bids-examples
bids-examples copied to clipboard
bids-standard
ENH: Add models for ds003, ds114
Thanks, you seem to have made this PR from your old fork ... are you sure that this will not reintroduce some bad history?
Why not delete your old fork, make a fresh one, and proceed?
I see ... I also just realized that we won't introduce bad history via PRs, but rather via direct pushes to master.
Is this ready to be merged? If you need a review, you'd have to tag a member that has some experience with the models spec (I am not familiar enough with that).
No, it's possible for someone to introduce bad history, but it should show up as a ton of commits in the PR. Typically just asking them to rebase on master will do the job. But re-forking is fine, too.
I'll request review from @tyarkoni and @adelavega.
I looked through the models (admittedly not very carefully, but short of trying to run them through pybids and fitlins, I don't know that my careful review is terribly helpful) and they LGTM.
Edit: one caveat is that the first model is missing intermediate steps between run and dataset, which I believe means that the dataset level should expect to receive multiple runs per session and/or subject. My recollection is that currently aggregating within these intermediate levels must be explicit (e.g., via autocontrast). Not sure if this is the intended behavior, and maybe my memory's wrong, but flagging it just in case.
