Classification of ERPs for several subjects

✉️ Mailing List Archive (read-only)
1

Hi All,
I am new to MNE and need your help to classify my ERPs in two different
conditions for 12 subjects. I would like to combine LDA and CSP for this
reason in the way that I use the ERPs of let's say 11 subjects and train my
classifier based on that and to test it on the rest of subjects.
My problem is in this line of my code:

scores = cross_val_score(clf, evoked_data_array, labels, cv=cv, n_jobs=1)

here, my evoked_data_array is an array of [2* number of subjects(24)*number
of channels (64)*number of samples of the ERP]. I get this error:
Found arrays with inconsistent numbers of samples: [24 64]

I guess that's because I am using a 3D array instead of 2D, right? Do you
have any suggestions how I can solve the issue?

Thank you very much in advance,
Forooz Shahbazi
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.nmr.mgh.harvard.edu/pipermail/mne_analysis/attachments/20151119/443493ef/attachment.html

2

hi,

did you follow this example:

http://martinos.org/mne/dev/auto_examples/decoding/plot_decoding_csp_space.html#sphx-glr-auto-examples-decoding-plot-decoding-csp-space-py

?

does it work for you?

Alex