3.2 Matrix Inverse: Zar97/09/03 Numerical Methods For Engineers Page 3-22
3.2 Matrix Inverse: Zar97/09/03 Numerical Methods For Engineers Page 3-22
3.2 Matrix Inverse: Zar97/09/03 Numerical Methods For Engineers Page 3-22
page 3-22
*
0 1 0 b2
0 0 1 b*
3
(3.17)
x2 = b*2
x3 = b*3
Procedures:
'
1 a12
0 1
0 0
'
a13
a'23
1
'
b1' R1' R1 a12
R2
'
b2
b'3
''
'
1 0 a13
b1'' R1'' R1' a13
R3
'
' ''
'
0 1 a 23 b2 R2 R2 a23R3
0 0 1 b'
3
''
1 0 a13
b1''
'
'
0 1 a 23 b2
0 0 1 b'
3
1 0 0 b1*
*
0 1 0 b2
0 0 1 b'
3
zar97/09/03
page 3-23
5
3
5
From example 3.3, the upper triangular matrix of the above augmented matrix is
1 1 1
4
4
0
1
0 0 1
1 1 1
4
4
0
1
0 0 1
3
4
1
1
3 R' R 1 R
1
1
4 2
4
'
1 R2 R2 R 3
1 0 0 1
0 1 0 2
0 0 1 1
zar97/09/03
page 3-24
Matrix Inversion
The elimination method can be used to get the inverse of the matrix. In the case of
our interest, If the matrix of the coefficient A is nonsingular, then there exists the
inverse of A denoted by A-1 such that
A-1 A = A A-1 = I
Then, we can show that (from
Ax = b ),
A-1 Ax = A-1 b
I x = A-1 b
x = A-1 b
(3.18)
From Gauss-Jordan elimination, we can represent the entire processes in the form
[A:I]
[ I : A-1 ]
(3.19)
in other words, the Gauss-Jordan elimination is used primarily for obtaining the
inverse of the matrix. Lets illustrate the method of matrix inverse through the
following example.
Remark:
The computation of an inverse requires about 3 times as many operations as the
Gauss elimination method and the effect of RErs is more pronounced. Therefore, it
is only used in cases when the matrix inverse is actually needed.1
zar97/09/03
page 3-25
Original matrix
Operation
5 2 1 1 0 0
4 1 1 0 1 0
2 3 3 0 0 1
Starting matrix
Operation
2
1
1
1
0 0
5
5
5
3
9
4
1 0
0 5
5
5
0 19 13 2 0 1
5
5
5
Starting matrix
Operation
1 2 1 1 0
0
5
5
5
4
5
0
0 1 3 3
3
0 0 1 1 19 1
3
42
14
Starting matrix
1 0 1 1
3
1
0
1
0
3
0 0 1 1
3
Answer:
2
3
13
42
19
42
Operation
0
3
14
1
14
Resultant matrix
1
1
1 2
5
5
5
3
9
4
0
5
5
5
0 19 13 2
5
5
5
0 0
1 0
0 1
Resultant matrix
1
1
1 2
5
5
5
4
0
1
3
3
0 0 14 14
3
Resultant matrix
1 0 1 1
3
1
0
1
0
3
0 0 1 1
3
0
1
0
5
3
19
3
2
3
13
42
19
42
Resultant matrix
1 0 0 0 3
14
1
13
0
1
0
3
42
0 0 1 1 19
3
42
0
3
14
1
14
1
14
3
14
1
14
zar97/09/03
page 3-25.2
Matrix Conditioning
Matrix conditioning is a concept introduced to avoid any ill-conditioned system.
Vector and matrix norms:
n
2
x
xi
i 1
n
A
i 1
Euclidean norm
Frobenius norm
aij
j 1