Potential improvements of the documentation about CSP

Dear MNE team,

First of all, thank you for creating and maintaining such a nice collection of tools for working with neurophysiological data!

I recently worked with this tutorial on applying CSP for decoding imaginary hand movements, and thought that it was lacking some notes about interpretation of the resulting patterns. When component_order is set to mutual_info, it might not be clear which patterns correspond to which class, and their order may be different for every fit-transform run.

The current tutorial is more focused on applying the CSP method for classification, and I thought that it could be useful to mention interpretation a bit. I would be happy to work on it and make a PR, just decided to start by creating a topic on this forum according to the guide for contributors. How should I proceed if you agree that it could be a useful addition to the documentation?

Best regards,
Nikolai Kapralov.



Lets open a PR and discuss content on GitHub