[Eeglablist] ICA problem
pwang.list at googlemail.com
Tue Feb 2 00:14:51 PST 2010
Thank you for your discussion. However, I did not see a direct answer to
my question. My problem is, when the number of component is smaller than the
number of channels, something wrong would happen. It seems that this problem
appears after computing ICA. I would appreciate to to know how to solve this
The ICA would converge in my situation. For Joe's question on how long,
it's about one night for one dataset (122 channels, 240 trials, 3000 data
points for each), if I take the same number of IC as the channels. I have
288 datasets, so it is too long to be acceptable. Probably, I can downsample
the dataset to reduce the time for computing ICA, but then I still need to
pick the noise components (eg. blinks) out of ~120 ones, this manual work
also takes a lot of time. Thus I wish I can have less ICs.
btw: I am using matlab R2009a on Vista SP1, both 64-bit.
On Tue, Feb 2, 2010 at 12:56 AM, Arnaud Delorme <arno at ucsd.edu> wrote:
> > Arno -- There was a problem with the Matlab rank() function on 64-
> > bit machines, I believe. Has this been solved and Is the auto rank
> > detection -> PCA option currently implemented in runica/binica?
> > Perhaps we could add a 'toy' rank() function pre-test (e.g. finding
> > the rank of a small full-rank matrix to detect if rank() is
> > working...) ? If so, run the rank test; if not, then warn the user
> > or build a work-around rank function that will work properly?
> This is correct. This was a problem with Matlab 64-bit in 2007. One of
> the EEGLAB user, Sven Hoffman, implemented an alternate computation of
> the rank which we used whenever the Matlab rank is deficient. However,
> the new Matlab rank is fine now.
> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
> To unsubscribe, send an empty email to
> eeglablist-unsubscribe at sccn.ucsd.edu
> For digest mode, send an email with the subject "set digest mime" to
> eeglablist-request at sccn.ucsd.edu
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the eeglablist