Math Help Forum

Math Help Forum Feed Site Feed

Go Back   Math Help Forum > University Math Help > Calculus
Reply
 
Thread Tools Display Modes
  #1  
Old May 9th, 2009, 02:54 PM
Junior Member
 
Join Date: Mar 2009
Posts: 26
Country:
Thanks: 5
Thanked 0 Times in 0 Posts
Collegeboy110 is on a distinguished road
Question Taylor Polynomial Error

I cant figure this one out either. Thank You.

Find an upper bound for the error in using a 3rd degree Taylor Polynomial to approximate (1.3)^(1/3).


Last edited by Collegeboy110; May 10th, 2009 at 01:25 PM.
Reply With Quote
Advertisement
 
  #2  
Old May 9th, 2009, 02:58 PM
TheEmptySet's Avatar
Chair of Approximate Accuracy

 
Join Date: Feb 2008
Location: Corvallis, Oregon, USA
Posts: 1,988
Country:
Thanks: 275
Thanked 908 Times in 854 Posts
TheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud of
Default

Quote:
Originally Posted by Collegeboy110 View Post
I cant figure this one out either. Thank You.

Find an upper bound for the error in using a 3rd degree Taylor Polynomial to approximate (1.3)^(1/3).
We need more info. Where is the series centered? that will change the error. 0,1,8...
Reply With Quote
  #3  
Old May 9th, 2009, 03:02 PM
Junior Member
 
Join Date: Mar 2009
Posts: 26
Country:
Thanks: 5
Thanked 0 Times in 0 Posts
Collegeboy110 is on a distinguished road
Default

The question is exactly written like this. It does not give me more info.
Reply With Quote
  #4  
Old May 9th, 2009, 03:16 PM
TheEmptySet's Avatar
Chair of Approximate Accuracy

 
Join Date: Feb 2008
Location: Corvallis, Oregon, USA
Posts: 1,988
Country:
Thanks: 275
Thanked 908 Times in 854 Posts
TheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud of
Default

Quote:
Originally Posted by Collegeboy110 View Post
The question is exactly written like this. It does not give me more info.
so the taylor expansion looks like this

T(x)=1+\frac{1}{3}(x-1)-\frac{1}{9}(x-1)^2+\frac{5}{81}(x-1)^3-\frac{10}{243}(x-1)^4+...

Since this is an alternating series (after the 1st term)

The error can be no larger than the next term in the series so we get

\frac{10}{243}(.3)^4 =0.000\bar3
Reply With Quote
  #5  
Old May 9th, 2009, 03:41 PM
Junior Member
 
Join Date: Mar 2009
Posts: 26
Country:
Thanks: 5
Thanked 0 Times in 0 Posts
Collegeboy110 is on a distinguished road
Default

Can u explain it to me further please. I cannot connect the answer to what they are asking me. The 3rd degree polynomial would it not be 5/81 (x-1)^3
Reply With Quote
  #6  
Old May 9th, 2009, 03:44 PM
TheEmptySet's Avatar
Chair of Approximate Accuracy

 
Join Date: Feb 2008
Location: Corvallis, Oregon, USA
Posts: 1,988
Country:
Thanks: 275
Thanked 908 Times in 854 Posts
TheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud of
Default

Quote:
Originally Posted by Collegeboy110 View Post
Can u explain it to me further please. I cannot connect the answer to what they are asking me. The 3rd degree polynomial would it not be 5/81 (x-1)^3

T_3(x)=1+\frac{1}{3}(x-1)-\frac{1}{9}(x-1)^2+\frac{5}{81}(x-1)^3

Is the 3rd degree taylor expansion centered at 1.

You need all of the terms of degree \le 3
Reply With Quote
The following users thank TheEmptySet for this useful post:
Donate to MHF
  #7  
Old May 9th, 2009, 03:51 PM
Junior Member
 
Join Date: Mar 2009
Posts: 26
Country:
Thanks: 5
Thanked 0 Times in 0 Posts
Collegeboy110 is on a distinguished road
Default

Ok, so for the formula i get that 1.3-1 = .3.. how did u get 10/243?

Quote:
Originally Posted by TheEmptySet View Post
so the taylor expansion looks like this

T(x)=1+\frac{1}{3}(x-1)-\frac{1}{9}(x-1)^2+\frac{5}{81}(x-1)^3-\frac{10}{243}(x-1)^4+...

Since this is an alternating series (after the 1st term)

The error can be no larger than the next term in the series so we get

\frac{10}{243}(.3)^4 =0.000\bar3
Reply With Quote
  #8  
Old May 9th, 2009, 04:00 PM
TheEmptySet's Avatar
Chair of Approximate Accuracy

 
Join Date: Feb 2008
Location: Corvallis, Oregon, USA
Posts: 1,988
Country:
Thanks: 275
Thanked 908 Times in 854 Posts
TheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud of
Default

Quote:
Originally Posted by Collegeboy110 View Post
Ok, so for the formula i get that 1.3-1 = .3.. how did u get 10/243?

It is the next term in the Taylor series

\frac{10}{243}(x-1)^4
Reply With Quote
  #9  
Old May 9th, 2009, 07:52 PM
Member
 
Join Date: May 2009
Posts: 127
Country:
Thanks: 6
Thanked 79 Times in 63 Posts
the_doc will become famous soon enough
Default

"TheEmptySet" just computed the Taylor expansion of T(x) = x^{1/3} centered at x=1. So to compute the coefficients you do the usual thing of sequentially evaluating each derivative of T(x) at x=1. So to work out the coefficient you are struggling with compute the fourth derivative i.e. (1/3)\cdot (-2/3) \cdot (-5/3) \cdot(-8/3) x^{-11/3} = -80/81 \, x^{-11/3}, evaluate at x=1 then finally divide by 4! to give you the desired coefficient.
Reply With Quote
  #10  
Old May 10th, 2009, 12:32 PM
Junior Member
 
Join Date: Mar 2009
Posts: 26
Country:
Thanks: 5
Thanked 0 Times in 0 Posts
Collegeboy110 is on a distinguished road
Default

So How would this series change if it is centered at x= 0?

