

MA8251 Engineering Mathematics II
Unit I Matrices
Problem 1. Find the eigen values and eigen vectors of the matrix
  2 2  3
A   2 1  6
  1  2 0 
Solution:
The characteristic equation is | A - I | = 0.
2- 2 3
i.e., 2 1-  6  0
1 2 0-
i.e., (-2 - ) [-(1 - ) -12] - 2[-2 - 6] -3[-4 + 1 - ] = 0
i.e., (-2 - ) [2 -  -12] + 4 + 12 + 9 + 3 = 0
i.e., 3 + 2 - 21 - 45 = 0 (1)
3 2
Now, (-3) + (-3) - 21(-3) - 45 = -27 + 9 + 63 – 45 = 0
 -3 is a root of equation (1).
Dividing 3 + 2 - 21 - 45 by  + 3
 3 1 1  21  45
0 3 6 45
1  2  15 0
Remaining roots are given by
2 - 2 - 15 = 0
i.e., ( + 3) ( - 5) = 0
i.e.,  = -3, 5.
 The eigen values are -3, -3, 5
 2  λ 2 - 3  x1  0
The eigen vectors of A are given by  2 1 - λ - 6  x 2   0
 - 1 - 2 - λ   x3  0
Case 1  = -3
 2  3 2 - 3  1 2 - 3
Now  2 1  3 - 6 ~  2 4 - 6
 - 1 -2 3  - 1 - 2 3 
1 2 - 3
~ 0 0 0 
0 0 0 
www.Vidyarthiplus.com
Matrices www.Vidyarthiplus.com 2
 x1 + 2x2 - 3x3 = 0
Put x2 = k1, x3 = k2
Then x1 = 3k2 - 2k1
3k 2  2k 1 
 The general eigen vectors corresponding to  = -3 is  k1 

 k 2 
3
When k1 = 0, k2 = 1, we get the eigen vector 0 
1 
 2
When k1 = 1, k2 = 0, we get the eigen vector  1 
 0 
 3   2
Hence the two eigen vectors corresponding to  = -3 are 0 and  1  .
 
1  0 
These two eigen vectors corresponding to  = -3 are linearly independent.
Case 2  = 5
  2  5 2 - 3   7 2 - 3
 2 1 - 5 - 6 ~  2 - 4 - 6


 - 1 - 2 - 5  - 1 - 2 - 5
 1  2  5 
~  0  8  16
 0 0 0 
 -x1 - 2x2 - 5x3 = 0
-8x2 - 16x3 = 0
A solution is x3 = 1, x2 = -2, x1 = -1
  1
 Eigen vector corresponding to  = 5 is  2 .
 1 
 1 1 2 
Problem 2. Find the characteristic equation of  2 1 3  and verify Cayley-
 3 2 3
Hamilton Theorem. Hence find the inverse of the matrix.
 1 1 2 
Solution: Let A   2 1 3   Characteristic eqn. of A is
 3 2 3
   1  1  3    9  9  1  26  0
3 2
i.e  3   2  19  26  0
By Cayley-Hamilton theorem  A3  A2  19 A  26I  0 .
Verification:
 1 1 2  1 1 2   9 2 7 
    
 A  A. A   2 1 3  2 1 3    5
2
9 10 
 3 2 3  3 2 3   10 7 21 
    
 9 2 7  1 1 2   16 21 45 
    
 A  A .A   5
3 2
9 10  2 1 3    43 16 67 
 10 7 21  3 2 3   67 45 104 
   
Substituting in the characteristic equation
 16 21 45   9 2 7   19 19 38   26 0 0   0 0 0 
         
 43 16 67    5 9 10    38 19 57    0 26 0    0 0 0 
 67 45 104   10 7 21   57 38 57   0 0 26   0 0 0 
         
Hence verified.
Now to find the inverse of the matrix A, premultiply the characteristic equation by A1
 A2  A  19 I  26 A1  0
 A1  19 I  A  A2 
1
26
19 0 0   1 1 2   9 2 7    9 5 5 
1       1  
   0 19 0    2 1 3    5 9 10     3 9 7 
26  26 
 0 0 19   3 2 3   10 7 21    7 5 1

