This is the mail archive of the
gsl-discuss@sources.redhat.com
mailing list for the GSL project.
Re: multifit
- To: mquinson at ens-lyon dot fr
- Subject: Re: multifit
- From: Brian Gough <bjg at network-theory dot co dot uk>
- Date: Fri, 13 Oct 2000 12:00:04 +0100
- CC: gsl-discuss at sourceware dot cygnus dot com
- References: <Pine.GSO.4.21.0010121809370.22481-100000@info>
- Reply-To: gsl-discuss at sourceware dot cygnus dot com
It is possible to do fitting for multiple parameters and data
dimensions -- see the chapters "Fitting" and "Nonlinear Fitting" in
the manual for details.
For linear models the function is gsl_multifit_linear, described there.
For nonlinear models there is an example program at the end of the
"Nonlinear Fitting" chapter, which can be generalized fairly easily.
If this doesn't help let us know the specific model/data that you want
to fit and I will be more specific.
regards
Brian Gough
Date: Fri, 13 Oct 2000 11:48:32 +0200 (MET DST)
From: Martin Quinson <mquinson@ens-lyon.fr>
Hello, I've checkout'ed the cvs tree to see the new fiting
functionalities, and I couldn't find how to fit functions of 3 variables,
ie things like f(x,y,z)=t where all x,y,z,t are scalars.
Is it possible ? What is the concerned function ?