[pymvpa] Selecting samples from a dataset

[James M. Hughes]

No, I haven't observed different behavior, just a surprising result ;)

I checked a few data points by hand (just by asking numpy to tell me  
if the values at all the indices matched), but this was for only one  
or two data labels and a random sample of datapoints from those  
labels; I end up with 240 in my experiment.

Thanks!

James.
On Apr 29, 2010, at 19:15 , Yaroslav Halchenko wrote:

> sorry about not-clear documentation -- samples order should be
> preserved.  have you observed different behavior?
>
> N.B. we are talking about 0.4.x functionality ;)
>
> On Thu, 29 Apr 2010, James M. Hughes wrote:
>
>> This is a silly question from someone who should know the answer
>> already, but the documentation is pretty unclear on this point:
>
>> if we select samples from a dataset according to labels,
>
>> e.g.,
>
>> samples = data.select(labels=[2,4,6]),
>
>> will the returned samples be grouped by label?
>
>> Also, will they be returned in groups in the order of label that
>> they were given?  Or are they returned in sorted order of increasing
>> label?
>
>> Thanks,
>> James.
> -- 
>                                  .-.
> =------------------------------   /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