Adding diagnostic functions

[christian-oreilly]

We should add functions that allow, given a config file and a raw object, to generate diagnostic figures for each of the steps. Representations shown in the first two figures of this paper https://pubmed.ncbi.nlm.nih.gov/33038417/ are probably a good starting point.

[scott-huberty]

It actually might be helpful for us to have some form of this done by paper submission (so end of June)?

[christian-oreilly]

Sure. But I think this is still milestone 0.02 (i.e., not blocking 0.01).

[scott-huberty]

This is surely far off (not 0.1) - but having some functionality similar to MNE's ica.property_plots Would be really great for a deeper inspection of ICA components!