[pymvpa] Regression estimates?

[Per B. Sederberg]

On a related note, any thoughts on how to tame a dataset where I have
20 samples and 3 million features?

P

On Wed, Mar 30, 2011 at 11:57 PM, Per B. Sederberg <psederberg at gmail.com> wrote:
> On Wed, Mar 30, 2011 at 11:44 PM, Yaroslav Halchenko
> <debian at onerussian.com> wrote:
>>
>> On Wed, 30 Mar 2011, Per B. Sederberg wrote:
>>> clfr = GLMNET_R(alpha=.5, model_type='naive', enable_ca=['estimates'])
>>
>> so it is a regression.  Wrap with RegressionAsClassifier to gain
>> classification ;)
>>
>> thus metrics for regressions:
>>
>>> -------> print(sclf.ca.stats)
>>> Statistics  Mean  Std   Min      Max
>>> ---------- ----- ----- -----    -----
>>> Data:
>>>    RMP_t   2.285 0.696  1.3       4
>>>    STD_t     0     0     0        0
>>>    >...<
>>
>
> I actually want the regression and not the classifier, but I now
> totally understand the stats printout.  I was confused because the CCe
> is 1 minus the correlation.
>
>>> Also, I really just want to get the predictions/estimates of the
>>> classifier from each fold, but I don't know how to get that out.
>>
>> those, since you have stats, are also available within
>> sclf.ca.stats.sets
>>
>
> Awesome!  Yes, I had just found that and the exact keys I needed were:
>
> cv.ca.stats.stats['RMP_p_all']
> cv.ca.stats.stats['RMP_t_all']
>
> Now I think I'm all set!
>
> Thanks,
> P
>
>
>> --
>> =------------------------------------------------------------------=
>> Keep in touch                                     www.onerussian.com
>> Yaroslav Halchenko                 www.ohloh.net/accounts/yarikoptic
>>
>> _______________________________________________
>> Pkg-ExpPsy-PyMVPA mailing list
>> Pkg-ExpPsy-PyMVPA at lists.alioth.debian.org
>> http://lists.alioth.debian.org/mailman/listinfo/pkg-exppsy-pymvpa
>>
>