Scribd
Upload
71 views3 pages

2 Homework Set No. 2: Problem 1

This document outlines 5 homework problems related to numerical analysis and interpolation using Matlab. Problem 1 involves deriving binomial series and using them to compute approximations. Problem 2 uses Lagrange interpolation to find a polynomial that matches given values. Problem 3 provides bounds on linear interpolation error. Problem 4 derives and tests a formula to approximate derivatives. Problem 5 involves writing Matlab functions to compute divided differences and polynomial values from them, and using these functions to compare interpolation using equally spaced and Chebyshev nodes.

Uploaded by

kelbmuts
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
71 views3 pages

2 Homework Set No. 2: Problem 1

This document outlines 5 homework problems related to numerical analysis and interpolation using Matlab. Problem 1 involves deriving binomial series and using them to compute approximations. Problem 2 uses Lagrange interpolation to find a polynomial that matches given values. Problem 3 provides bounds on linear interpolation error. Problem 4 derives and tests a formula to approximate derivatives. Problem 5 involves writing Matlab functions to compute divided differences and polynomial values from them, and using these functions to compare interpolation using equally spaced and Chebyshev nodes.

Uploaded by

kelbmuts
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/ 3

2 Homework Set No.

2
NB! If a problem is not specified to use Matlab, then you should not use Matlab.

Preparation for Matlab home works: Read chapter 3 in the notes of Gockenbach.

Problem 1

(a). The Taylor series for (1 + x)n is also known as the binomial series. It states
that
n(n − 1) 2 n(n − 1)(n − 2) 3
(1 + x)n = 1 + nx + x + x + · · · (x2 < 1)
2! 3!
Derive the series and for n√= 2, n = 3 and n = 1/2, give its particular form. Then use
the last form to compute 1.0001 correct to 15 decimal places (rounded).

(b). (continuation) Use the previous problem to obtain a series for (1 + x2 )−1 .

Problem 2

(a). Use the Lagrange interpolation process to obtain a polynomial of least degree that
assumes these values:
x 0 2 3 4
y 7 11 28 63

(b). (Continuation) For the points in the table of (a), find the Newton form of the in-
terpolating polynomial. Show that the two polynomials obtained are identical, although
their forms may differ.

(c). The polynomial p(x) = x4 − x3 + x2 − x + 1 has the values shown.


x -2 -1 0 1 2 3
p(x) 31 5 1 1 11 61
Find a polynomial q(x) that takes these values (you don’t need expand it):
x -2 -1 0 1 2 3
q(x) 31 5 1 1 11 30
(Hint: This can be done with little work.)

Problem 3

(a). Show directly that the maximum error associated with linear interpolation is
bounded by 18 (x1 − x0 )2 M , where M = maxx0 ≤x≤x1 |f 00 (x)|.

1
(b). An interpolating polynomial of degree 20 is to be used to approximate e−x on the
interval [0, 2]. How accurate will it be?

Problem 4

(a). Derive the approximation formula


1
f 0 (x) ≈ [4f (x + h) − 3f (x) − f (x + 2h)]
2h
and show that its local truncation error is of the form 31 h2 f 000 (ξ).
(Hints: Consider the polynomial that interpolates the function at the following 3 points:
x, x + h, x + 2h.)

(b). Then, use the formula to find the derivative of f (x) = tan(x) at x = 1.0 for
h = 0.1, 0.01 and 0.001. Compare the error with the error bound for this formula.

(c). Use Matlab to compute the approximation to tan(x) at x = 1.0, with this formula,
with smaller and smaller values of h. Plot the error as a function of h, use log scale
(loglog). What is the best result you manage to get? What is your comments to the
result?
What to hand in for part c): Hand in the Matlab script file, the plot of error, and
your comments.

Problem 5: Divided difference in Matlab.

Part (A): Write two functions in Matlab. The first one should read in two vectors x
and y, and return a table (a matrix) of the divided difference values. This means, the
first line of your divdiff.m file should be:

function a=divdiff(x,y)

The second function should read in the table of divided difference values, the x-vector,
and a vector t, and return the polynomial values computed at points given in the vector
t. This means, your file polyvalue.m should start with the following line:

function v=polyvalue(a,x,t)

What to hand in: Hand in the files divdiff.m and polyvalue.m.

Part (B) : Here you can test your Matlab functions in (A). Using 21 equally spaced
nodes on the interval [−5, 5], find the interpolating polynomial p of degree 20 for the
function f (x) = (x2 + 1)−1 . Plot the functions f (x) and p(x) together at 41 equally

2
spaced points, including the nodes. Plot the error e(x) = |f (x) − p(x)| also. Give your
comments.
What to hand in: Hand in your Matlab script file, and the plots of f (x) with p(x),
and the error, and your comments.

Part (C): With the same problems in (B), now using Chebyshev nodes xi = 5 cos(iπ/20),
where 0 ≤ i ≤ 20, and nodes xi = 5 cos[(2i + 1)π/42], where 0 ≤ i ≤ 20. Plot f (x) with
p(x) for both cases, and the errors. Give your comments.
What to hand in: Hand in your Matlab script file, and the plots of f (x) with p(x),
and the error, and your comments.

You might also like