WRCC2023_MI_A/B/C from Harvard Dataverse, and the card at 32 datasets
#2
by bruAristimunha - opened
Adds the three WRCC2023 motor imagery sets and updates the card to 32 datasets.
- WRCC2023_MI_A, _B, _C: Harvard Dataverse J9JFES, HFNWRX, G8FBHH v1.0, CC0 1.0; 9 + 9 + 8 files, 4.4 GB, stored as MOABB at 555d14a caches them (api/access/datafile/). Each file matched its upstream MD5 when staged and its SHA-256 again before this upload. MOABB at 555d14a loads all 26 subjects from these files with downloads blocked.
- Card: headline counts, At a glance, Verification, the Datasets table (classes, trials per class per session, channels, rate, participants, from mirror.py's integrity facts with downloads blocked; 0 disagreements with the facts committed on 8-9 Oct over 124 shared sessions), Loading sections and citations for WRCC2023. Built on e9216ec.
- Ma2022 is not included: figshare's current release keeps its files in password-protected zips and MOABB reads it only from NEMAR nm000288, which is not public.
- Every existing file is unchanged.
bkowshik changed pull request status to merged