Math 3607: Homework 9 2021
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
Math 3607: Homework 9
2021
TOTAL: 30 points
• Problems marked with are to be done by hand; those marked with are to be solved using a computer.
• Important note. Do not use Symbolic Math Toolbox. Any work done using sym or syms will receive NO credit.
• Another important note. When asked write a MATLAB function, write one at the end of your live script.
1. (Low-rank approximation using SVD; image compression) Load hubble_gray.jpg, which is a grayscale image taken by the Hubble Space Telescope, convert it to a matrix of floating point pixel intensities, and then display the image in MATLAB by
A = imread(’hubble_gray . jpg’);
imshow(A);
Following the demo in Lecture 28 as a guide,
(a) Plot the singular values σ 1 , σ2 , . . . , σn of A on a log scale (using semilogy).
(b) Plot the accumulation of singular values of A.
(c) Compute the best approximations of A of rank 2, 20, and 120 and display the corresponding images using subplot.
Figure 1: NGC 3603 (Hubble Space Telescope).
104
103
102
101
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
Accumulation of singular values
100 200 300 400 500 600 700 800 900 1000
matrix rank
Figure 2: Example outputs for part (a) on the left and part (b) on the right.
rank = 2, ratio = 0.004
rank = 20, ratio = 0.037
rank = 120, ratio = 0.219
Figure 3: Example output for part (c)
2. (Annuity with fzero; FNC 4.1.4) A basic type of investment is an annuity: One makes monthly deposits of size P for n months at a fixed annual interest rate r, and at maturity collects the amount
˜ ˆ1 ` ˙n ´ 1¸ .
Say you want to create an annuity for a term of 300 months and final value of $1,000,000. Using fzero, make a table of the interst rate you will need to get for each of the different contribution values P “ 500, 550, . . . , 1000.
3. (Lambert’s W function; FNC 4.1.6) Lambert’s W function is defined as the inverse of xex . That is, y “ W pxq if and only if x “ yey . Write a function y = lambertW(x) that computes W using fzero. Make a plot of W pxq for 0 ď x ď 4.
4. (Fixed-point iteration; adapted from FNC 4.2.1 and 4.2.2.) In each case below,
• gpxq “ ´x ` ¯ , r “ 3.
• gpxq “ π ` sinpxq, r “ π .
• gpxq “ x ` 1 ´ tanpx{4q, r “ π .
(a) Show that the given gpxq has a fixed point at the given r and that fixed point iteration can converge to it.
(b) Apply fixed point iteration in MATLAB and use a log-linear graph (using semilogy) of the error to verify (linear) convergence. Then use numerical values of the error to determine an approximate value for the rate σ .
5. (Convergence of Newton’s method) Answer the following questions by hand, without using MATLAB.
(a) Discuss what happens when Newton’s method is applied to find a root of fpxq “ signpxqa|x|,
starting at x0 ‰ 0. 1
1 sign pxq is 1 if x ą 0, ´1 if x ă 0, and 0 if x “ 0.
(b) In the case of a multiple root, where f prq “ f1 prq “ 0, the derivation of the quadratic error convergence is invalid. Redo the derivation to show that in this circumstance and with f2 prq ‰ 0 the error converges only linearly.
2022-04-06