flywheel-gear-extensions

Gather Form Data

This gear gathers form data for participants across centers and writes files for each module.

It takes a CSV file listing participants and gathers data for each.

Input

The gear requires an input_file input — a CSV file with a column named naccid containing the NACCID for each participant.

Note: use the identifier-lookup gear if your input source only has adcid, ptid.

Example input file:

"naccid"
"NACC000001"
"NACC000002"

The file may have other columns.

Gear Configuration

The gear manifest config includes the following parameters:

Field	Type	Default	Description
`project_names`	string	`"ingest-form"`	Comma-separated list of projects to search for form data
`include_derived`	boolean	`false`	Whether to include derived variables or missingness information
`modules`	string	`"UDS,FTLD,LBD"`	Comma-separated list of form module names to include. Valid values: `UDS`, `FTLD`, `LBD`
`study_id`	string	`"adrc"`	The study ID. Should be set if any participants have data from an affiliated study
`formver_split`	boolean	`false`	Split output CSVs by form version. When enabled, output is split by form version producing files named `{prefix}-{module}-{formver_label}-{date}.csv`
`output_file_prefix`	string	`""`	Prefix for output file names. If empty, defaults to `study_id`. Output files are named `{prefix}-{module}-{date}.csv` (or `{prefix}-{module}-{formver}-{date}.csv` when `formver_split` is enabled)
`dry_run`	boolean	`false`	Whether to do a dry run
`apikey_path_prefix`	string	`"/prod/flywheel/gearbot"`	AWS parameter path prefix for apikey

File Metadata and Tagging

The gear updates the input file with the following metadata after processing. See the QC Conventions reference for details on the data models and conventions used.

QC Result: A validation QC result is added to the file’s file.info.qc metadata with:
- name: "validation"
- state: "PASS" or "FAIL" depending on whether all participant lookups succeeded
- data: List of FileError objects with error details if any errors occurred
File Tag: The gear name is added as a simple tag to the input file, indicating the file has been processed by this gear.

Output

A file is written for each module for which participant data is found. Columns depend on the module and whether include_derived is true.

File names have the format {prefix}-{module_name}-{date}.csv, where {prefix} is the value of output_file_prefix (or study_id if output_file_prefix is empty). For instance, allftd-uds-10-20-2025.csv.

When formver_split is enabled, output is split by form version, producing one file per module/version pair: {prefix}-{module_name}-{formver_label}-{date}.csv. Each file has a column set restricted to that exact form version; rows are not column-unioned across versions.

This site is open source. Improve this page.