[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Help-gsl] non linear fitting with constraints

From: Hansen, T.A.R.
Subject: [Help-gsl] non linear fitting with constraints
Date: Sun, 24 Aug 2008 20:38:46 +0200

Dear gsl users,

I am new to this list and to gsl, so pardon me this newbie question.

I want to fit a number (user-defined) of gauss functions to a dataset. The user 
(mostly me) will specify start values for the parameters, as well as 
constraints on these parameters.
For fitting non linear functions, I can use the functions as described in the 
manual "37.9" (e.g. gsl_multifit_fdfsolver, gsl_multifit_fdfsolver_iterate, 
expb_df, expb_f...) However, these functions do not deal with constraints. 
Looking around with google I found an old post "least-squares fitting or 
minimization with constrained parameters" from dec. 2005 explaining that one 
way is add a penalty value when the new parameters fall outside the constraints.

Unfortunately, I am uncertain as to which functions to adjust or which values 
to change, or how to reconfigure the solver. Mathematical fitting algorithms 
are not my area of expertise.
Although I know obj-c, with plain c I am not very familiar.

With regards,

reply via email to

[Prev in Thread] Current Thread [Next in Thread]