[Eeglablist] storing and processing averaged data in EEGLAB

[Arnaud Delorme]

Darren,

the best way to process ERPs under EEGLAB is to create studies. This way 
the ERPs are saved under separate files and loaded on the fly. They are 
also stored in the STUDY structure under

STUDY.changrp(X).erpdata
STUDY.changrp(X).erptimes

X represents the index of the channel (the structure also contains the 
channel label). This can contain ERP for an arbitrary large number of 
subjects (since the raw data is not in memory - only the ERP data is 
present there - there is little memory limitation). Also, this structure 
"erpdata" usually contains a cell array of ERP for different 
conditions/groups. For more information, see the different fields of the 
STUDY super-structure below

http://www.sccn.ucsd.edu/eeglab/clusttut/clustertut.html#studyset

You should try creating a simple study with a couple of subjects, 
precompute the ERPs (using menu item "STUDY > Precompute data measure"), 
plot them (so they are read in memory) and look at this structure. The 
plotting option also allows you to compute basic parametric and 
permutation statistics on ERPs.

Note however, that you should not expect EEGLAB to be ideal for 
processing ERP data the "standard ERP way". It was not designed in this 
spirit.

Hope this helps,

Arno

ps: the pop_comperp() function also computes ERP for different 
conditions but it is obsolete and will not be used in the future.

Darren Weber wrote:
>
> What is the best way to calculate and process ERPs in EEGLAB?
>
> If I want to convert an epoch dataset to an ERP dataset, should I 
> simply change EEG.data to an Nchan x Nsamp matrix?  For example:
>
> EEG.data = mean(EEG.data, 3);
>
> Can we store the variance somewhere?  Do we have a data structure 
> separate from EEG.data, maybe EEG.avg that contains the mean, var, 
> Ntrials or Nepochs in the avg, etc?  What is the best way to combine 
> ERP data across subjects?
>
> Best, Darren
>
> PS, I can do all this with custom scripts, but I want to know if there 
> is a specific EEGLAB way of doing all this.
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> 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/20080307/dbacc044/attachment.html>