When using proc_csp with the following parameters:
'CovFcn',@cov,
'ScoreFcn',@score_medianvar
the whitening matrix M (proc_csp, l.74) might not have full rank anymore.
This results in an error in score_medianvar when doing the for loop over all channels, since not all channels are accessible anymore (because of proc_linearDerivation(dat,W)).
I added some code into score_medianvar to adapt the number of channels in case W is not quadratic.
When using
proc_csp
with the following parameters:the whitening matrix M (proc_csp, l.74) might not have full rank anymore. This results in an error in
score_medianvar
when doing the for loop over all channels, since not all channels are accessible anymore (because ofproc_linearDerivation(dat,W)
). I added some code intoscore_medianvar
to adapt the number of channels in case W is not quadratic.