[Eeglablist] ICA running slowly and returning error

Scott Makeig smakeig at gmail.com
Tue Feb 23 08:37:27 PST 2021

Anaeile -  This means that the problem dataset is rank-deficient. One
channel might be repeated, or be a sum of two or more other channels, etc.
Use the 'pca' flag to reduce the rank by one when running runica. Notice on
the Matlab command line if the step size is decreasing as normal. If not,
try reducing the dimension by 2, etc.

Scott Makeig

On Tue, Feb 23, 2021 at 11:34 AM Anaelle Alouit <anaelle.alouit at gmail.com>

> Dear all,
> I have ran ICA on a few datasets of 64 channels without any problem, but
> for one dataset, the ICA decomposition takes hours (an entire weekend
> instead of 1-2 hours), and in the end, I can’t use the ICLabel tool because
> it is returning an error saying “the ica decomposition must be real”.
> I tried cleaning the data with the CleanData/ASR plugin, but it didn’t
> help. I downsample to 256Hz and high-pass filter at 1Hz.
> Since I have the same type of dataset and exactly the same kind of
> pre-processing, do you know what may cause this ?
> Best regards,
> Anaëlle ALOUIT
> _______________________________________________
> 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

Scott Makeig, Research Scientist and Director, Swartz Center for
Computational Neuroscience, Institute for Neural Computation, University of
California San Diego, La Jolla CA 92093-0559, http://sccn.ucsd.edu/~scott

More information about the eeglablist mailing list