Scribd
Upload
41 views

Problem Set 3

This document contains instructions for Problem Set 3 in a Numerical Computing course. It includes: 1) Three exercises involving analyzing the convergence of fixed point iterations with different functions. 2) An exercise showing a function stays within a given interval and has an attracting fixed point. 3) Solving a system of equations using Gaussian elimination and finding the number of iterations for Newton's method to solve a polynomial.

Uploaded by

Brian Dunn
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
41 views

Problem Set 3

This document contains instructions for Problem Set 3 in a Numerical Computing course. It includes: 1) Three exercises involving analyzing the convergence of fixed point iterations with different functions. 2) An exercise showing a function stays within a given interval and has an attracting fixed point. 3) Solving a system of equations using Gaussian elimination and finding the number of iterations for Newton's method to solve a polynomial.

Uploaded by

Brian Dunn
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

D.

Schwendeman
Numerical Computing

Problem Set 3

Due:
Tuesday, 2/18/14

1. Text exercises 16 on page 41.


1
Wed like to see whether the fixed point iteration converges to the point x = 4 3
1
It is not hard to verify that x = 4 3 is indeed the fixed point for all of the three g(x),
should the iteration converge.
1. (a) g(x) = 2x , g 0 (x) = x

3
2

= 12

2. (b) g(x) = 34 x + x2 , g 0 (x) = 41


3. (c) g(x) = 23 x+ 3x42 , g 0 (x) = 23 83 x3 = 0
We see that in all three cases, |g 0 (x)| < 1, and that 1 > |ga0 | > |gb0 | > |gc0 | = 0. Hence we
have that gc (x) converges faster than gb (x) than ga (x).
2. g(x) = x + ln(2 x)
1. We wish to show that g(x) stays in the interval [1/2, 4/3] when taking x from that
interval.
1
= 0 x = 1 g(1) = 1. Since g(x) is continuous on [1/2, 4/3], closed,
g 0 (x) = 1+ x2
bounded, we know that g(x) takes maximun and minimun values from end point
values and local extrema. Therefore possible candidates for maximum and minimum
are: g( 12 ) = 0.9055, g( 34 ) = 0.9279 and g(1) = 1. Hence, 21 < g( 12 ) < g(1) < 43 . We
can now argue that g(x) stays in [1/2, 4/3] whenever taking x from that interval. By
theorem, we are sure to have a fixed point.

2. We wish to show that the fixed point is attracting.


1
0
g 00 (x) = (x2)
2 < 0 for all x [1/2, 4/3] g (x) is monotone we just need to
verify the value at two endpoints and check whether they lie within (-1, 1).

g 0 ( 21 ) = 31 , and g 0 ( 43 ) = 21 . Hence |g 0 (x)| < 1 for all x [1/2, 4/3] the fixed point
is attracting.
3. Text exercise 4(b) on page 59.
f (x) = x3 x2 5x 3 = x3 x2 2x 3(x + 1) = x(x 2)(x + 1) 3(x + 1) =
(x + 1)(x2 2x 3) = (x + 1)2 (x 3). We see that r = 1 is a double root and r = 3 is a
single root.
1. r = 1
1
1
m = 2 S = (21)
2 = 2 ei+1 = 2 ei . We have linear convergence rate.

2. r = 3
00

(3) 1
M = ff 0 (3)
= 2 ei+1 = 12 e2i . We have quadratic convergence rate.

4. Again there are two parts for this question the code for Newtons method could be found
on the course website.
The code for the second question is as follows.

5. Text exercises 2(a) and 6 on page 78


1. Gaussian elimination:
The original system of equations is equivalent to

1 1
2 2 1 2
1 1 12 1
4
0
5 0 1
1 2 1 0 5
0 1 2 5
0 0
2 1 1
3

1 0 1 1
0 1 0 1
0 0 1 2
Therefore we have x = 1,y = 1,z = 2.
3

2. Number of time needed = 32 5000


0.005 = 50
3 = 16.67 = 17s
50002

21
0
2

1
1
1 0
4
0

1 21
1
0
0
1

1
1
2

You might also like