[WIP] Participants data always null. Issue 49 never thrown.

[rwblair]

• closes Participant data not being returned #1755

Participants was being defined as a const and assigned null. Usual use case for consts is to give it an actual array or object with reference and then later on push to that array or object. For whatever reason tsv validation attempts to reassign the participants reference it gets.

[codecov]

Codecov Report

Patch coverage: 100.00% and project coverage change: -2.32% ⚠️

Comparison is base (5ec0b3d) 85.71% compared to head (8122bcb) 83.40%.

❗ Current head 8122bcb differs from pull request most recent head cf46045. Consider uploading reports for the commit cf46045 to get more accurate results

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1772      +/-   ##
==========================================
- Coverage   85.71%   83.40%   -2.32%     
==========================================
  Files         130       92      -38     
  Lines        6575     3893    -2682     
  Branches     1524     1270     -254     
==========================================
- Hits         5636     3247    -2389     
+ Misses        833      548     -285     
+ Partials      106       98       -8     

Files Changed	Coverage Δ
bids-validator/validators/bids/fullTest.js	98.92% <100.00%> (ø)

... and 38 files with indirect coverage changes

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[rwblair]

We now get code 67 NO_VALID_DATA_FOUND_FOR_SUBJECT for any bids-examples with a participants.tsv.

ieeg_motorMiller2007 has rows in participants.tsv that have no directories in the dataset. Throws code 49 PARTICIPANT_ID_MISMATCH

[sappelhoff]

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?

[rwblair]

Suggestion from maintainers meeting: for legacy validator automatically add 67 to ignored codes if --ignore-nifti-headers used, prevent us from having to edit ignored codes in bids-examples tests.

[dorahermes]

It seems that 3 subjects are missing folders (jf, gf, jp). Data from these subjects is located in the original ECoG library. Let me know if you prefer option 1 or 2:

1. Remove these three subjects from the BIDS example participants.tsv file
2. I could convert these 3 subjects to BIDS and do a pull request to update accordingly

(I am not sure why I missed these subjects in the conversion. One thing I vaguely remember is that there was a mismatch in electrodes and channels, but the current version of the ECoG library does not have any mismatches. There was an update in 2022 though.)

[effigies]

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.

[sappelhoff]

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