Closed neuromechanist closed 4 months ago
Looks good.
Thanks Yahya. I usually say save the dataset first (as 2 files) so the datfile field is filled in and no writing of data is necessary. Is the new approach recommended because it's tidier?
Jason
From: Arnaud Delorme @.> Sent: Sunday, April 28, 2024 11:44:55 PM To: sccn/amica @.> Cc: Jason Palmer @.>; Review requested @.> Subject: Re: [sccn/amica] Create data file when EEG.datfile is not present. (PR #43)
Merged #43https://github.com/sccn/amica/pull/43 into master.
— Reply to this email directly, view it on GitHubhttps://github.com/sccn/amica/pull/43#event-12637194975, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ACRBESXTGMKLP5P5DHUEC2LY7W63PAVCNFSM6AAAAABG5R7MR2VHI2DSMVQWIX3LMV45UABCJFZXG5LFIV3GK3TUJZXXI2LGNFRWC5DJN5XDWMJSGYZTOMJZGQ4TONI. You are receiving this because your review was requested.Message ID: @.***>
Anytime. Yes, the current default and recommendation for BIDS datasets is to save them as a single SET file, since there are already several auxiliary files for each task/run in the eeg/session directory.
I believe this change also resolves one of @dungscout96's issues with using AMICA in his NEMAR processing pipeline.
If
EEG.datfile
is empty, it seems that the parameter file just would point to the file directory, which makes AMICA not working.The fix is that if the EEG.datfile is absent, the script creates the data file from
EEG.data
. This is especially compatible with the new approach to keep the EEG files in only one SET file, (without the FDT file)