Validate SDTM datasets against the SDTMIG rule catalog

Evaluate a collection of SDTM domain datasets against the bundled CDISC SDTMIG conformance rule catalog. Rules are loaded from JSON files shipped with pointblank and evaluated directly against data frames—no external tools, Docker containers, or API calls are needed at runtime.

The engine supports seven rule types: record checks (per-row value validation), dataset metadata checks (column presence, variable order), variable metadata checks, domain presence checks, dataset contents checks, and Define-XML item/codelist checks (stubbed as always-pass until a Define-XML importer is available in R).

Usage

validate_sdtmig(
  datasets,
  version = "3-4",
  ct_packages = NULL,
  rule_types = NULL
)

Arguments

datasets

list // required

A named list of data frames, where each name is a domain name (e.g., "DM", "AE", "LB") and each value is the corresponding data frame. Names are matched case-insensitively.

version

scalar<character> // default: "3-4"

The SDTMIG version string. Use hyphens (e.g., "3-4") or dots (e.g., "3.4"); dots are converted to hyphens automatically. Currently only version "3-4" is bundled.

ct_packages

vector<character> // default: NULL (optional)

Character vector of controlled terminology package slugs to load (e.g., "sdtm-ct-2024-09-27"). When NULL, the most recent bundled CT package is loaded automatically.

rule_types

vector<character> // default: NULL (optional)

Optional character vector of rule types to evaluate (e.g., "RECORD_CHECK"). When NULL, all supported rule types are run.

Value

A cdisc_conformance_result object. Use cdisc_all_passed(), cdisc_status_counts(), cdisc_issues(), cdisc_findings_df(), and print() to inspect the results.

Function ID

13-1

See also

Other CDISC: cdisc_all_passed(), cdisc_findings(), cdisc_findings_df(), cdisc_issues(), cdisc_n_total_issues(), cdisc_rules(), cdisc_status_counts()