Unrecognized field name "X" during 1st level LIMO
Opened this issue · 4 comments
Modeling 1st level on power spectra.
Each subject has a session 1 and 2. Trying to compare pre-post changes between the two sessions.
I get this error:
Would it make more sense to merge the 2 PRE/POST intervention files for each subject and have 1 event marking beginning of PRE and 1 event marking beginning of POST, and consider them 2 conditions for paired t-test? Problem is files have sometimes different number of channels and cannot interpolate (low-density montage, not enough channels).
There is an issue when some datasets are missing some channels and there are 2 sessions (for the same subject, some sessions have some channels and the other session has others). The the code in std_limo line 102 (below) generates an error because the
if size(data1,1) ~= size(data2,1) || size(data1,2) ~= size(data2,2)
error('Dataset have different dimensions - reprocess sessions with the same parameters')
end
Cedric, you can interpolate channels for now.
@arnodelorme with 4 channels? Sometimes they have only 1 or 2 channels left...
Not completed. Cyril needs to look at the issue with the different sessions.