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've run into what might be an edge case. I have one participant with missing imaging data but questionnaires filled in. This participant is thus present in the phenotype folder and participants.tsv. This raises the error PHENOTYPE_SUBJECTS_MISSING when running the BIDS validator. Maybe there is something that I am missing, but it seems like the validator does not expect this scenario.
Yes. It is also causing errored validation on my old OpenNeuro dataset ds004215 (now ds005752). In our case there are 1,608 subjects with survey (phenotype/) data, but without sub-*/ directory data.
Activity
sappelhoff commentedon Oct 8, 2022
Hi @JonteP, this may be of interest to you:
effigies commentedon Aug 29, 2024
Is this an issue in the current (schema) validator?
ericearl commentedon Jan 22, 2025
Yes. It is also causing errored validation on my old OpenNeuro dataset ds004215 (now ds005752). In our case there are 1,608 subjects with survey (
phenotype/) data, but withoutsub-*/directory data.effigies commentedon Jan 22, 2025
@ericearl My understanding is that this needs to be resolved in the specification (and schema). bids-standard/bids-specification#914 is where this needs to be finalized.
My reading is that we came to a position that seemed to work, and nobody objected, but a PR was never opened.
With the rule as written at the bottom, I don't think the expression language can express that rule yet, but we can figure that out.
ericearl commentedon Jan 22, 2025
@effigies Cool, I will work on a PR to the spec now.
ericearl commentedon Feb 5, 2025
PR added here: bids-standard/bids-specification#2044.