Plot all channels with raw.plot_psd

[cbrnr]

We have already discussed and fixed the problem that it was impossible to plot the PSD of bad channels (#6637, #6678). However, I find the solution still not ideal. I frequently need to plot PSDs of all channels, and the only way to do that is to pass all channel indices or names to picks. I'd prefer a string value, but unfortunately "all" and "data" are already taken and just include good channels. There's also the default third option (None), which seems to be equal to "all".

Does anyone have a good suggestion how to solve this issue?

[agramfort]

I would do it such that:

raw.plot_psd(spatial_colors=False, exclude=None)

plots all channels including bads. Woudl you also plot simple with different line style or color?

[cbrnr]

I would do it such that:

raw.plot_psd(spatial_colors=False, exclude=None)

I guess you mean raw.plot_psd(spatial_colors=False, picks=None)? Yes, this would work, but then we need to change the default to picks="all" in order not to break current behavior.

Woudl you also plot simple with different line style or color?

Not sure what you mean, but usually I use the default spatial_colors=True but with average=False (otherwise I'd get an average over all channels).

[agramfort]

I guess you mean raw.plot_psd(spatial_colors=False, picks=None)? Yes, this would work, but then we need to change the default to picks="all" in order not to break current behavior.

presently the default is picks=None and this plots all good channels. I propose to add an exclude param that would default to 'bads' and that would not change the default behavior. So when you change the default with raw.plot_psd(spatial_colors=False, picks=None, exclude=[]) it would plot the bads. This is consistent with evoked.plot for example.

Woudl you also plot simple with different line style or color? Not sure what you mean, but usually I use the default spatial_colors=True but with average=False (otherwise I'd get an average over all channels).

I would copy the behavior of evoked.plot (colors when exclude is []) makes sense?

[cbrnr]

Sounds good. Any objections? Otherwise I'll make a PR.

[larsoner]

Sounds reasonable to me

[drammock]

+1 from me anytime we're make API + behavior more consistent between different plotting functions :)