[MISC] Expand entity table for MEG/EEG/iEEG specific files

[sappelhoff]

closes #173

labelled misc because it's not really a new feature but was simply forgotten when merging EEG and iEEG

A human readable version of the changes is here: https://520-150465237-gh.circle-artifacts.com/0/home/circleci/project/site/99-appendices/04-entity-table.html (NOTE: this link is outdated, check the fresh circleCI build to see / ask me how to do that)

or probably easier, see the google doc (NOTE: gdoc is now outdated)

[choldgraf]

Two quick things (that are technically out of scope for this PR, so feel free to say "let's do this in another PR").

• I think the table should be transposed. The columns will continue growing while the rows probably won't, and already we have a wider-than-higher table, which I think would render better if it were transposed. This is pretty easy to do with a little pandas, happy to try helping out here.
• Could we add lists to the bottom of this table that include the list of extensions associated with each type of file? I think that'd make the table a bit easier to parse since currently some cells are a lot bigger than others.

The changes directly in this PR look good to me though!

[sappelhoff]

Those sound like to excellent suggestions to me! A little help would be greatly appreciated: Ideally in the form of a new PR, as you suggested :-)

[choldgraf]

done #200

[choldgraf]

This LGTM - a few changes I think we should make in this table (but not related to this PR, issue opened in #200 )

[sappelhoff]

@robertoostenveld @dorahermes would you have time to review this in the next week? Or can you suggest someone to review in your stead?

[jasmainak]

It's really hard to know looking at the rendering what was added and what wasn't :-(

To answer your question from the google doc, proc is only for sss or tsss data AFAIK and that too it is not clear because BIDS is not supposed to cover derived data for MEG at the moment

[monkeyman192]

It's really hard to know looking at the rendering what was added and what wasn't :-(

I agree and I would say the table itself is getting a bit hard to read since it is so large, but I can't really think of a better way to format it...

[sappelhoff]

Yep, the bad readability will be taken care of in a different PR, Chris H has already opened an issue: #200

Unless there are complaints soon, I will go ahead and merge this ... and then we can improve it even further in future PRs. At least we will have EEG and iEEG modalities/files represented.

[robertoostenveld]

I had a deja vu when recent emails on this PR came in, but now realize that this was also discussed here.

The label task should not be required for eeg/ieeg electrodes (it is just like headshape and photo). It should also not be optional. Looking at the published spec for ieeg and eeg, I see that ieeg does not have it, but eeg has. I see that meg also does not have it.

So task-<label> for electrodes should not only be removed here, but also in the EEG spec.

[sappelhoff]

Thanks @robertoostenveld I removed the task parameter for electrodes. In the spec, it will be removed in #188, which is also ready to be merged from my perspective

[robertoostenveld]

acq-<label> will be removed from meg, eeg, ieeg as per ##188. That should also be reflected here. Consequently, acq-<label> should also not be optional for meg/eeg/ieeg events.

Other than that I an fine, so let me approve already without requiring another pass at it.

[sappelhoff]

Thanks for the reviews. I am merging this now and with #200 we will further improve on this.