[Eeglablist] EEGLAB BDF script

[Julie Onton]

Thank you Brad, this should be helpful. But I also want to add one more comment to this. If you
intend to epoch your data after importing the continuous file, which often results in overlapping
epochs, you may multiply your data to an extent that exceeds matlab memory limits.

Best,  Julie

-- 
Julie Onton, PhD
Swartz Center for Computational Neuroscience
Institute for Neural Computation
University of California, San Diego
(858) 458-1927 ext 17
http://sccn.ucsd.edu/~julie

> I sent this to Arno, and he thought I should just send it on to the whole list.
>
> I've seen a number of messages on here, and have spoken with several
> folks here at Berkeley, about memory issues when importing BioSemi BDF
> files. I'm sure you guys already know this, but the issues stem from
> the fact that multiple versions of the dataset are created. If the
> original dataset is > ~400MB or so, then the import fails.
>
> The BioSemi system can indeed downsample data for you, however even at
> a modest 256Hz sampling rate, a long recording session can easily go
> over this 400 MB limit.
>
> I've written a pretty dirty, but functioning script that takes all the
> raw BDF files in the pwd, imports them in segments, saves the
> segments, reloads them (weird I know, but it avoids some memory
> issues), merges them (within the script, to avoid the fact that
> pop_mergeset creates a copy of the dataset as well), saves the merged
> version, and removes the temporary files.
>
> This lets you import BDF files as big as at least 1GB I'd say, though
> in theory if you make the segments small enough, you could import
> files close to 2GB (MatLab's memory limit on non-Linux machines).
>
> There are probably some things at the bottom of the script other
> people would want to change: I have this script look up channel
> locations on the Mac I'm using, and I also have it create a standard
> naming structure for the files. But those are minor tweaks.
>
> Hopefully someone on the list could find this useful.
>
> Cheers!
>
> ::brad
>
> --
>
> Bradley Voytek
> PhD Candidate: Neuroscience
> Helen Wills Neuroscience Institute
> University of California, Berkeley
> 132 Barker Hall
> Berkeley, CA 94720-3190
>
> btvoytek at berkeley.edu
> 213.840.4560 (cell)
> _______________________________________________
> eeglablist mailing list eeglablist at sccn.ucsd.edu
> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
> To unsubscribe, send an empty email to eeglablist-unsubscribe at sccn.ucsd.edu