[BioPython] Support Vector Machine

[Michael Sorich]

I have noticed that inputing the xs and ys to the train function of the SVM 
module as arrays instead of lists results in a memory leak. Has anyone else 
noticed this? I thing the problem is somewhere in the _F function of the 
Trainer class.

Has anyone done any work to optimising the training code? It seems a little 
slow. Could someone who understands the code comment on the easy of doing 
this (eg using numeric arrays wherever possible)?

Thanks

Michael




_________________________________________________________________
STOP MORE SPAM with the new MSN 8 and get 2 months FREE* 
http://join.msn.com/?page=features/junkmail