Re: Numerous Functional Forms
- To: mathgroup at christensen.cybernetics.net
- Subject: [mg1265] Re: Numerous Functional Forms
- From: jarle.brinchmann at astro.uio.no (Jarle Brinchmann)
- Date: Wed, 31 May 1995 01:55:36 -0400
- Organization: University of Oslo
In article <3q6d8b$854 at news0.cybernetics.net> wiestt at rl.af.mil (Todd E. Wiest) writes:
Does Mma. have any method to fit data to a list of functions if the exact
functional form is unknown?? I don't want to have to guess at all the
functional forms that may fit the data I have, and I have not located any
type of package that has a predetermined list of function forms to fit to.
You probably have to have some idea of what functions might be
suitable for fitting. Otherwise you will have to go around looking for
a packages starting like this:
(* :Name: TOE` *)
(* :Author: God *)
But seriously, unless you have a strong suspicion of how the data are
distributed and you want to get a better understanding of that you
should probably start plotting them with different types of axes. Do
also try to simply plot data on paper and sketch curves that seems to
fit - by hand. It should not be too difficult to guess at some
suitable functions to do the fit. And it will also enable you to
narrow down the search space when looking for a suitable package.
Jarle.
--
---------------------------------------------------------------------
Nuke the Whales ! | Jarle Brinchmann,
| Email: Jarle.Brinchmann at astro.uio.no
International Krill Union. | or : jarleb at astro.uio.no