This is the mail archive of the
gsl-discuss@sources.redhat.com
mailing list for the GSL project.
Re: forwarded message from Sergey Kolos
- From: Fabian Bastin <fabian dot bastin at math dot fundp dot ac dot be>
- To: gsl-discuss at sources dot redhat dot com
- Date: Tue, 16 Mar 2004 11:47:52 +0100
- Subject: Re: forwarded message from Sergey Kolos
- References: <OFA1121013.1305430C-ON80256E55.00492AB6@db.com>
Hi,
Just some suggestions, but not necessarily the best ones. Instead of
rewritten routine for nonlinear optimization, could it be possible to
link GSL and Opt++ in some way? I think that Opt++ is a very nice
library, published under LGPL, for nonlinear nonconvex optimization:
http://acts.nersc.gov/opt++/ However, it is written in C++ instead of C.
You could also look at the package IPOPT, maintained by Andreas Wachter
(IBM, New York):
http://www-124.ibm.com/developerworks/opensource/coin/Ipopt/
IPOPT is published under the Common Public Licence, and is very efficient.
Regards,
Fabian
Przemyslaw Sliwa wrote:
Hi,
Excellent!
This is one of the main issues missing in gsl.
But which algorithms did you use fur the local minimization?
See: http://sources.redhat.com/ml/gsl-discuss/2003-q3/msg00173.html
Fab promised to fix the bfgs code, but is still inefficient.
It could be nice to have a comparison with another routines.
I could do this for NAG > they have an excellent module for minimization with nonlinear constrains.
Look forward to hearing from you.
Regards,
Przemyslaw
--
This e-mail may contain confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and destroy this e-mail. Any unauthorized copying, disclosure or distribution of the material in this e-mail is strictly forbidden.
--
Dr Fabian Bastin | Tel: +32 (0)81 724923
Aspirant FNRS, Dpt Mathématiques | Fax: +32 (0)81 724914
FUNDP Namur, Belgium | email: fbas@math.fundp.ac.be
http://www.fundp.ac.be/~fbastin | JabberID: slashbin@jabber.org