Scribd
Upload
32 views

UECM1693/UECM2623/UGCM2623 Tutorial N3: Solution of Linear Systems

This document discusses solving linear systems using Jacobi, Gauss-Seidel, and power iteration methods. It provides examples applying these methods to solve systems and estimating eigenvalues and eigenvectors of matrices. Specific problems addressed include solving systems, estimating dominant eigenvalues, applying Gerschgorin's theorem, and using inverse and shifted inverse power methods.

Uploaded by

freeload
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
32 views

UECM1693/UECM2623/UGCM2623 Tutorial N3: Solution of Linear Systems

This document discusses solving linear systems using Jacobi, Gauss-Seidel, and power iteration methods. It provides examples applying these methods to solve systems and estimating eigenvalues and eigenvectors of matrices. Specific problems addressed include solving systems, estimating dominant eigenvalues, applying Gerschgorin's theorem, and using inverse and shifted inverse power methods.

Uploaded by

freeload
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

UECM1693/UECM2623/UGCM2623

Tutorial N3: Solution of Linear Systems


1. Perform three iterations of the methods of Jacobi and Gauss-Seidel to obtain approximate
solutions of the following linear systems. In each case, use an initial guess of
x(0) = y (0) = z (0) = 0.
Make sure the matrix in each case is strictly diagonally dominant (rearrange it if necessary).

4x + y + z = −1
(a) x + 6y + 2z = 0
x + 2y + 4z = 1
x − 4y + z = −4
(b) 5x + y − z = 4
2x + 2y − 5z = −6

Answer: (a) (−0.3177, −0.0521, 0.3490) , (−0.3281, −0.0634, 0.3637)


(b) (0.8840, 1.6900, 2.1360) , (0.9080, 1.7970, 2.2820)

2. Use Gauss-Seidel method to solve the system


2x − 7y = −34
.
5x + y = −11

Continue the iterations until two successive approximations agree to two decimal places.
Answer: x = −3.00, y = 4.00
 
3 4
3. Let A = .
1 3
(a) Use the power method with scaling
 to find a dominant eigenvector of A. Start with the
1
initial approximation x0 = . Round off all computations to 3 decimal places and stop
1
after 3 iterations.
(b) Use the result of part (a) and the Rayleigh quotient to approximate the dominant eigenvalue.
(c) Find an estimate for the percentage error in the approximation of the dominant eigenvalue.
 
1.000
Answer: (a) (b)5.007 (c) 0.499%
0.503
 
5 1
4. (a) Apply Gerschgorin’s theorem to conclude the eigenvalues of A = .
3 2
(b) Apply one iteration of Inverse Power Method with  scaling
 to find the smallest
 eigenvalue
5 1 0.25
and the corresponding eigenvector of matrix A = with x0 = .
3 2 −1
 
0.2609
Answer: (a) 4 ≤ λ1 ≤ 6, −1 ≤ λ2 ≤ 5 (b) 1.2099,
−1

1
 
3 10
5. Apply one iteration of Shifted Inverse Power Method with scaling to matrix A = to find
2 4 
5
the eigenvalue nearest to a = −0.9 and its corresponding eigenvector. Start with x0 = .
  −2
−1
Answer: −1,
0.4

 
2 −1 1
6. Apply the shifted inverse power method with scaling (3 steps) to A = −1 3 2 to find the
1 2 3  
1
eigenvalue nearest to a = 4.9 and a corresponding eigenvector. Start with the vector x0 = 1 .

1
 −1  
−2.9 −1 1 −0.4189 0.1074 −0.1074
(Hint: −1 −1.9
 2  = 0.1074 4.8443 5.1557 )

1 2 −1.9 −0.1074 5.1557 4.8443
 
−0.0001
Answer: 5,  1 
0.9999

You might also like