Quote:
Originally Posted by the_doc View Post
"TheEmptySet" just computed the Taylor expansion of T(x) = x^{1/3} centered at x=1. So to compute the coefficients you do the usual thing of sequentially evaluating each derivative of T(x) at x=1. So to work out the coefficient you are struggling with compute the fourth derivative i.e. (1/3)\cdot (-2/3) \cdot (-5/3) \cdot(-8/3) x^{-11/3} = -80/81 \, x^{-11/3}, evaluate at x=1 then finally divide by 4! to give you the desired coefficient.
Reply With Quote
  #11  
Old May 10th, 2009, 12:43 PM
TheEmptySet's Avatar
Chair of Approximate Accuracy

 
Join Date: Feb 2008
Location: Corvallis, Oregon, USA
Posts: 1,988
Country:
Thanks: 275
Thanked 908 Times in 854 Posts
TheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud of
Default

Quote:
Originally Posted by Collegeboy110 View Post
So How would this series change if it is centered at x= 0?

It cannot be centered at zero becuase f(x)=x^{1/3} is not differentable at x=0

f'(x)=\frac{1}{3}\frac{1}{x^{2/3}} is undefined at zero.
Reply With Quote
  #12  
Old May 10th, 2009, 12:49 PM
Junior Member
 
Join Date: Mar 2009
Posts: 26
Country:
Thanks: 5
Thanked 0 Times in 0 Posts
Collegeboy110 is on a distinguished road
Default

The question says Find an upper bound for the error in using the 3rd degree polynomial to approximate the cubed root of 1.3.. which I assume is (1.3)^(1/3).. i asked my proffesor he said centered at 0.
Reply With Quote
  #13  
Old May 10th, 2009, 12:57 PM
TheEmptySet's Avatar
Chair of Approximate Accuracy

 
Join Date: Feb 2008
Location: Corvallis, Oregon, USA
Posts: 1,988
Country:
Thanks: 275
Thanked 908 Times in 854 Posts
TheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud ofTheEmptySet has much to be proud of
Default

ummmm... I don't want to contradict your prof. but there is not a taylor expansion at 0 for f(x) = x^{1/3}

A taylor polynomial centered at a is given by the formula

T(x)=\sum_{n=0}^{\infty}\frac{f^{n}(a)}{n!}(x-a)^n

if we try to center it a zero we get


T(x)=\sum_{n=0}^{\infty}\frac{f^{n}(0)}{n!}(x)^n

This is a big problem becuse

f'(x)=\frac{1}{3}\frac{1}{x^{2/3}}

Is undefined at zero and so is every derivative after this one
Reply With Quote
  #14  
Old May 10th, 2009, 01:29 PM
Junior Member
 
Join Date: Mar 2009
Posts: 26
Country:
Thanks: 5
Thanked 0 Times in 0 Posts
Collegeboy110 is on a distinguished road
Default

I attached a similar problem (0.5)^(1/3) at x= 0..
Reply With Quote
  #15  
Old May 10th, 2009, 01:45 PM
Member
 
Join Date: May 2009
Posts: 127
Country:
Thanks: 6
Thanked 79 Times in 63 Posts
the_doc will become famous soon enough
Default

Firstly, centred at zero simply means it is in a power series of x rather than x minus something.
The above post is correct but I suspect that your Prof. is actually wanting you to find a Taylor expansion centered at x=0 for the function T(x) = (1+x)^{1/3} hence there is no contradiction. This is just a simple binomial expansion i.e.
(1+x)^{1/3} = 1 + \frac{1}{3} x + \left( \frac{1}{3} \right) \cdot \left(-\frac{2}{3} \right)\frac{x^2}{2!} + \left( \frac{1}{3} \right) \cdot \left(-\frac{2}{3} \right) \cdot \left(-\frac{5}{3} \right) \frac{x^3}{3!} + \left( \frac{1}{3} \right) \cdot \left(-\frac{2}{3} \right) \cdot \left(-\frac{5}{3} \right) \cdot\left(-\frac{8}{3} \right) \frac{x^4}{4!} +\ldots
which after tidying up gives you
(1+x)^{1/3} = 1 + \frac{1}{3} x - \frac{1}{9} x^2 + \frac{5}{81} x^3 - \frac{10}{243} x^4 +\ldots
which as you can see has all the same coefficients and in fact numerically amounts to the same power series (x=0.3 instead of 1.3)!

So in conclusion you can centre at 0 but you have to choose a function which is differentiable at 0.
Reply With Quote
The following users thank the_doc for this useful post:
Donate to MHF
Reply

Tags
error, taylor series

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off
Forum Jump


All times are GMT -7. The time now is 09:25 PM.


Powered by vBulletin® Version 3.7.3
Copyright ©2000 - 2009, Jelsoft Enterprises Ltd.
SEO by vBSEO 3.2.0 ©2008, Crawlability, Inc.
©2005 - 2009 Math Help Forum


Math Help Forum is a community of maths forums with an emphasis on maths help in all levels of mathematics.
Register to post your math questions or just hang out and try some of our math games or visit the arcade.