1 0 3 
Problem 3. Given A  2 1  1 , use Cayley-Hamilton Theorem to find the inverse of A
1  1 1 
and also find A4
Solution:
The characteristic equation of A is
1 λ 0 3
2 1 λ 1  0
1 1 1 λ
i.e., (1-) [(1 - ) (1 - ) -1] + 3[-2 - (1 - )] = 0
i.e., (1 - )3 - (1 - ) – 6 -3 + 3 = 0
i.e., 1 - 3 + 32 - 3 – 1 +  - 9 + 3 = 0
i.e., -3 + 32 +  - 9 = 0
i.e., 3 - 32 -  + 9 = 0
By Cayley-Hamilton theorem, A3-3A2 – A + 9I = 0
To find A , multiplying by A , A2-3A - I + 9A-1 = 0
-1 -1
1
 A-1 = [-A2 + 3A + I]
9
1 0 3  1 0 3   4  3 6 
A  2 1  1 2 1  1  3 2 4

2
1  1 1  1  1 1  0  2 5

  4 3  6  3 0 9  1 0 0
A    3  2  4  6 3  3  0 1 0
1 1  
9
 0 2  5 3  3 3  0 0 1
0 3 3
 3 2  7
1
9
3  1  1
To find A4:
We have A3- 3A2 – A + 9I = 0
i.e., A3 = 3A2 + A - 9I (1)
Multiplying (1) by A, we get,
A4 = 3A3 + A2-9A 
= 3(3A2 + A - 9I) + A2 - 9A using (1)
= 10A2 - 6A - 27I
 4  3 6  1 0 3  1 0 0
 10 3 2 4  62 1  1  27 0 1 0
   
0  2 5 1  1 1  0 0 1
 7  30 42
  18  13 46
 6  14 17
 0 0 2
Problem 4. . If A   2 1 0  express A6  25 A2  122 A as a single matrix
 1 1 3 
 
Solution: To avoid higher powers of A like A6 we use Cayley Hamilton Theorem.
Characteristic equation is  3  4 2  5  2  0
By Cayley Hamilton Theorem A3  4 A2  5 A  2I  0
To find A6  25 A2  122 A we will express this in terms of smaller powers of A using the
characteristics equation. We know that (Divisor) X (Quotient) + Remainder = Dividend
Assuming A3  4 A2  5 A  2I as the divisor we get,
A3  4 A2  11A  22I
A6  0 A5  0 A4  25 A2  122 A  0 I
A3  4 A2  5 A  2I
A6  4 A5  5 A4  2 A3
4 A5  5 A4  2 A3  25 A2  122 A
4 A5  16 A4  20 A3  8 A2
11A4  22 A3  33 A2  122 A
11A4  44 A3  55 A2  22 A
22 A3  88 A2  100 A
22 A3  88 A2  110 A  44 I
10 A  44I
 A6  25 A2  122 A   A3  4 A2  5 A  2I  A3  4 A2  11A  22I    10 A  44I 

But A3  4 A2  5 A  2I  0
A6  25 A2  122 A  0  10 A  44 I
  10 A  44 I 
 0 0 20   44 0 0  
   
   20 10 0    0 44 0  
 10 10 20   0 0 44  
 44 0 20 
 
   20 54 0
 10 10 74 

 44 0 20 
 
   20 54 0 
 10 10 74 

Problem 5. If i are the eigen values of the matrix A, then prove that
i k i are the eigen values of kA where ‘k’ is a nonzero scalar.
ii. im are the eigen value of Am and
1
iii. are the eigen values of A1 .
i
Solution: Let i be the eigen values of matrix A and Xi be the corresponding eigen
vectors. Then by defn: AXi  iXi......( I ) ( i.e by defn. of eigen vectors)
i. Premultiply ( I ) with the scalar k. Then
k  AXi   k  iXi 
i.e.  kA X i   k i  Xi
 k i are the eigen values of kA (comparing with ( I ) i.e by defn.)
ii. Premultiply ( I ) with A, then

A  AXi   A  iXi 
i.e. A2 X i  i  AXi 
 i  i Xi  from (I)
  i  Xi
2
1y we can prove that A3 Xi   i  Xi and so on Am Xi   i  Xi

3 m
i m are the eigen values of the Am (comparing with ( I ) i.e. by defn.)
iii. Premultiply ( I ) with A1 , then

A1  AXi   A1  iXi 
i.e.  A1 A  Xi  i  A1 Xi 
i.e. IXi  i  A1 Xi 
1
i.e. A1 Xi  Xi
i
1
are the eigen values of A1 (comparing with ( I ) ).

