<p>That's a great idea (on both fronts). I tried running it through an SvdMapper last night and it failed to converge. </p>
<p>Isn't there an IcaMapper in pymvpa (I'm currently not at a computer to check) or do you suggest trying the fastica in mdp? </p>
<p>Thanks,<br>
P</p>
<div class="gmail_quote">On Mar 31, 2011 7:34 AM, "Yaroslav Halchenko" <<a href="mailto:debian@onerussian.com">debian@onerussian.com</a>> wrote:<br type="attribution">> talk to a machine learning guy and have a good laugh together ;)<br>
> <br>> try running spatial ICA/SVD and try classifying among those 20 modes<br>> <br>> On Wed, 30 Mar 2011, Per B. Sederberg wrote:<br>> <br>>> On a related note, any thoughts on how to tame a dataset where I have<br>
>> 20 samples and 3 million features?<br>> <br>> -- <br>> =------------------------------------------------------------------=<br>> Keep in touch <a href="http://www.onerussian.com">www.onerussian.com</a><br>
> Yaroslav Halchenko <a href="http://www.ohloh.net/accounts/yarikoptic">www.ohloh.net/accounts/yarikoptic</a><br>> <br>> _______________________________________________<br>> Pkg-ExpPsy-PyMVPA mailing list<br>
> <a href="mailto:Pkg-ExpPsy-PyMVPA@lists.alioth.debian.org">Pkg-ExpPsy-PyMVPA@lists.alioth.debian.org</a><br>> <a href="http://lists.alioth.debian.org/mailman/listinfo/pkg-exppsy-pymvpa">http://lists.alioth.debian.org/mailman/listinfo/pkg-exppsy-pymvpa</a><br>
</div>