[pymvpa] how to train/test on a single partition?

Yaroslav Halchenko debian at onerussian.com
Mon Jul 15 00:15:32 UTC 2013

On Sun, 14 Jul 2013, Nick Oosterhof wrote:

> Hello,

> I'm trying to do MVPA where training and testing are done one
> different parts of a dataset using a partitioner and a sifter. I
> define the test chunk manually in a for loop (for reasons not
> important for the present question), and given the value for test_run,
> the partitioner is defined by:

> sifter=Sifter([('partitions', 2),
>              ('chunks', [test_run])])

> par=ChainNode([NFoldPartitioner(attr='chunks'),
>                                            sifter])

brr -- are you just need a single cross-validation estimate from some
specific all-but-X chunks to the left out X chunk?

then use CustomPartitioner

and specify splitting you like

if X doesn't have to be specific, then use "count=1" for whatever stock
partitioner (NFoldPartitioner) you like to use.

or am I missing smth?

Yaroslav O. Halchenko, Ph.D.
http://neuro.debian.net http://www.pymvpa.org http://www.fail2ban.org
Senior Research Associate,     Psychological and Brain Sciences Dept.
Dartmouth College, 419 Moore Hall, Hinman Box 6207, Hanover, NH 03755
Phone: +1 (603) 646-9834                       Fax: +1 (603) 646-1419
WWW:   http://www.linkedin.com/in/yarik        

More information about the Pkg-ExpPsy-PyMVPA mailing list