I’ve just released Math::NLopt, a Perl interface to NLopt <https://github.com/stevengj/nlopt>, which is a
library for nonlinear local and global optimization, for functions with and without gradient information. It is designed as a simple, unified interface and packaging of several free/open-source nonlinear optimization libraries.
The interface uses plain Perl arrays to pass parameters[1], but the objective functions can use PDL if they want to. Eventually there will be support for PDL ndarrays for parameters, which may ease up on some of the data copying between Perl and C (which is pretty minimal, as data are tiny and the transfer buffers cached, so this is most likely dwarfed by the compute time).
Cheers,
Diab
[1] I find the tools to generate an XS interface to a library better for libraries with a large API.
_______________________________________________ pdl-general mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/pdl-general
