Splits a CSV of participant data by ADCID, and writes the results into projects for the corresponding centers.
Along with the input CSV to split, the gear takes in a config YAML with the following fields:
adcid_key: <column name from the input CSV with the ADCID>
target_project: <name of the target Flywheel project to write results to per center>
staging_project_id: <ID of the staging Flywheel project to stage results to; will override target_project if specified>
include: <comma-delimited list of ADCIDs to include in the split; will ignore all others>
exclude: <comma-delimited list of ADCIDs to exclude in the split; will evaluate all others>
batch_size: <number of centers to batch; will wait for all downstream pipelines to finish running for a given batch before writing others>
downstream_gears: <If scheduling, comma-delimited string of downstream gears to wait for>
delimiter: <delimiter of the CSV, defaults to ','>
local_run: <true if running on a local input file>
dry_run: <whether or not this is a dry run - if so, will do everything except upload to Flywheel>
Some additional notes:
metadata project.staging_project_id is specified, it will write all split files to the specified staging project instead of each center’s target_project, effectively overriding the former. This can be used for preliminary review/testingadcid_key: ADCID
target_project: distribution-ncrad-biomarker