Updated to show a comprehensive HED example.

[VisLab]

This commit adds a new eeg_ds003645s_hed_demo dataset, which demonstrates various uses of HED in a BIDS dataset. In addition to annotating the events.tsv files, the dataset demonstrates the use of HED to annotate participants.tsv, samples.tsv, scans.tsv, and phenotypes/*.tsv.

Validation of HED in these BIDS files is supported starting in hed-validator 3.13.3.

This example supersedes:

• eeg_ds003645s_hed
• eeg_ds003645s_hed_inheritance
• eeg_ds003645s_hed_longform,

all of which have now been removed. @sappelhoff this the consolidation of HED examples as mentioned previously.

[VisLab]

@rwblair @sappelhoff:

I believe that the error in this is incorrect handling by the BIDS validator. The phenotypes does not validate if there is more than one entry for a subject (which I think it should allow according to the spec). Further, when there is more than one .tsv file in the phenotypes subdirectory, the dataset won't validate. (I gave in and took out multiple .tsv's for the example, but I think this is also an error.) Thx!

[sappelhoff]

Is this the issue you reported in

• https://github.com/bids-standard/bids-validator/issues/1874

?

[VisLab]

Is this the issue you reported in

* [phenotype doesn't appear to allow more than one assessment. bids-validator#1874](https://github.com/bids-standard/bids-validator/issues/1874)

?

This is related, but an additional issue. I guess I will put an addendum on issue #1874. This issue is that when the dataset has sessions, multiple entries in a phenotype/*.tsv for the same subject in a phenotype file won't validate. It has nothing to do with HED. I'll remove the multiple entry on this example or the entire phenotype file if necessary to move this example forward. Thx

[VisLab]

Yes...we think it is good to go...

[VisLab]

hed-javascript version 3.13.3 was merged with bids-validator master on January 24, 2024. It fixes the issues that caused this example not to validate.