[pymvpa] help gnbsearchlight

[Patrik Andersson]

2011/2/10 Yaroslav Halchenko <debian at onerussian.com>

> Ha -- Thanks Patrik for chiming in, and sorry for not buzzing you
> immediately as soon as GNBSearchlight was fixed up in 0.6.
>
> No worries! You are quite active in the mailing list, must take a lot of
time.


> >    I have installed your 0.6 dev version to try it out. Though I from
> time to
> >    time get a bit perplex from the new parameter names etc, I have
> managed to
> >    get fairly familiar with it.
>
> sorry and thanks for digging into it ;)
>
> It seems to be a good way of being forced to learn Python.


> >    I now wanted to, just for fun, try the GNB searchlight, but fail.
> >    >...<
> >    What is wrong here? I guess the 'generator' is not whats expected(?)
> everything looks cool in your example... it runs for me without crashing
> (although a bit longer than I expected, had to wait nearly a few
> minutes) and results in
>
> <Dataset: 2x39912 at float64, <a: mapper>>
>
> So, the question is -- what is your mvpa version of 0.6? up to date
> master?  mine was
>
> $> git describe
> upstream/0.6.0.dev-111-gbe82724
> 0.60
>

In Synaptic it says 0.6.0~dev-69-g7afe4f7-1~nd60+1
both installed and latest. Is there another repository?

if it fails with an up-to-date version -- what is the failure message?
>
> /home/brain/Desktop/Data/tutorial_data/tutorial.py in <module>()
     16 gnbsl = sphere_gnbsearchlight(clf, gen)#, postproc=mean_sample())
     17
---> 18 res = gnbsl(ds)
     19
     20

/usr/lib/pymodules/python2.6/mvpa/base/learner.pyc in __call__(self, ds)
    220                                    "used and auto training is
disabled."
    221                                    % str(self))
--> 222         return super(Learner, self).__call__(ds)
    223
    224

/usr/lib/pymodules/python2.6/mvpa/base/node.pyc in __call__(self, ds)
     67         """
     68         self._precall(ds)
---> 69         result = self._call(ds)
     70         result = self._postcall(ds, result)
     71         return result

/usr/lib/pymodules/python2.6/mvpa/measures/searchlight.pyc in _call(self,
dataset)
    108
    109         # pass to subclass

--> 110         results, roi_sizes = self._sl_call(dataset, roi_ids, nproc)
    111
    112         if not roi_sizes is None:

/usr/lib/pymodules/python2.6/mvpa/measures/gnbsearchlight.py in
_sl_call(self, dataset, roi_ids, nproc)
    281         #       splitcfg just returns what to split into the other
in terms

    282         #       of chunks... and we need actual indicies

--> 283         if splitter.permute_attr is not None:
    284             raise NotImplementedError, \
    285                   "Splitters which permute targets aren't supported
here"

AttributeError: 'NFoldPartitioner' object has no attribute 'permute_attr'
WARNING: Failure executing file: <tutorial.py>

Thanks for helping out!





> --
>                                  .-.
> =------------------------------   /v\  ----------------------------=
> Keep in touch                    // \\     (yoh@|www.)onerussian.com
> Yaroslav Halchenko              /(   )\               ICQ#: 60653192
>                   Linux User    ^^-^^    [175555]
>
>
>
> _______________________________________________
> Pkg-ExpPsy-PyMVPA mailing list
> Pkg-ExpPsy-PyMVPA at lists.alioth.debian.org
> http://lists.alioth.debian.org/mailman/listinfo/pkg-exppsy-pymvpa
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.alioth.debian.org/pipermail/pkg-exppsy-pymvpa/attachments/20110210/f1e8aa6b/attachment.htm>