Re: hc3 - order of channels in .eeg file

I'm analyzing the LFP recordings in some of the sessions in hc3. I'm a bit concerned about the channel order in the .eeg files. Are they always ordered such that the 1st 8 correspond to e1 (in the epos table in hc3-tables); the 2nd 8 correspond to e2, and so on? I don't see this specified anywhere, but I want to make sure I am not making a mistake.

I am concerned since the order of electrodes seems different in some of the xml files (I would have attached it but limited to very small images). In /gor01-6-7/2006-6-7_16-40-19 the channels are listed in order, but in /pin01-11-04/11-05_0-06-51, the channel listings are not in order (starts from channel 16).

Additionally, I understand that some sessions have electrodes missing. In this case, how would I know how the channels in the .eeg file correspond to each shank?

Further, are the contacts ordered from deep to superficial within each shank (as in crcns-hc2-shank_maps.pdf)?

Thanks for your time!

I think the necessary information is in the .xml file provided with each recording. In the documentation, this file is described as the "configuration file used with neuroscope software (http://neuroscope.sourceforge.net)."

However, this xml file seems to have the information I'm looking for to parse the eeg data. I've written a script, attached, to output each eeg channel as separate mat files in a way that I'm reasonably confident reflects the true electrode configuration. But I can't say 100% confident because this xml file is not mentioned in the documentation with respect to the eeg files.

That script had a bug in it, so it didn't loop through channels properly. Correct script is attached.

Also, note that there are still some bad channels: e.g. from shank 9 (1-indexing) in rats ec013 (LFP is not good) and f01 (nothing in some of the clu files).