schemas

Schemas for digest input files

Overview

digest supports TSV files as inputs. Expected columns in a digest file are defined in JSON files called digest schemas. There are different schemas for digest files containing different modalities of data (e.g. imaging vs. phenotypic).

Schema	Modality of data in corresponding digest file
imaging_digest_schema.json	Processing pipeline output and derivative availability
phenotypic_digest_schema.json	Demographic and phenotypic assessment data

How to read the schema

Column categories

Within a schema, columns are grouped into two semantic categories. These categories are purely for organizational purposes and do not appear in a digest file.

Global columns: Columns describing basic metadata that should have the same meaning regardless of the specific event described by a given record (e.g., a certain processing pipeline or phenotypic assessment), and do not depend on event outputs.

Event-specific columns: Columns whose values may have event-specific meanings. e.g., in the schema for an imaging digest, the "PIPELINE_STATUS_COLUMNS" convey info about processing pipeline completion that depends on pipeline-specific outputs and may have varying values depending on the pipeline tracker used.

Column attributes

Recognized columns are described in the schema using the following attributes:

• Description
  • Describes the column contents, along with the meaning of different acceptable values for categorical columns.
• dtype
  • The expected type for values in the column.
• IsRequired
  • Whether or not the column is required to be present in the digest file. If true, the dashboard will throw an error if the column is missing.
• Range
  • Acceptable values for data in the column. Only present for categorical columns.