Per-subject results, analysis code and verification scripts for "Session Alignment Placement and Sparse Representation-Based Classification in Cross-Session Motor Imagery EEG Decoding: A Two-Dataset Comparison"

Per-subject results, analysis code and verification scripts accompanying the manuscript "Session Alignment Placement and Sparse Representation-Based Classification in Cross-Session Motor Imagery EEG Decoding: A Two-Dataset Comparison". Contents: per-subject cross-session accuracies for every arm on both datasets (OpenBMI, 54 subjects; WBCIC-SHU, 51), from which every table and figure in the manuscript is computed; the MATLAB analysis code, including the preprocessing script for OpenBMI and the conversion script for WBCIC-SHU; the written analysis plan, fixed before the runs, with its three dated amendments and three dated result records; the number register, which holds every reported quantity with its source; the scripts that regenerate every deposited result; and the scripts that check the register against the deposited results and draw the figures. A README documents the layout, the columns of each file, and how to rerun. Source data: OpenBMI (Lee et al. 2019), https://doi.org/10.5524/100542, CC0 1.0; WBCIC-SHU (Yang et al. 2024), https://doi.org/10.25452/figshare.plus.22671172, CC BY 4.0. Neither dataset was collected by the authors and neither is redistributed here. The deposited results and documentation are under CC BY 4.0; the analysis code is under the MIT licence, as stated in code/LICENSE.