Our MEG Center has just upgraded the MEG machine control computer to a Linux box.
I ran a study yesterday and saved both averages and raw data for each block of trials.
The averaged data looks fine but the raw data files do not appear to contain the digitizer or head position information. I have tried both Elekta's show_fiff and mne_show_fiff without success. Both blow up on a "segmentation fault."
Also both mne_browse_raw and mne_process_raw fail with the same error. I download the latest nightly build and tried the versions in there with the same result.
Does it make sense that the absence of head position info in a raw data file would cause this fault. If that's it, is it difficult to trap the error so that the mne_process_raw, etc. can just keep going? And what is the easiest way to insert the digitizer and head position information into the raw data files, since I have it from the average files, fortunately.
Thanks.
Don
Don Krieger, Ph.D.
Department of Neurological Surgery
University of Pittsburgh
(412)648-9654 Office
Sorry to hear about the errant saving. One option would be to try using
mne-python -- it might be possible to load the raw file there, add the
necessary info to the header (stored in mne-python in raw.info), and save
it.
If you're not too familiar with python, if you can make a sample errant
file available, I could try loading it in mne-python myself to see if it's
possible (and thus worth your time to get a python setup going). In any
case, it's possible that an error thrown in python would be more
informative than a segfault regarding the problem.
We figured out Don's problem locally, it turned out that the files were corrupted during transfer. Thanks for your help.
Best Wishes,
Avniel
Avniel Ghuman, Ph.D.
Laboratory of Cognitive Neurodynamics
Director of MEG Research
Assistant Professor of Neurological Surgery and Neurobiology
Faculty in the Center for the Neural Basis of Cognition
University of Pittsburgh