i
2 0 1
Problem 6. Find the characteristic vectors of  0 2 0  and verify that they are
1 0 2
 
mutually orthogonal.
2 0 1
 
Solution: A =  0 2 0  Characteristic equation is  3  6 2  11  6  0
1 0 2
 
Solving:   1, 2,3
Consider the matrix equation  A   I  X  0
Case (i) when   1;
 1 0 1  x1   0  1x1  0 x2  1x3  0  1
    
 0 1 0  x2    0  i.e. 0 x1  1x2  0 x3  0   2  equation (1) & (3) are identical.
 1 0 1  x   0  1x1  0 x2  1x3  0   3
  3   
Solving (1) and (2) using the rule of cross multiplication
 1
x1 x2 x3 x1 x2 x3  
  i.e.    X 1   0 
0 1 0 1 0 1 1 0 1 1
 
Case (ii) when   2;
 1 0 1  x1   0  0 x1  0 x2  1x3  0 x3  0
    
 0 1 0  x2    0  i.e. 0 x1  0 x2  0 x3  0 i.e. x2 is arbitrary say k
 1 0 1  x   0  1x1  0 x2  0 x3  0 x1  0
  3   
 0 0
   
 X 2   k  i.e  1  .
 0 0
   
Case (ii) when   3;
 1 0 1  x1   0   x1  0 x2  1x3  0
    
 0 1 0  x2    0  i.e. 0 x1  1x2  0 x3  0 Solving (1) and (2)
 1 0 1 x   0  1x1  0 x2  1x3  0
  3   
1
x1 x2 x3  
   X3   0
1 0 1 1
 
Thus the eigen values are 1,2,3 and the correspondent eigen vectors are
 1  0  1
     
 0  ,  1  and  0  . To check orthogonallity, X1 X 2  0
T
 1  0 1
     
X 2T X 3  0
X 1T X 3  0
 X1, X 2 , X 3
are mutually orthogonal.
 6 6 5 
Problem 7. Find the latent vectors of 14 13 10 
 7 6 4 
 
Solution: Characteristic equation is    1  0    1, 1, 1
3
When   1 (repeated 3 times)  we have to find 3 corresponding latent vectors.

 7 6 5  x1   0  7 x1  6 x2  5 x3  0
    
14 12 10  x2    0  i.e. 14 x1  12 x2  10 x3  0 All three equation are identical
 7 6 5  x   0  7 x1  6 x2  5 x3  0
  3   
.i.e. we get only one equation, but we have to find three vectors that are linearly
independent.
0
x2 x3  
 Assume x1  0  6 x2  5 x3  0 i.e.  6 x2  5 x3 i.e.   X 1   5 
5 6 6
 
 5 
x1 x3  
Assume x2  0  7 x2  5 x3  0 i.e. 7 x1  5 x3i.e..   X 2   0 
5 7 7
 
6
x1 x2  
And assume x2  0  7 x2  6 x3  0 i.e. 7 x1  6 x2 0i.e..   X 3   7 
6 7 0
 
X1, X2 and X3 are linearly independent.
 1 1 1
Problem 8. Find the eigen vectors of the matrix A   0 2 1
 4 4 3
Solution:
1 -  1 1 
The characteristic equation of A is  0 2- 1   0
  4 4 3 -  
i.e., (1 - ) [(2 - ) (3 - ) - 4] -1[0 + 4] +1[0 + 4(2 - )] =0

i.e., (1 - )(2 - 5 + 6 - 4) – 4 + 8 - 4 =0
i.e., (1 - )(2 - 5 + 2) + 4 - 4 =0
i.e., (1 - )(2 - 5 + 2 + 4) =0
i.e., ( -1)(2 - 5 + 6) =0
i.e., ( -1)( - 2)( - 3) =0
 The eigen values of A are  = 1, 2, 3.
1 - λ 1 1   x 1  0 

The eigen vectors are given by  0 2 - λ 1   x 2   0
  4 4 3 - λ   x 3  0
Case 1 =1
 0 1 1    4 4 2
 0 1 1 ~  0 1 1
   
 4 4 2  0 0 0
-4x1 + 4x2 + 2x3 = 0
x2 + x 3 = 0
A solution is, x3 = 2, x2 = -2, x1= -1
  1
 Eigen vector X1 =   2
 2 
Case 2 =2
  1 1 1  1 1 1
 0 0 1 ~  0 0 1
   
