[Eeglablist] Problem with pop_rejepoch

[Arnaud Delorme]

Dear Peter,

I think this issue arises because some of your time locking events do not have latencies that end up on sample data points. For instance, if you have a stimulus at latency 1034.67 ms, it might not end up on a data sample. However, after epoching (i.e. realigning the data to this event for instance), EEGLAB functions consider the stimulus to be at latency 0 and therefore realign the event with the closest sample EEG data point. As a result, latencies of other events in the epoch have to be slightly shifted to preserve the absolute relationship between all the events.

This is OK because the time precision of the EEG data is at most 1/2 of a sample point. Also, it does not make much sense to have a stimulus at time 0.3 ms and other at another wide range of latencies while they are all supposed to be at latency 0. This used to create all sorts of problems in earlier (<2005) versions of EEGLAB. However, if anybody has some comments or suggestions for better practice, do not hesitate to share it with us.

Best,

Arno

On Mar 28, 2011, at 5:19 AM, Peter Murphy wrote:

> Dear all,
> 
> I'm having an issue with the pop_rejepoch function and associated artefact rejection functions (pop_eegthresh, etc). Whenever I attempt to apply these to particular datasets as part of a script, they consistently appear to tamper with the latency information contained in the EEG.epoch.eventlatency substructure. Specifically, they seem to shift the latency of each trigger in each epoch by a very small amount of time (sub-millisecond). While this is a very small difference, it interferes substantially with my further attempts to automate the analysis of these data. Further, there's always the worry that some other aspects of the dataset, latency or otherwise, are being affected by use of these functions but I just havn't noticed it yet.
> 
> Has anyone come across a similar issue before, and possibly have any causes/solutions? Any advice would be much appreciated.
> 
> Thanks,
> 
> Peter Murphy
> 
> 
> _______________________________________________
> 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