[Eeglablist] ICA related fields on the EEG struct

Arnaud Delorme arno at ucsd.edu
Thu Nov 27 23:25:05 PST 2014

To make it short: yes you should reshape the matrix to make it 3-D. Note that, in your code, you should also select only these channels which are used for ICA

EEG.icaact = (EEG.icaweights*EEG.icasphere)*reshape(EEG.data(EEG.icachansind,:), EEG.nbchan, EEG.trials*EEG.pnts);
EEG.icaact = reshape(EEG.icaact, size(EEG.icaact,1), EEG.trials, EEG.pnts);


On Nov 17, 2014, at 12:06 PM, Yamil Vidal Dos Santos <hvidaldossantos at gmail.com> wrote:

> Dear all,
> I would like to know what are each ICA realted fields on the EEG struct (such as icaact, icaweights, icasphere). I remember finding this in a tutorial a long time ago, but now I cannot find it.
> In particular I would like to know this. The field EEG.icaact is usually empty and can be filled by using the function icaact.m or by running something like the following line:
> EEG.icaact = (EEG.icaweights*EEG.icasphere)*reshape(EEG.data, EEG.nbchan, EEG.trials*EEG.pnts);
> But the resulting matrix is 2 dimensional (with all the trials concatenated). Is this correct? Or should I reshape it to have dimensions component x time x trial.
> I'm asking these because I'm trying to use an automatic classifier to mark components as artifactual. The paper presenting the classifier is this one: http://www.behavioralandbrainfunctions.com/content/7/1/30
> Does anyone has experience using this classifier?
> Any alternative methods suggested?
> Thanks,
> Yamil Vidal Dos Santos
> _______________________________________________
> 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...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20141127/e260b6f3/attachment.html>

More information about the eeglablist mailing list