Currently metadata is structured by experiment_type (ex. 'reinforcement', 'photometry', etc.) and maps from session uuid --> metadata fields.
BUT some animals have sessions in different dataframes with different subject-level metadata. Ex. 'dlight-chrimson-1' has a 'sex' field in optoda_raw_data/closed_loop_behavior.parquet but no 'sex' field in dlight_raw_data/dlight_photometry_processed_full.parquet.
Obviously the sex of 'dlight-chrimson-1` doesn't change for different sessions, so we need to propagate these subject-level metadata fields to all sessions related to each subject.
Currently metadata is structured by
experiment_type(ex. 'reinforcement', 'photometry', etc.) and maps from session uuid --> metadata fields.BUT some animals have sessions in different dataframes with different subject-level metadata. Ex. 'dlight-chrimson-1' has a 'sex' field in
optoda_raw_data/closed_loop_behavior.parquetbut no 'sex' field indlight_raw_data/dlight_photometry_processed_full.parquet.Obviously the sex of 'dlight-chrimson-1` doesn't change for different sessions, so we need to propagate these subject-level metadata fields to all sessions related to each subject.