Skip to contents

Build a Union Cohort Definition

Source: R/buildDependentCohorts.R
buildUnionCohort.Rd

Creates a SQL file and metadata for a union cohort that combines multiple input cohorts. Returns a CohortDef object ready to add to a CohortManifest.

Usage

buildUnionCohort(
  label,
  cohortIds,
  unionRule = "any",
  atLeastN = 2L,
  cohortsDirectory = NULL,
  manifest = NULL
)

Arguments

label

Character. User-friendly name for the union (e.g., "Chronic Kidney Disease Phenotypes")

cohortIds

Numeric vector (minimum 2). Cohort IDs to union.

unionRule

Character. One of 'any', 'all', 'at_least_n'. Default: 'any'

  • 'any': subjects appearing in ANY input cohort

  • 'all': subjects appearing in ALL input cohorts

  • 'at_least_n': subjects appearing in at least N cohorts

atLeastN

Integer. Number of cohorts required (only if unionRule='at_least_n'). Default: 2

cohortsDirectory

Character. Path to inputs/cohorts/. Uses study hierarchy if not provided.

manifest

CohortManifest object (optional). If provided, validates that all input cohorts exist. Recommended to ensure referential integrity. If NULL, a warning is issued.

Value

A CohortDef object with cohortType='union' and dependencies set.

Details

Creates three files:

  • SQL file: inputs/cohorts/derived/union/union_cohorts_{cohort_id_list}.sql

  • Metadata JSON: Same path with .json extension

  • Context file: .metadata with rule description