Subject: Re: Curve fitting
From: RODA (jean-marc.roda@cirad.fr)
Date: mar fév 19 2002 - 10:23:24 CET
Do you know the following softwares ?
=> Tablecurve 2D or Tablecurve 3D
=> Statistica
They are really convenient for curve fitting. Tablecurve is more adequate for
an exploratory work, and Statistica is the best tool after SAS (almost the
same power, and fare more easy to use).
But wat is exactly your need ?
- Do you need to process separately but automaticly the same regression
equation for a great number of data sets, in order to get different
regression parameters for each set (which is called local regression, or
local modelling) ?
- Do you need to apply simultaneously the same regression to a great number
of data sets (which is called global regression, or global modelling) ?
- Do you need to test several regression equations for the same data set ?
- or something else ?
Depending your exact need, the different existing tools are more or less
adequate...
all the best,
Jean-Marc
Louise Erasmus a écrit :
> Hi everybody
>
> I need to do a number of regression analyses (curve fitting), but I don't
> want to do a heuristic search for each analysis. Can any one think of any
> short cuts?
>
> Best wishes
>
> Louise
>
> Centre for Environmental Studies / Department of Zoology and Entomology
> Geography Building 2-1
> University of Pretoria
> Pretoria
> 0002
>
> Tel. +27-12-420 4048
> Fax. +27-12-420 3210
>
> http://www.up.ac.za/academic/centre-environmental-studies/
>
> "Criticism comes easier than craftsmanship."
>
> -Zeuxis
-- Jean-Marc Roda Economie des produits forestiers // Forest products economics Programme Bois // Forest Products Programme CIRAD-Forêt - TA 10/16 - 73 rue J-F Breton 34398 Montpellier Cedex 5 - France Tel +33 (0)4 67 61 44 99 - Fax + 33 (0)4 67 61 57 25 e-mail jean-marc.roda@cirad.fr web http://www.cirad.fr/activites/bois/index.html http://www.cirad.fr/