[Eeglablist] Data rank reduction before ICA

[Gian Marco Duma]

Dear EEGlab community, I have a problem. I used to interpolate channels
before doing ICA, and it always worked well. I have 128 channels, and even
if I interpolated the channels before ICA, I have always obtained 128
components. I downloaded the Prep-pipeline to clean the data before ICA,
just to try an automatized cleaning method, than I deleted the plug-in. But
now I have a problem, if I interpolate a channel before ICA, the rank of my
matrix is reduced of a number equal to the number of channels that I have
interpolated. For example I have 128 channels, now if I interpolate 5
channels the rank of my matrix becomes 123 and I obtain 123 components
instead of 128. This has never happened before. I tried with a pc in wich
Prep-pipelin plug-in has never been downloaded, and it works well, without
reducing the number of components. So I think that Prep-pipeline toolbox
changed some function that I don't know. Have you some suggestion to solve
this problem?
thank you
Gian Marco Duma
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20170421/004c6a0e/attachment.html>