[pymvpa] extracting selected feature IDs using a Split Classifier

[Yaroslav Halchenko]

every classifier has a state variable feature_ids (disabled by default)
which keeps ids of the features which classifier was trained to use.
and unless classifier does some feature selection (e.g. SMLR,
FeatureSelectionClassifier), then it is a range(dataset.nfeatures).

so for split classifier feature_ids is a union of feature_ids used by
each classifier (ie for each split). if you need to know feature_ids of
a classifier for each split, just enable that state variable for a slave
classifier which you provided to SplitClassifier's constructor (so
called slave classifier), and then look into

splitclf.clfs[?].feature_ids
(where ? is an index of a classifier)

if I am not clear anywhere -- don't hesitate to ask for clarification
;-)

On Wed, 23 Jul 2008, James M. Hughes wrote:

> Hi all,

> The subject line pretty much sums it up; I'm curious what the  
> convenient way is to actually get the list of selected features after  
> performing RFE using a SplitClassifier with an embedded  
> FeatureSelectionClassifier (like the example given in warehouse.py).

> Thanks,
> James.

> _______________________________________________
> Pkg-ExpPsy-PyMVPA mailing list
> Pkg-ExpPsy-PyMVPA at lists.alioth.debian.org
> http://lists.alioth.debian.org/mailman/listinfo/pkg-exppsy-pymvpa


-- 
Yaroslav Halchenko
Research Assistant, Psychology Department, Rutgers-Newark
Student  Ph.D. @ CS Dept. NJIT
Office: (973) 353-5440x263 | FWD: 82823 | Fax: (973) 353-1171
        101 Warren Str, Smith Hall, Rm 4-105, Newark NJ 07102
WWW:     http://www.linkedin.com/in/yarik