[pymvpa] PyMVPA Question

[Yaroslav Halchenko]

On Tue, 29 Oct 2013, Taku Ito wrote:

>    Awesome, thanks so much. So far so good... Hopefully it turns out ok!

sure -- just bake it a bit longer for better flavor ;)

>     Just out of curiosity, how does it choose the number of attributes to
>    balance?  For each fold, is it a random number of balanced target labels
>    each fold?  Or is it a fixed number (e.g., the maximum number of labels
>    such that the number of target labels remains equal)?

I think it is the latter...  I would encourage you to verify what such a
ChainNode would actually do to your dataset before plugging it into
CrossValidation ;)

-- 
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