 4 4 1  0 0 0
-x1 + x2 + x3 = 0
x3 = 0
A solution is, x3 = 0, x2 = 1, x1 = 1
1 
 Eigen vector X2 = 1 
0 
Case 3 =3
 2 1 1  2 1 1
 0  1 1 ~  0  1 1
   
 4 4 0  0 0 0
-2x1 + x2 + x3 = 0
-x2 + x3 = 0
A solution is, x3 = 1, x2 = 1, x1 = 1
1
 Eigen vector X3 = 1
1
 2 2 0
Problem 9. Diagonalise the matrix  2 5 0  using orthogonal transformation.
 0 0 3
 
Solution: Characteristic equation is   10  27 18  0
3 2
Solving we get the eigen value as   1,3,6

 2  0 1
     
When   1, X 1   1  ; When   3, X 2   0  ; When   6, X 3   2 
 0 1 0
     
 2   1 
 5  0  5
 1    2 
Normalizing each vector, we get 
5   
, 0 and 
5
  1  
 0   0 
   
 2 0 1   2 1 0
 5 5  5 5 
  
 Normalized Modal Matrix, N   1 2 . N  N 
T
1,
5
0 0 0
5  
   1 2 0 
 0 1 0 
 5 5 
 
Then by the orthogonal transformation,
 2 
 2 1 0  0 1
 5 5  2 2 0 5 5
  
N AN   0 0 1 2 5 0 0 0 2  . On simplifying, we get
 
 1   5
2 0   0 0 3   1
 5 5   1 0 
 5 
N AN  D  1 , 2 , 3 
1 0 0
  which is diagonal matrix with eigen values along the
 D 1,3, 6    0 3 0 
0 0 6
 
diagonal (in order).
 6 2 2 
Problem 10. Reduce  2 3 1 to a diagonal matrix by orthogonal reduction.
 2 1 3 
 
Solution: Characteristic equation is   12 2  36  32  0    8, 2, 2
3
When   8
 2 2 2  x1   0 
    
 2 5 1  x2    0 
 2 1 5  x   0 
  3   
i.e 2 x1  2 x2  2 x3  0
2 x1  5x2  1x3  0
2 x1  1x2  5x3  0
2
x1 x2 x3  
Solving any two equations    X 1   1
2 1 1 1
 
When   2 (repeated twice)
 4 2 2  x1   0 
    
 2 1 1 x2    0  i.e 2 x1  2 x2  2 x3  0 . All the equations are identical.
 2 1 1  x   0 
  3   
0
x2 x3  
To get one of the vectors, assume x1  0  x2  x3  0 i.e.   X 2   1 
1 1 1
 
a
 
X1 X 2  0 . Therefore X 1 and X 2 are orthogonal. Now assume X 3   b  to be mutually
T
c
 
orthogonal with X1 and X2.
a 
  
X 1T X 3  0 i.e.  2 1 1  b   0 i.e.2a  b  c  0 
c
   a b c
 i.e  
a  2 2 2
 
and X 2T X 3  0 i.e.  0 1 1  b   0 i.e.0a  b  c  0 
c 
  
1
 
 X3   1  .
 
 1
After normalizing these 3 mutually orthogonal vectors, we get the normalized Modal
 2 0 1 
 6 3
 
Matrix N   1 1 1 
 6 2 3
 1 1 1 

 6 2 3
Diagonalizing we get
2 1 1   2 1 1 
 6 6 6  6 2 2  
 6 6 3
    
D  N AN   0
T 1 1  2 3 1  1 1 1 
 2 2   
 6 2 3
 2 1 3  
 1
 1 1  
 1 1 1 
 3 3 3  6 3 3
on simplifying we get D  D  1 , 2 , 3 
8 0 0
 
0 2 0
0 0 2 

 D  8, 2, 2 
3 1 1 
Problem 11. Diagonalise the matrix A  1 3 -1
1 -1 3 
Solution:
3- 1 1 
The characteristic equation of A is  1 3- -1   0

 1 -1 3- 
i.e., (-1)(2 - 8 + 16) = 0

 The eigen values of A are  = 1, 4, 4.
3-λ 1 1   x1   0 
The eigen vectors are given by  1 3-λ -1   x 2   0

 1 -1 3-λ   x 3  0
Case 1 =1
 1
Eigen vector X1 =  1 
 1 
Case 2  = 4
0
Eigen vector X2 =  1
 1 
a
 
