[pymvpa] using Balancer()

[Edmund Chong]

Ah, works perfectly. So it should be:

datasets = list(partitioner.generate(dataset))
then
print datasets[n].summary(chunks_attr='partitions')

I took a look and indeed I end up with individual test runs with empty
conditions, the same way Tren pointed out in the previous message. But
Balancer does work as advertised for the training set, because I don't have
empty conditions when multiple runs are summed together...

Thanks!
-Edmund

On Thu, Jul 26, 2012 at 2:40 PM, Michael Hanke <mih at debian.org> wrote:

> On Thu, Jul 26, 2012 at 02:16:50PM -0400, Edmund Chong wrote:
> > foo1 = partitioner.generate(dataset)
>
> You could to
>
>   datasets = list(partitioner.generate(dataset))
>
> to get them all at once.
> >
> > print foo2.summary()
> >
> > ...I tried that, but it gives me a summary table of the number of samples
> > per condition per chunk. How do I view the samples per condition per
> > partition (classification fold)?
>
> That a look at the datasets. They have a sample attribute 'partition'
> (or something like that). Now look at the help for Dataset.summary()
> (just do ``ds.summary?`` in IPython). That will list a 'chunks_attr'
> arguments that you can set to be the partitions argument. Which in turn
> will cause the summary to be per-partition.
>
> Hope that helps,
>
> Michael
>
>
> --
> Michael Hanke
> http://mih.voxindeserto.de
>
> _______________________________________________
> Pkg-ExpPsy-PyMVPA mailing list
> Pkg-ExpPsy-PyMVPA at lists.alioth.debian.org
> http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/pkg-exppsy-pymvpa
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.alioth.debian.org/pipermail/pkg-exppsy-pymvpa/attachments/20120726/fdff682b/attachment.html>