Scribd
Upload
36 views9 pages

Polynomial Approximation

This document discusses polynomial approximation methods for numerical analysis. It introduces polynomial approximation for a known function, using Taylor series approximations to represent a function as an infinite polynomial series. The document provides examples of using Taylor series approximations to represent the functions e^x and sin(x)+cos(2x) as polynomials.

Uploaded by

Sergio H
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
36 views9 pages

Polynomial Approximation

This document discusses polynomial approximation methods for numerical analysis. It introduces polynomial approximation for a known function, using Taylor series approximations to represent a function as an infinite polynomial series. The document provides examples of using Taylor series approximations to represent the functions e^x and sin(x)+cos(2x) as polynomials.

Uploaded by

Sergio H
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/ 9

Numerical Methods

Polynomial Approximation

Universidad Panamericana Numerical Methods Dr. José Herrera Santos 1


Polynomial approximation for a known function
Consider a function f(x) for which, a
polynomial approximation in x = x0 is
f ( x) sought:

PN ( x) = c0 + c1 x + c2 x 2 + ... + cn x N
In order to simplify the algebra, the
following form for the polynomial is used:
( x0 , f ( x0 ))
PN ( x) = b0 + b1 ( x − x0 ) + b2 ( x − x0 ) 2 + ... + bn ( x − x0 ) N

If f(x) must be equal to Pn(x), then the n derivatives of both in x = x0 must be equal

PN(0) ( x) = b0 + b1 ( x − x0 ) + b2 ( x − x0 ) 2 + ... + bn ( x − x0 ) N → PN(0) ( x0 ) = b0


PN(1) ( x) = 1  b1 + 2  b2 ( x − x0 ) + 3  b3 ( x − x0 ) 2 + ... + N  bn ( x − x0 ) N −1 → PN(1) ( x0 ) = 1  b1
PN(2) ( x) = 1  2b2 + 2  3b3 ( x − x0 )1 + ... + (N − 1)  N  bn ( x − x0 ) N − 2 → PN(2) ( x0 ) = 1 2b2
PN(3) ( x) = 1  2  3b3 + ... + (N − 2)  (N − 1)  N  bn ( x − x0 ) N −3 → PN(3) ( x0 ) = 1  2  3b3

PN( k ) ( x) = 1  2  ...  (k − 1)  k  bk + ... + (N − k + 1)  ...  (N − 1)  N  bn ( x − x0 ) N − k → PN(k) ( x0 ) = k !bk

Universidad Panamericana Numerical Methods Dr. José Herrera Santos 2


Polynomial approximation for a known function
The coefficients of the polynomial are found from the last equation:

f ( k ) ( x0 )
P ( x0 ) = f
(k)
N
(k )
( x0 ) → bk =
k!

Therefore:

Polynomial approximation of degree n for the


N (k )
f ( x0 )
PN ( x) =  ( x − x0 )
k function f(x). As the evaluation point get away
k =0 k! from x0, more terms are necessary to have a
good approximation

f ( x)  PN ( x)
Remember that the number of terms is always
finite no matter how many of them are

f ( k ) ( x0 ) calculated. The series expansion is expected to
f ( x) =  ( x − x0 )
k

k! be equal to f(x) only if there an infinite number


k =0
of terms

Universidad Panamericana Numerical Methods Dr. José Herrera Santos 3


Taylor’s Theorem

Assume that f  C n +1[a, b] and let x0  [a, b]. Then, for every x  (a, b), there exist a number
c = c( x) (the number c depends on the value of x) that lies between x 0 and x such that

f ( x) = PN ( x) + RN ( x)

where,
N
f ( k ) ( x0 )
PN ( x) =  ( x − x0 )
k

k =0 k!
is the Taylor's polynomial of degree N around x0 and RN ( x) is the error of the approximation

Calculation of the error.


RN ( x) = f ( x) − PN ( x)
Notice that in x = x0 , RN ( x0 ) = 0

f ( N +1) (c)
( x - x0 )
N +1
RN ( x) = Estimation of the error
( N + 1)!

Universidad Panamericana Numerical Methods Dr. José Herrera Santos 4


Example 1

f ( x) = e x

Find the Taylor′s series for 𝑥0 = 0 up to 𝑁 = 15


For every single polynomial, evaluate 𝑥 = 1 and
calculate the error of the approximation

Universidad Panamericana Numerical Methods Dr. José Herrera Santos 5


Example 1 – cont.

Universidad Panamericana Numerical Methods Dr. José Herrera Santos 6


Example 2
f ( x) = Sin( x) + Cos (2 x)

Find the Taylor′s series for 𝑥0 = 0 up to 𝑁 = 15


For every single polynomial, evaluate 𝑥 = 1 and
calculate the error of the approximation

Universidad Panamericana Numerical Methods Dr. José Herrera Santos 7


Example 2 – cont.

Universidad Panamericana Numerical Methods Dr. José Herrera Santos 8


Problem

For the function 𝑓 𝑥 = 𝑒 𝑥

a) Find the polynomial approximation of degree 3 with center in 𝑥 = 1


b) Find the polynomial approximation of degree 4 with center in 𝑥 = 1
c) Find the polynomial approximation of degree 2 with center in 𝑥 = 3
d) For the cases a) , b) and c), calculate the relative error in x = 2

Universidad Panamericana Numerical Methods Dr. José Herrera Santos 9

You might also like