Automated OHDSI PatientLevelPrediction Study Package Generator — export_plp_study

Scaffolds an end-to-end reproducible OHDSI PatientLevelPrediction (PLP) R study package with pre-configured Shiny diagnostics, cohort queries, and cross-validation workflows.

Usage

export_plp_study_package(
  output_dir,
  package_name = "PlpPredictionStudy",
  target_id = 1L,
  outcome_id = 2L,
  db_path = "omop_cdm.duckdb",
  model_type = "lasso",
  cv_folds = 5L,
  test_fraction = 0.25,
  study_title = "OHDSI Patient-Level Prediction Study",
  author = "OHDSI Study Consortium",
  version = "0.1.0"
)

Arguments

output_dir: Destination directory where the study package will be created.
package_name: Name of the R package (default "PlpPredictionStudy").
target_id: Cohort definition ID of the target population (default 1L).
outcome_id: Cohort definition ID of the outcome event (default 2L).
db_path: Default DuckDB database file path (default "omop_cdm.duckdb").
model_type: Predictive model type: "lasso", "gradient_boosting", "random_forest", or "logistic".
cv_folds: Number of cross-validation folds (default 5L).
test_fraction: Fraction of patients in test split (default 0.25).
study_title: Human-readable study title.
author: Author or consortium name.
version: Package version string.

Value

Character string containing the absolute path to the generated study package directory.