[pymvpa] Individual measures for subjects

[Yaroslav Halchenko]

On Fri, 31 Jan 2014, Arman Eshaghi wrote:

>    Dear all,
>    I have the following�snippet� and�according�to documentations I can get
>    different statistics from 'results' object.�
>    However, I was wondering how I can get individual (subject by subject)


>    labels, or in other word how I can see the predictions for each subject?�
>    Hopefully I will add it somewhere in�documentation�for posterity.
>    All the best,
>    Arman
>    � �baseclf= LinearNuSVMC(nu=0.3)
>    � �metaclf=MappedClassifier(baseclf, SVDMapper())
>    � �cvte=CrossValidation(metaclf, HalfPartitioner(attr='chunks'),
>    errorfx=lambda p, t: np.mean(p == t), � � �enable_ca=['stats'])
>    � �results=cvte(MyData)
>    #cvte.predictions() does not work

and your MyData is the one containing the subjects or where do they
come? (not sure since you do partitioning still based on chunks
attribute not e.g. based on 'subjects' ;-) )

-- 
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