Assignment: Orthogonal Polynomials and Approximation Theory - Prof. Peter Wolfe, Assignments of Mathematical Methods for Numerical Analysis and Optimization

The second assignment for the course amsc/cmsc 666, taught by dr. Wolfe. The assignment covers topics such as the gram-schmidt procedure, orthogonal polynomials, chebychev polynomials, and polynomial approximation. Students are required to perform calculations related to these topics, including finding orthogonal polynomials, minimizing errors, and computing q-r factorizations.

Typology: Assignments

Pre 2010

Uploaded on 07/30/2009

koofers-user-7dt
koofers-user-7dt 🇺🇸

10 documents

1 / 1

Toggle sidebar

This page cannot be seen from the preview

Don't miss anything!

bg1
AMSC/CMSC 666 Dr. Wolfe ASSIGNMENT #2 Due March 9, 2009
1. Use the Gram-Schmidt procedure to determine the first three orthogonal polynomials
for [0,1] with weight function w(x) = ln(1/x).
2. Define Sn(x) = 1
n+1 T
n+1(x), n 0,with Tn(x) the Chebychev polynomial of degree
n. The polynomials Sn(x) are called Chebychev polynomials of the second kind.
(a) Show that {Sn(x)|n0}is an orthogonal family on [1,1] with respect to the
weight function w(x) = 1x2.
(b) Show that the family {Sn(x)}satisfies the same triple recursion formula as the
family {Tn(x)}.
(c) Given fC[1,1] solve the problem: Minimize
Z1
1p1x2[f(x)p(x)]2dx
where p(x) is allowed to range over all polynomials of degree n.
3. Exercise 15, p.186, Sto er and Bulirsch.
4. Compute (by hand) the Q-R factorization of A where
A=5 9
12 7 .
5. Let f(x) = ex,0x1, π1= the set of all polynomials of degree 1.Find the
following.
(a) p
1, the minimax approximation to fin π1.
(b) f, the least squares approximation to fin π1.Here we use the norm
||f||2=qR1
0f(x)2dx.
(c) The function in π1interpolating fat the Chebychev points 1
2(1 ±1
2).
Compare the three errors in the uniform () norm and the 2-norm. Note: You
can use MATLAB to compute the integrals (either numerically or symbolically).
6. (MATLAB) Let xi=i1
10 , i = 1,···11,yi=exi.Find the polynomial of degree n
fitting this data in the sense of least squares for n= 1,2,3. Plot these polynomials
along with exfor 0 x1.The MATLAB commands POLYFIT and POLYVAL
give you exactly what you need.

Partial preview of the text

Download Assignment: Orthogonal Polynomials and Approximation Theory - Prof. Peter Wolfe and more Assignments Mathematical Methods for Numerical Analysis and Optimization in PDF only on Docsity!

AMSC/CMSC 666 Dr. Wolfe ASSIGNMENT #2 Due March 9, 2009

  1. Use the Gram-Schmidt procedure to determine the first three orthogonal polynomials for [0, 1] with weight function w(x) = ln(1/x).
  2. Define Sn(x) = (^) n+1^1 T (^) n′+1(x), n ≥ 0 , with Tn(x) the Chebychev polynomial of degree n. The polynomials Sn(x) are called Chebychev polynomials of the second kind. (a) Show that {Sn(x) | n ≥ 0 } is an orthogonal family on [− 1 , 1] with respect to the weight function w(x) =

1 − x^2. (b) Show that the family {Sn(x)} satisfies the same triple recursion formula as the family {Tn(x)}. (c) Given f ∈ C[− 1 , 1] solve the problem: Minimize ∫ (^1)

− 1

1 − x^2 [f (x) − p(x)]^2 dx

where p(x) is allowed to range over all polynomials of degree ≤ n.

  1. Exercise 15, p.186, Stoer and Bulirsch.
  2. Compute (by hand) the Q-R factorization of A where

A =

  1. Let f (x) = ex, 0 ≤ x ≤ 1 , π 1 = the set of all polynomials of degree ≤ 1. Find the following. (a) p∗ 1 , the minimax approximation to f in π 1. (b) f ∗, the least squares approximation to f in π 1. Here we use the norm ||f || 2 =

1 0 f^ (x)

(^2) dx. (c) The function in π 1 interpolating f at the Chebychev points 12 (1 ± √^12 ). Compare the three errors in the uniform (∞) norm and the 2-norm. Note: You can use MATLAB to compute the integrals (either numerically or symbolically).

  1. (MATLAB) Let xi = i− 101 , i = 1, · · · 11 , yi = exi^. Find the polynomial of degree ≤ n fitting this data in the sense of least squares for n = 1, 2 , 3. Plot these polynomials along with ex^ for 0 ≤ x ≤ 1. The MATLAB commands POLYFIT and POLYVAL give you exactly what you need.