|
[Date Index]
[Thread Index]
[Author Index]
Re: Fitting a function to a list (newbie)
- To: mathgroup at smc.vnet.net
- Subject: [mg21884] Re: Fitting a function to a list (newbie)
- From: "Hans Michel" <hans at dorsai.org>
- Date: Wed, 2 Feb 2000 22:54:40 -0500 (EST)
- References: <872jej$civ@smc.vnet.net>
- Sender: owner-wri-mathgroup at wolfram.com
Math Prof wrote
<mathprof at bigfoot.com> wrote in message news:872jej$civ at smc.vnet.net...
> I'm new to Mathematica 4.0, so I apologize if this question is overly
> trivial (and please let me know if there's a place for trivial
> question).
>
> I have a list that I'd like to fit to the function:
>
> A + B*Sin[C*x+D]
>
> In other words, I'd like to find the (approximate numeric) values of
> A,B,C,D that yield the best fit to my data.
>
> I tried using Fit[], but it only applies to linear combinations of
> functions, and couldn't optomize the values of C and D.
>
> For this specific problem, I'm guessing some sort of Fourier transform
> might work, but I'm interested in a more general solution -- for example
> if Sin[] were replaced with a more complicated function.
>
> Any help appreciated!
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
>
I don't have an example for you but I would suggest checking a copy of
Exploring Mathematics With Mathematica by Theodore Gray and Jerry Glynn
Chapter 14 Fourier for Sound Page 240-251 may be relevant. The version off
my shelf is from 1991. In any case some of the functions they wrote are
available in MathSource. [NicerFourier and Resample (with some caveats i.e.,
x values must range between 0 and 2Pi)]
Also I would suggest looking into NonlinearFit
Statistics`NonlinearFit`
And also Using the Help file.
Prev by Date:
Re: ContourPlot in arbitrary shape domain.
Next by Date:
Double-sided printing, one side at a time ....
Previous by thread:
Re: Fitting a function to a list (newbie)
Next by thread:
Re: Fitting a function to a list (newbie)
|