[pymvpa] To compute the prediction accuracy under null distribution

cfygj cfygj at 126.com
Wed Aug 20 07:35:39 UTC 2014

Now I have some command lines to run MVPA, to compute the prediction error rate under null distribution.
clf = LinerCSVMC()
partitioner = NFoldPartitioner()
repeater = Repeater(count=10000)
permutator = AttributePermutator('targets', limit={'partitions': 1}, count=1)
null_cv = CrossValidation(clf, ChainNode( [partitioner, permutator], space=partitioner.get_space, postproc=mean_sample())
distr_est = MCNullDist(repeater, tail='left', measure=null_cv, enable_ca=['dist_samples'])
cv_mc_corr = CrossValidation(clf, partitioner, postproc=mean_sample(), null_dist=distr_est, enable_ca=['stats'])
error = cv_mc_corr(ds)
My Problems: the command lines above is to compute the prediciton error rate, but I want to compute the prediction accuracy. How to modify these command lines?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.alioth.debian.org/pipermail/pkg-exppsy-pymvpa/attachments/20140820/6fb3c1f6/attachment.html>

More information about the Pkg-ExpPsy-PyMVPA mailing list