Now assume X 3   b  to be mutually orthogonal with X1 and X2.
c
 
X 1 X 3  0 i.e.  a  b  c  0 
T
a b c
 i.e  
and X 2 X 3  0 i.e.  b  c  0 
T
2 1 1
 2
 
 X3   1 .
1
 
 1 0 2 
Hence the modal matrix M   1 1 1 
 1 1 1 
 1 0 2 
 3 6
 
The Normalized Modal Matrix is N   1 1 1 
 3 2 6
 1 1 1 

 3 2 6
Diagonalizing, we get
 1 1 1   1 0 2 
 3 3 3 3 1 1  3 6
   
D  N T AN   0 1 1   1 3 1  1 1 1 
 2 2   
 3 2 6
 1 1 3   1
 2
 1 1   1 1 
 6 6 6  3 2 6
1 0 0 
 0 4 0  = D(1, 4, 4)
0 0 4 
Problem 12. Reduce the Quadratic From 10 x12  2 x22  5x32  6 x2 x3 10 x3 x1  4 x1x2 into
canonical form by orthogonal reduction. Hence find the nature, rank, index and the
signature of the Q.F. Find also a nonzero set of values of X which will make the Q.F.
vanish.
 10 2 5 
 
Solution: Matrix of the given Q.F. is A   2 2 3  , which is a real and symmetric
 5 3 5 
 
matrix. The characteristic equation is   17  42  0
3 2
Solving, we get   0, 3, 14
1  1  3 
     
When   0, X 1   5  ; When   3, X 2  1 ; When   14, X 3   1 
 4  1 2
     
and X1 , X 2 , X 3 are mutually orthogonal since X1 , X 2  0, X 2 X 3  0 andX 3 X1  0
T T T
Normalizing these vectors we get the normalized model matrix

 1 1 3 
 42 3 14 
 
N   5 1 1 
 42 3 14 
 4 1 2 
 
 42 3 14 
Diagonalising we get D  N T AN
 D  12 , 3  in order
 D  0, 3, 14 
0 0 0 
 
i.e D   0 3 0  (i.e. the eigen values in order along the principal
 0 0 14 
 
diagonal).
Now to reduce the Q.F to C.F (.i.e Canonical form)
 y1 
 
Consider the orthogonal transformation X = NY where Y   y2 
y 
 3
Then the Q.F. X T AX becomes  NY  A  NY   Y T  N T AN  Y
T
= Y T DY since N T AN  D
0 0 0  y1 
  
  y1 y2 y3   0 3 0  y2 
0 0 14  
  y3 
 0 y1  3 y2  14 y3
2 2 2
Thus = 0 y1  3 y2  14 y3 is the Canonical form of the given Q.F. And the equations of
2 2 2
this transformation are got from X= NY.

 1 1 3 
 x1   42 3 14   y1 
    
x  NY   5 1 1   y2 
 2
x   42 3 14  
 3  4 1 2   y3 
 
 42 3 14 
1 1 3
 x1  y1  y2  y3
42 3 14
5 1 3
x2   y1  y2  y3
42 3 14
4 1 3
x3  y1  y2  y3
42 3 14
To get the non-zero set of values of x which make the Q.F zero we assume values
for y1 , y2 and y3 such that the C.F. vanishes.
i.e 0 y1  3 y2  14 y3 will vanish if y2  0, y3  0and y1 is any arbitrary value (for

2 2 2
simplicity sake, assume y1 as the denominator of the coeff. of y1 in the equations) let
y1  42
 x1 
1
42

42 
1
3
 0 
3
14
(0)
i.e. x1  1  0  0  1
III 1 y x2  5  0  0  5
and x3  4  0  0  4
Thus the set of values of x i.e 1, 5, 4  will reduce the given Q.F. to zero.
To find the rank, index, signature and nature using canonical form:
C.F. is 0 y1  3 y2  14 y3
2 2 2
 rank is 2 (no. of terms in C.F)

Index is 2 (no. of positive terms)
Signature of Q.F. = ( no. of positive terms) – (no. of negative terms) = 2
Nature of the Q.F. is positive semi definite.
Problem 13. Reduce the Q.F. 2 xy  2 yz  2 zx into a form of sum of squares. Find the
rank, index and signature of it. Find also the nature of the Q.F.
0 1 1
 
