Skip to contents

Load Cohort Manifest from Database or Cohort Files

Source: R/manifest_cohorts.R
loadCohortManifest.Rd

Loads a CohortManifest R6 object by either reading from an existing cohortManifest.sqlite database or by scanning the inputs/cohorts directories. ExecutionSettings are optional and only required if you plan to generate cohorts or retrieve cohort counts. You can load the manifest without them to review metadata.

Usage

loadCohortManifest(
  cohortsFolderPath = here::here("inputs/cohorts"),
  executionSettings = NULL,
  verbose = TRUE
)

Arguments

cohortsFolderPath

Character. Path to the cohorts folder containing the manifest database and cohort definition files. Defaults to "inputs/cohorts". The function will look for:

  • cohortManifest.sqlite in this folder for existing manifest data

  • json/ subfolder for CIRCE JSON cohort definitions

  • sql/ subfolder for SQL cohort definitions

executionSettings

An ExecutionSettings object containing database configuration for cohort generation. Optional; only required if you plan to generate cohorts or retrieve cohort counts. Defaults to NULL. You can add settings later using setExecutionSettings() on the returned CohortManifest object.

Value

A CohortManifest R6 object initialized with all cohorts found.

Details

If database exists: Loads cohort paths and metadata from the cohortManifest.sqlite database, verifies files still exist, and checks if any files have changed by comparing the stored hash with the current file hash.

If database doesn't exist: Scans cohortsFolderPath/json and cohortsFolderPath/sql directories to find cohort definition files and creates a new CohortDef for each file with:

  • label: The basename of the file without extension

  • tags: Empty list

  • filePath: The full path to the cohort file

Metadata Enrichment (optional): If a cohortsLoad.csv file exists in cohortsFolderPath, the function will automatically enrich CohortDef objects with tags by matching the file_name column from the load file with the filePath of each entry. For matching entries, tags are added from the following columns:

  • atlasId: Added as an "atlasId" tag

  • category: Added as a "category" tag

  • subCategory: Added as a "subCategory" tag

Hash comparison alerts:

  • ✓ Unchanged: Hash matches stored value

  • ⚠ Changed: Hash differs from stored value (file was modified)

Examples

if (FALSE) {
  # Load manifest for metadata review (no settings required)
  manifest <- loadCohortManifest()
  
  # Or load from custom path
  manifest <- loadCohortManifest(cohortsFolderPath = "path/to/cohorts")
  
  # Add execution settings later if needed for cohort generation
  settings <- ExecutionSettings$new(
    databaseName = "mydb",
    dbms = "postgresql",
    connectionDetails = list(...)
  )
  manifest$setExecutionSettings(settings)
}