In my MsC I used very large sparse matrices. The software was implemented in C, and it gave me very nice headaches.
Now, I'm preparing my PhD based on the same area (Natural Language Processing) and I'm thinking if I can gain anything with Perl Data Language. Anybody using PDL (well, I know Gimp uses it) can give me examples of where it is beeing used?
It has special toolboxes (C optimized) for sparse matrices, by the way. Oh, and a special toolboox for NLP as well.
Re:if you're dealing with matrices
ambs on 2004-07-12T11:17:32
Matlab has a big problem: license:-| Re:if you're dealing with matrices
spur on 2004-07-12T16:25:44
Uni's usually have licenses. Besides, Matlab's licenses for students are very cheap. The cheapest licenses can be found on Emule:-) You can use Octave then...
Re:if you're dealing with matrices
ambs on 2004-07-12T16:47:10
Yes. It is true that the University has the license. I just wanted to prepare a toolkit to anyone to download and use, freely. I'll look at Octave and see what I can gain with him. Thanks.Octave!
educated_foo on 2004-07-13T05:04:23
I highly recommend octave. Make sure to get the latest version (2.1.57), and also to get the "octave-forge" tools off sourceforge, which have a sparse matrix implementation and some Matlab compatibility routines. Octave diverges from Matlab mostly in not supporting the latter's unpleasant C interface (mex) and even more repugnant "object orientation".
Re:Octave
spur on 2004-07-12T16:24:17
Octave is a good idea for a free alternative to Matlab. But I highly doubt that Inline::Octave will provide the performance he needs.