Solution: Matrix of the Q.F. is A   1 0 1 
1 1 0
 
Characteristic equation is   3  2  0 solving   2, 1, 1
3
 1
 
When   2, X 1  1
 1
 
When   1 (repeated twice) we get identical equations as x1  x2  x3  0
x x
x1  0  x2  x3  0 i.e. x2   x3 i.e. 2  3
1 1
Assume 0
 
 X 2  1
1
 
which is orthogonal with X 1.
a
 
Now to find X 3 orthogonal with both X1 and X 2 assume X 3   b 
c
 
if X 2T X 3  0, a  b  c  0 

if X 2T X 3  0, 0a  b  c  0 
a b c

i.e. 
2 1 1
2  2 
   
 X 3   1 i.e.  1 
 1 1
   
which is orthogonal with X1 and X 2 .
 1 0 3 
 3 2 6
 
Normalising these vectors we get N   1 1 1  and D  N AN
 3 2 6
 1 1 2 
 
 3 2 6 
2 0 0 
 
= D  1 , 2 , 3    0 1 0  .Consider the orthonormal transformation X = NY
 0 0 1
 
such that Q.F.is reduced to C.F.
The Q.F. is reduced as

X T AX   NY  A  NY 
T
 Y T  N T AN  Y
 Y T DY
 2 0 0  y1 
  
  y1 , y2 , y3 ,   0 1 0  y2 
 0 0 1 y 
  3 
 The C.F. is 2y12  y2 2  y32

rank of Q.F.is = no. of terms in C.F=3
index of Q.F. = no. of positive terms in C.F. = 1
signature of Q.F. = ( no. of positive terms) – (no. of negative terms)
= 1-2 = -1
Nature of the Q.F. is indefinite.
Problem 14. Reduce the quadratic form 8x12  7 x22  3x32  12x1 x2  4 x1 x3  8x2 x3 to the
canonical form by an orthogonal transformation. Find also the rank, index, signature and
the nature of the quadratic form.
Solution:
 8 6 2 
The matrix of the quadratic form is A   6 7  4
 2  4 3 
The eigen values of this matrix are 0, 3 and 15 and the corresponding eigen vectors are
1  2 2
 
X 1   2 , 
X2   1  , X 3   2 , which are mutually orthogonal.
2  2  1 
1/3 2/3 2/3 

The normalized modal matrix is N  2/3 1/3  2/3

2/3  2/3 1/3 
0 0 0 
and N AN = D  0 3 0 
T
0 0 15
Now the orthogonal transformation X = NY will reduce the given quadratic form to the
canonical form 0y12  3y22  15y32 .
Also rank = 2, index = 2, signature = 2. The quadratic form is positive semi definite.
Problem 15. Find the orthogonal transformation which reduces the quadratic form
2 x12 
2 x22  2 x1 x2  2 x2 x3  2 x1 x3 into the canonical form. Determine the rank, index,
2 x32
signature and the nature of the quadratic form.
Solution:
 2 1 1 
The matrix of the quadratic form is A   1 2  1
 1  1 2 
2 -  -1 1
The characteristic equation of A is -1 2- -1  0
1 -1 2-
Expanding  - 6 + 9 - 4 = 0
3 2
 = 1 is a root
Dividing 3 - 62 + 9 - 4 by  -1,
1 6 9 4
0 1 5 4
1 5 4 0
The remaining roots are given by 2-5 + 4 = 0
2 - 5 + 4 = ( - 1) ( - 4) = 0
i.e.,  = 1, 4
The eigen values of A are  = 4, 1, 1
Case 1 =4
2 - 4 - 1 1   x 1  0 
The eigen vectors are given by  - 1 2 - 4 - 1   x 2   0

 1 - 1 2 - 4  x 3  0
 2  1 1  1 - 1 - 2
  1  2  1  ~ 0 - 3 - 3 
   
 1  1  2 0 0 0 
 x1 - x2 - 2x3 = 0
-3x2 - 3x3 = 0
A solution is x3 = 1, x2 = -1, x1 = 1.
1
 The corresponding eigen vector is X1 =   1
 1 
Case 2 =1
2 - 1 - 1 1   x1  0
The eigen vectors are given by  - 1 2 - 1 - 1   x 2   0

