[Eeglablist] CORRMAP with missing channels

[Arnaud Delorme]

Dear Berry,

do you have to remove line 272 as well. This one seems important. Would you mind to test without removing Line 272 and let us know. This part of the code was added by Romain (copied) but I do not know the purpose of it.
Thanks,

Arno

On 20 Feb 2013, at 10:19, berry van den berg wrote:

> Dear all, 
> 
> I did some debugging based on Stefan's advice and it seems indeed that the channel indexes gets messed up. The channel location structure going into the eeg_interp is not ascending with respect to urchan index/ locs file.
> 
> I think the error lies in the CORRMAP.m function and probably around here: I dont really understand the purpose of line 251 through 272 but I assume to fix inconsistencies with respect to channelstructure. I would assume that the channel indexes/location across participants would be consistent with the .loc file and thereby, you should not have to check the channel indexes and can just feed in the .loc information into eeg_interp?.... Or you might want to check if for each participant the eeg.chanlocs.urchan match at the end of interpolation?
> 
> But in the end I commented out line 251 - 272   which seems to fix the weird looking IC's. and CORRMAP works fine now :)
> 
> Thanks!
> 
> Berry
> 
> 
> On 20 February 2013 13:03, Makoto Miyakoshi <mmiyakoshi at ucsd.edu> wrote:
> Dear Arno,
> 
> Here is a question for you asking about CORRMAP. When you have time please help him.
> 
> Makoto
> 
> 2013/2/18 berry van den berg <berryv.dberg at gmail.com>
> Dear all,
> 
> I am trying to use CORRMAP (eeglab12_0_1_0b) on a dataset to acquire a more data driven (and automatic) selection components that I want to remove as artifacts. Unfortunately I am having trouble if channels from certain participants are missing (bad channels removed before running ICA). If I am running CORRMAP on subjects without removed channels it works fine: 
> 
> https://www.dropbox.com/s/0y6718uezamllak/corrmap_works_well.png
> 
> However, if I add a subject with removed channels it looks a bit weird.... CORRMAP first interpolates the channels before it runs, and the selected ICs look completely different (also the template which should be the same template as when it works well). The "components" it selects also are not in the data.
> 
> https://www.dropbox.com/s/0kp0cd2exsk6apq/CORRMAP_didnt_work_well.png
> 
> Am I doing something wrong?
> 
> Thanks!
> 
> 
> -- 
> Berry van den Berg
> berryv.dberg at gmail.com 
> _______________________________________________
> 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
> 
> 
> 
> -- 
> Makoto Miyakoshi
> JSPS Postdoctral Fellow for Research Abroad
> Swartz Center for Computational Neuroscience
> Institute for Neural Computation, University of California San Diego
> 
> 
> 
> -- 
> Berry van den Berg
> berryv.dberg at gmail.com

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20130224/12a4e686/attachment.html>