[Eeglablist] Combining datasets, with ICA weights, by stacking NOT concatenating

[Makoto Miyakoshi]

Dear James,

My colleague John Iversen did something similar/related. Maybe you can find
the exact method in this paper.
http://sccn.ucsd.edu/~scott/pdf/IversenMakeig_MEG_Chapter13lo.pdf

Makoto

On Tue, Jan 5, 2016 at 9:46 AM, James Jones-Rounds <jj324 at cornell.edu>
wrote:

> Hello all,
>
> I have a question about stacking datasets that have ICA weights. It might
> seem a bit blasphemous. I want to use SIFT (in component space) to assess
> connectivity between two different individuals' brains, who performed a
> speaking-listening task. Because they were communicating, their datasets
> share the same time-course.
>
> Can I just stack the ICA weights of dataset B, below dataset A? For
> example, if the EEG.icaact field for dataset A is 13 x 10000, and if that's
> the same for dataset B, can I just assign dataset B's EEG.icaact field to
> rows 14:26, in a new "frankenstein" EEG.icaact field that has dimensions
> [26 x 10000] ?
>
> Besides some obvious issues resulting from ICA being run on two separate
> datasets, are there any basic matrix-math issues caused by doing this?
>
> Finally, how would I stack the EEG.icasphere, EEG.icaweights, and
> EEG.icawinv fields? Do I even need to if the EEG.icaact field is already
> computed separately for each dataset?
>
> Thank you for your help!!
>
> James
>
> --
> James Jones-Rounds
> Laboratory Manager
> Human Development EEG and Psychophysiology (HEP) Laboratory,
> Department of Human Development,
> --------------------------------------------
> Cornell University | Ithaca, NY
> 607-255-9883
> eeg at cornell.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/20160108/2badfa64/attachment.html>