 1 - 1 2 - 1  x3  0
 1 - 1 1  1  1 1
- 1 1 - 1 ~ 0 0 0
   
 1 - 1 1  0 0 0
 x1- x2 + x3 = 0
Put x3 = 0. We get x1 = x2 = 1. Let x1 = x2 = 1
1 
 The eigen vector corresponding to  = 1 is X2 = 1 
0 
X1 and X2 are orthogonal as X1T X 2 = 10 + (-1) 1 + 11 = 0.
a 
To find another vector X3 =  b  corresponding to  =1 such that it is orthogonal to both
 c 
X1 and X2 and satisfies x1- x2 + x3 = 0
i.e., X1.X3 = 0, X2.X3 = 0 and a – b + c = 0
i.e., 1.a -1.b + 1.c = 0, 1.a + 1.b + 0.c = 0 and a – b + c = 0.
i.e., a – b + c = 0 and a+b=0
i.e., a = -b and c = 2b
Put b =1, so that a = -1, c = 2
 1
 X 3   1 
 2 
 1 1  1
The modal matrix is  1 1 1 
 1 0 2 
 1/ 3 1/ 2  1/ 6
 
Hence the normalized modal matrix is N   1 / 3 1 / 2 1/ 6 
 1/ 3 0 2/ 
 6

 The required orthogonal transformation is X = NY will reduce the given quadratic
form to the canonical form.
C.F= 4 y12  y 22  y32
Rank of the quadratic form = 3, index = 3, signature = 3. The quadratic form is positive
definite.
St.JOSEPH’S COLLEGE OF ENGINEERING,CHENNAI-119
MA8251 – MATHEMATICS-II
UNIT I VECTOR CALCULUS
     
1.Vector differential operator   i j k .
x y z
     
2.Gradient of     i j k .
x y z
 
3. Divergence of F    F .
  
i j k
    
4. curl F    F 
x y z
F1 F2 F3
 
5.If F is solenoidal then   F  0
 
6. If F is irrotation al then   F  0
7.Maximum directional derivative = 


   a
8.Directional derivative of  in the direction of a = 
a
1  2
9.Angle between two normal to the surface is cos  = ( given points)
1 2
 
10.Unit normal vector , n 

dx dy dy dz dz dx
11.In the surface integral ds   , ds   , ds  
nˆ  k nˆ  i nˆ  j
12. Green’s Theorem
If P(x ,y) and Q(x ,y) are continuous function with continuous partial derivatives in a region R
 Q P 
of the xy plane bounded by a simple closed curve C, then  Pdx  Qdy    x  y  dxdy
C R 
Where C is the curve described in positive direction
13.Gauss Divergence Theorem
The surface integral of the normal component of a vector function F over a closed
surface S enclosing a volume V is equal to the volume integral of the divergence of F
taken throughout the volume
   

S
F  n dS =  divF dv =
V

V
  F dv where , n is the unit out ward normal to the
surface S .
14. Stoke’s Theorem


If S is an open surface bounded by a simple closed curve C and if a vector function F is
continuous and has continuous partial derivatives in S and on C, then
  
 curlF  ˆ
n ds =   dr
F
S C
Where n̂ is the unit vector normal to the surface.

Matrices Eigenvalues and Inverse

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Matrices Eigenvalues and Inverse

Uploaded by

Copyright:

Available Formats

www.Vidyarthiplus.

St. Joseph’s College of Engineering, Chennai 600 119.

1  1 1  1  1 1  0  2 5

 A6  25 A2  122 A   A3  4 A2  5 A  2I  A3  4 A2  11A  22I    10 A  44I 

ii. Premultiply ( I ) with A, then

1y we can prove that A3 Xi   i  Xi and so on Am Xi   i  Xi

i m are the eigen values of the Am (comparing with ( I ) i.e. by defn.)

iii. Premultiply ( I ) with A1 , then

When   1 (repeated 3 times)  we have to find 3 corresponding latent vectors.

i.e., (1 - ) [(2 - ) (3 - ) - 4] -1[0 + 4] +1[0 + 4(2 - )] =0

Solving we get the eigen value as   1,3,6

i.e., (-1)(2 - 8 + 16) = 0

Normalizing these vectors we get the normalized model matrix

this transformation are got from X= NY.

i.e 0 y1  3 y2  14 y3 will vanish if y2  0, y3  0and y1 is any arbitrary value (for

 rank is 2 (no. of terms in C.F)

The Q.F. is reduced as

 The C.F. is 2y12  y2 2  y32

1/3 2/3 2/3 

The eigen values of A are  = 4, 1, 1

UNIT I VECTOR CALCULUS

7.Maximum directional derivative = 

12. Green’s Theorem

14. Stoke’s Theorem

You might also like