[Eeglablist] about the epoch code

[Tarik S Bel-Bahar]

Hello Edward, The pipeline is a good recommendation but it does not cover
all possible cases. It just a rough map. One needs to use it smartly and
know why one is doing each step. That being said...

What's recommended in that pipeline is -1 second to +2 seconds around an
event at 0 time.
If you don't have a event-related paradigm to which you're locking your
epochs, and if your data is continuous,
then you just want to epoch your data as you wish (into 2 or 3 second
chunks).

Note that ICA does not care if your data is epoched or not, it mixes up all
time points for it's purposes.

Note that some people epoch continuous data in order to remove the worst
periods.
So for example, you can run something like the pop_rejcont function, which
will remove big bad periods from the continuous data, and then epoch using
reg_epoch, and then drop the worst epochs, and then ICA.
One can also skip the continuous rejection, as the bad data will be
detected in some epochs anyway.








On Sun, Jan 10, 2016 at 10:59 PM, Edward Wiskers <edwardwiskers at gmail.com>
wrote:

> Hi everyone,
>
> I'm wondering whether this code "EEG = eeg_regepochs (EEG, 'recurrence',
> 2, 'rmbase', NaN)" applies the recommendation of Makoto's preprocessing
> pipeline (epoch length of -1 to 2 sec) or not? If not, so how to adjust it
> to make it suits the recommendation?
>
> Thanks.
>
> Edward
>
>
> _______________________________________________
> 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/20160111/c16b2244/attachment.html>