[Eeglablist] ICA problema Massimiliano Papera

[Massimiliano Papera]

Dear all,

I am just about to run the ICA and I just incurred in this statement in the
tutorial:

'Note: If you try to run ICA on this new dataset, the number of dimensions
of the data will have been reduced by the number of components subtracted.
We suggest again 'baseline-zeroing' the data (if it is epoched when some
components have been removed, data epoch-baseline means may change). To run
ICA on the reduced dataset, use the *pca* option under the Tools > Run
ICApop-up window, type
* 'pca', '10' * in the Commandline options box to reduce the data
dimensions to the number of remaining components (here 10), before running
ICA (see runica.m <http://sccn.ucsd.edu/eeglab/locatefile.php?file=runica.m>.
If the amount of data has not changed, ICA will typically return the same
(remaining) independent components -- which were, after all, already found
to be maximally independent for these data.'

this is from
http://sccn.ucsd.edu/wiki/Chapter_09:_Decomposing_Data_Using_ICA

Let's us suppose that I have just removed, say, 4 ICS comps. Then how do I
check if the amount of data has changed?

Thank you very much for your help.


Cheers
Massimiliano Papera
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20121010/a783ba8d/attachment.html>