Re: C Code required
- From: Roman Werpachowski <"r o m a nNOSPAM"@theta1.cft.edu.pl>
- Date: Sat, 2 Apr 2005 19:08:50 +0000 (UTC)
On the 2 Apr 2005 10:27:36 -0800, beliavsky@xxxxxxx wrote:
> Stephen Clark wrote:
>> Can someone recommend some C source code for matrix operations
>> (multiplication, inverse) and one/two dimensional optimisation.
> Preferably
>> freeware. I am looking at the Numerical Recipes code but I've been
> told that
>> isn't necessarily "stable". Thanks.
>
> I would look in the GNU Scientific Library
> http://www.gnu.org/software/gsl/ , but note the licensing conditions
> listed there. Optimization codes usually come in two variates,
> univariate and multivariate (which can handle 1 or more).
I advise against gsl. It is overloaded with special types for even the
simplest objects (like gsl_vector when in most cases a simple double *
would do), special interfaces (you think you need to reference an array
to read the 2nd coordinate of the vector? wrong! you need to call a
macro to do this). To top all this, it still has weird errors. I coded a
simple multidimensional root-finding program, using the docs as a basis,
and I receive an error, although everything is correct and almost
identical to the documentation example.
--
Roman Werpachowski
/--------==============--------\
| http://www.cft.edu.pl/~roman |
\--------==============--------/
.
- Follow-Ups:
- Re: C Code required
- From: BEC Charlie
- Re: C Code required
- References:
- C Code required
- From: Stephen Clark
- Re: C Code required
- From: beliavsky
- C Code required
- Prev by Date: Re: time-series smoothing
- Next by Date: Laplace Transform of an integral
- Previous by thread: Re: C Code required
- Next by thread: Re: C Code required
- Index(es):
Relevant Pages
|