Skip to content

Thapa2025: declare sessions per the loader-default rule (b) - #1222

Merged
bruAristimunha merged 1 commit into
NeuroTechX:developfrom
bruAristimunha:fix/rule-b-sessions
Oct 5, 2026
Merged

bruAristimunha merged 1 commit into
NeuroTechX:developfrom
bruAristimunha:fix/rule-b-sessions

Conversation

@bruAristimunha

Copy link
Copy Markdown
Collaborator

Applies the counting rule Bruno fixed on 2026-10-05: sessions_per_subject describes what the loader returns by default, the cohort distribution goes in the docstring. Thapa2025 released 23 subjects with 49 sessions ({1 session: 2 subjects, 2: 16, 3: 5}, matching the paper and the converted BIDS tree on nm000324); the docstring and metadata now say so instead of a flat '3'. The NEMAR deposit README (nm000324) and PardoGarcia2026's (nm000309) were updated the same way; PardoGarcia2026's loader already followed the rule. Synthetic test adjusted; no behaviour change.

…rule b)

sub-02 and sub-17 stopped after a single session; the released cohort has
{1: 2 subjects, 2: 16 subjects, 3: 5 subjects}, 49 sessions total. Per
Bruno's rule (b), the declared count must describe what the loader
returns by default. Follow the minimum-sessions convention already used
by PardoGarcia2026 and declare sessions_per_subject=1 (both in METADATA
and super().__init__), with the actual distribution kept in the
docstring. Behaviour-preserving: data_path still returns every session
present on disk, and the n_sessions-dependent get_data iteration is
unchanged. Test test_thapa_transport_and_missing_sessions now asserts
n_sessions == 1.
@bruAristimunha
bruAristimunha merged commit 019dea0 into NeuroTechX:develop Oct 5, 2026
12 of 14 checks passed
@bruAristimunha
bruAristimunha deleted the fix/rule-b-sessions branch October 6, 2026 10:19
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant