Stefan Appelhoff

+1 for making it at least RECOMMENDED as well (for all, EEG, iEEG, MEG). I would be okay with making it REQUIRED for iEEG (depends on what @dorahermes thinks is...

I see, is that an argument to **not** make electrodes.tsv *required*? How do you respond to this comment by @dorahermes in https://github.com/bids-standard/bids-validator/issues/1771#issuecomment-1714213732? > iEEG data often do not make any...

IMHO the examples described above rather call for a RECOMMENDED status, where the validator would send a warning if no `electrodes.tsv` file is found. The warning may serve the purpose...

okay, fine by me then 👍

@effigies @rwblair should we close this issue as it's targeted to the legacy-validator, and just keep the one I just opened on the bids-specification repo as a reference/todo? - https://github.com/bids-standard/bids-specification/issues/1854

Thanks, that's a good approach.

Thanks for the summary, Chris. Do I understand 1. correctly, that **in order to validate that an iEEG data file has a corresponding electrodes file** , we need to apply...

Thanks for the bug report. - the datasets in question: - https://openneuro.org/datasets/ds003620/versions/1.1.1 - https://openneuro.org/datasets/ds003800/versions/1.0.0 The mismatch between what's listed under `participant_id` in `participants.tsv` and the actual subject directories seems like...

> ieeg_motorMiller2007 has rows in participants.tsv that have no directories in the dataset. Throws code 49 PARTICIPANT_ID_MISMATCH @dorahermes could you look into this, please?

> IMO if the usefulness of the dataset as an example isn't damaged by dropping the subjects, removing them from participants.tsv seems sensible to me. agreed