[Eeglablist] variable times running ICA

Makoto Miyakoshi mmiyakoshi at ucsd.edu
Thu Dec 25 18:14:28 PST 2014

Dear Baker,

I know that runica() speed variess depending on data quality (AMICA does
not do this). Watch the runica log when you start it and if it says
'lowering learning rate' or something like that, it is 'having hard time
figuring out how to proceed'. If your data is 1) full-rank and 2) clean
(i.e. stationary), it goes fast.

For data rank, see below.


On Tue, Dec 16, 2014 at 7:42 AM, Baker, Joshua <joshua.baker at ntu.ac.uk>

> Dear EEGlablist recipients.
> Could somebody please explain to me the large variation in the time it
> takes to do ICA on datasets of roughly the same size? I am using 64 channel
> Biosemi data at 1024 Hz sampling rate filtered from 0.1 to 500Hz. All
> datasets in this particular experiment are processed using the exact same
> parameters and have the exact same epoch length (1 epoch of 600 seconds).
> Some channels have been omitted from some of the datasets. Some datasets
> take around 20 minutes to run ICA, others take 4 hours and above. What
> could be the reason for this?
> Many thanks,
> Joshua.
> DISCLAIMER: This email is intended solely for the addressee. It may
> contain private and confidential information. If you are not the intended
> addressee, please take no action based on it nor show a copy to anyone. In
> this case, please reply to this email to highlight the error. Opinions and
> information in this email that do not relate to the official business of
> Nottingham Trent University shall be understood as neither given nor
> endorsed by the University. Nottingham Trent University has taken steps to
> ensure that this email and any attachments are virus-free, but we do advise
> that the recipient should check that the email and its attachments are
> actually virus free. This is in keeping with good computing practice.
> _______________________________________________
> 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

Makoto Miyakoshi
Swartz Center for Computational Neuroscience
Institute for Neural Computation, University of California San Diego
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20141225/baaeb867/attachment.html>

More information about the eeglablist mailing list