-
Jurjen Broeke authored
The single long/melted aggregate CSV padded every row with the union of columns across incompatible feature types, hitting 2.5GB+ with the majority of cells blank. AggregateExporter replaces it with six wide, rectangular tables (animals, eeg_timeseries, eeg_events, eeg_duration_summary, video_timeseries, video_events) -- band power, time-frequency, and epoch diagnostics merge exactly since they share the same per-chunk window grid. Co-Authored-By:Claude Sonnet 5 <noreply@anthropic.com>
6604137eJurjen Broeke authoredThe single long/melted aggregate CSV padded every row with the union of columns across incompatible feature types, hitting 2.5GB+ with the majority of cells blank. AggregateExporter replaces it with six wide, rectangular tables (animals, eeg_timeseries, eeg_events, eeg_duration_summary, video_timeseries, video_events) -- band power, time-frequency, and epoch diagnostics merge exactly since they share the same per-chunk window grid. Co-Authored-By:Claude Sonnet 5 <noreply@anthropic.com>
Loading