Calculating the inverse solution using the fusion method properly

[young-x-skyee]

The noise covariance matrices for Russian and English dataset seem to have different structure - the entries dedicated to the EEG and MEG parts are different. How is that information reflected in the fif files and then used properly for noise covariance matrix calculation?

[young-x-skyee]

Checked with @neukym that the orders of channels for the English and Russian datasets are different from the very beginning. We need to make sure the ordering is correct for the all the channels when calculating the noise covariance matrix using fusion method.

[neukym]

Closed with above fix.