Save ECG and EOG evokeds created during ICA artifact detection to disk

[hoechenberger]

I needed access to the EOG evoked for diagnostic purposes, but figured we didn't save this data anywhere … This PR now saves the evokeds of ECG and EOG data. I considered saving the epochs instead, but at least for my usecase, I would always want to average first anyway…

Before merging …

• [x] Changelog has been updated (docs/source/changes.md)

[larsoner]

Yeah my first impulse was to say "why not save the Epochs" but then realized that I, too, always work with the averages :laughing:

So okay for me to save the evokeds for now and add epochs later if we want them!

[larsoner]

I think another dataset will come back out of date based on a failure I saw recently -- if/when it does can you try bumping the cache number (e.g., -3 to -4 or whatever) on CircleCI? 246 will probably also be broken but we can just ignore it for now.

[larsoner]

Yeah looks like 247 needs a new cache number bump

https://app.circleci.com/pipelines/github/mne-tools/mne-bids-pipeline/4527/workflows/a3292c92-606c-4f66-ba2a-7b91ddced838/jobs/65426

[hoechenberger]

@larsoner I've bumped the cache ID, let's see if it comes back green.

[larsoner]

I'll be surprised if anything works at the moment given https://www.githubstatus.com/ today but I'll restart CIs once things settle a bit

[hoechenberger]

@larsoner I just found myself wanting to do an ICA properties plot and this one requires epochs… 😶‍🌫️🫣

[larsoner]

Hah, feel free to change it to Epochs instead

[hoechenberger]

I decided that plot_components() was sufficient after all. Let's keep the averages until somebody complains :)

[larsoner]

Just 246 that's failing so in it goes, thanks @hoechenberger !