[Eeglablist] ICA: number of components and channels

[John Fredy Ochoa Gómez .]

Hello Irene, the infomax included in the eeglab is a complete ICA
algorithm, that is, the number of "channels" correspond to the number of
ICs. I use "channels" instead of channels because is possible to reduce the
number of real channels using PCA, even the pop_runica detects if the rank
of your data is less than the number of channels and automatically reduces
the dimension of your data if is necessary

Exclude some channels might be problematic because the distribution of the
channels is related to the topography and in some cases to the clinical
standard montages like the 10-20 system, etc.

If your channels are noisy maybe is better recover them using interpolation
and later use the PCA algorithm

There are other algorithms like fastica that uses an deflationary approach
that enables the acquisition of a predetermined number of ICs

Best wishes,
John

On Thu, Mar 17, 2016 at 3:50 AM, Irene Sophia Mayer <mi_mayer at gmx.de> wrote:

> Dear EEGlabers,
>
> I have data with 64 channels and want to perform ICA. However, I would
> like to compute only 35 components and not 64 and also exclude certain
> channels from being considered without deleting them from the data.
>
> Is there a way to do that?
>
> Cheers,
> Irene
>
> _______________________________________________
> 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
>



-- 
John Ochoa
Docente de Bioingeniería
Universidad de Antioquia
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20160317/018d6a8b/attachment.html>