Open yarikoptic opened 2 years ago
NWB is a permitted file format for iEEG only (currently), perhaps someone has stumbled over this issue already? How did you deal with it? @bids-standard/raw-electrophys-ieeg
Personally, I don't know anyone who has worked with NWB in the context of BIDS.
This does not only apply to NWB, but to all formats where the original header is difficult or impossible to edit by the neuroscientists that acquire the data and do the BIDS curation.
In discussions with @hoechenberger and @cbrnr et al on the MNE-Python repository we (as software developers) agreed that the BIDS sidecars overrule the (binary) headers of the MEG, EEG and iEEG files. The discussion starts more or less around here.
Ah, that reminds me of:
which would be great to finally agree on and merge
NWB is a permitted file format for iEEG only (currently), perhaps someone has stumbled over this issue already? How did you deal with it? @bids-standard/raw-electrophys-ieeg
Note that in this case specifically, NWB is 'permitted', but not 'supported', which means that the validator does not validate the NWB metadata.
@sappelhoff
Personally, I don't know anyone who has worked with NWB in the context of BIDS.
we hope to possibly/eventually converge the https://dandiarchive.org custom bids-inspired layout to fully fledged BIDS. BEP032: BIDS-animal-ephys is working with .nwb being one of the two supported data formats.
@dorahermes
NWB is 'permitted', but not 'supported', which means that the validator does not validate the NWB metadata.
oh, I think I was not aware of such distinction. Is it (supported
vs permitted
) described anywhere within BIDS, e.g. is it just a matter of validating internal structure/data/metadata of the data files or there is more to it? e.g. are .jpg
supported
or permitted
?
Ah, that reminds me of:
which would be great to finally agree on and merge
I added Closes #1133
to that PR description since it would really resolve this question ;) Thank you @sappelhoff
.nwb files are supported by BIDS since awhile. NWB files are hdf5 files following quite elaborate nwb schema which includes metadata which is also covered by BIDS itself, such as subject_id, session_id, etc. I wonder, are we to allow such metadata be "overloaded" by BIDS (so it is
sub-001
in BIDS while .nwb file says "mouse3") or they must be consistent (and thus bids-validator should error out)?