hi,
can you post a code snippet in a gist to reproduce the problem?
Alex
On Wed, Jun 27, 2012 at 11:44 PM, Igor Nikolskiy wrote:
> Hi Andy,
>
> Thanks for the fast response!
>>
>> Could you please say which version of sklearn you are using?
>
> I am using sklearn version 0.11
>
>> About the na
Hi Andy,
Thanks for the fast response!
> Could you please say which version of sklearn you are using?
>
I am using sklearn version 0.11
About the nans, I have no idea.
>
Yep the nans is where the problem is.
> Have you tried to not do your own whitening, i.e. whiten=True?
>
Yes and in that ca
Hi Igor.
Could you please say which version of sklearn you are using?
There was a minor change recently to improve consistency.
Now, the independent components of size [n_components, n_features are in
ICA.components_.
ICA.sources_ should contain the [n_samples, n_components] estimated
source sig
Hello all,
I've got a FastICA question. I am interested in doing my own whitening of
the data before feeding it to fastICA, but I'm running into trouble. Here's
the code:
from pylab import *
from pandas import *
from sklearn.decomposition import FastICA, PCA
ms2= read_csv('L-Glu.csv', index_col=
