You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I missed it during implementation but the runs are expected to be indexes in BIDS (see https://bids-specification.readthedocs.io/en/stable/appendices/entities.html#run).
But in a few datasets (such as Lee2019), run indices are followed by a description string (like 1train or 2restingState). This makes it impossible to save these datasets in cache rn:
ValueError: run is not an index (Got 1train)
A possible solution would be, when saving such datasets, to split the run name, and encode the index part as run index and the string part as another entity field. But the problem is that no field perfectly matches this need. The best is probably the task field IMHO but we already use it to specify the paradigm.
I would argue we don’t actually need to specify the paradigm because, in every case, the whole dataset uses the same paradigm.
The text was updated successfully, but these errors were encountered:
I missed it during implementation but the runs are expected to be indexes in BIDS (see https://bids-specification.readthedocs.io/en/stable/appendices/entities.html#run).
But in a few datasets (such as
Lee2019
), run indices are followed by a description string (like1train
or2restingState
). This makes it impossible to save these datasets in cache rn:A possible solution would be, when saving such datasets, to split the run name, and encode the index part as run index and the string part as another entity field. But the problem is that no field perfectly matches this need. The best is probably the
task
field IMHO but we already use it to specify the paradigm.I would argue we don’t actually need to specify the paradigm because, in every case, the whole dataset uses the same paradigm.
The text was updated successfully, but these errors were encountered: