[pymvpa] top-n match lists
tg at cs.york.ac.uk
Thu Jan 7 17:32:51 UTC 2010
In the classification problem I have, it is useful to be able to find the top 10
(or top-n) best matches to a single sample, i.e. a ranked list of the closest
classes rather than just the top-1/closest class label.
At the moment Classifier.predict() does not appear to be able to do this and I
have not been able to find this functionality in the pymvpa classifier base
class. Is it elsewhere, or is there any other way to extract this information?
More information about the Pkg-ExpPsy-PyMVPA