Perform Simulation Analysis — simulation

This function performs a simulation analysis for a given environment using the specified analysis configuration and configuration directory. It reads the output configuration, loads results, applies frequentist power analysis, and computes the sweet spot.

Usage

simulation_analysis(
  env,
  analysis_config,
  config_dir,
  frequentist_metrics,
  case_studies = "all",
  to_compute = c("frequentist_power_at_equivalent_tie",
    "frequentist_power_at_nominal_tie", "sweet_spot", "bayesian_ocs"),
  methods = "all",
  case_studies_config_dir = NULL,
  parallelization = NULL,
  n_replicates = NULL
)

Arguments

env: A character string specifying the environment for the analysis (e.g., "development", "production").
analysis_config: A list containing the analysis configuration parameters.
config_dir: A character string specifying the directory containing the configuration files.
frequentist_metrics: List of frequentist metrics.
case_studies: Case studies to analyse, or "all".
to_compute: Which analyses to run: any of "frequentist_power_at_equivalent_tie", "frequentist_power_at_nominal_tie", "sweet_spot" and "bayesian_ocs".
methods: Methods to analyse, or "all".
case_studies_config_dir: Directory holding the case study YAMLs.
parallelization: Whether the analysis runs in parallel: a single logical, or a list of method names, in the same form as the parallelization entry of scenarios_config.yml. Defaults to that entry, read from config_dir.
n_replicates: Number of Monte Carlo replicates the simulated power estimates are built from. Defaults to the n_replicates entry of scenarios_config.yml, read from config_dir.

Value

This function does not return a value. It writes the results and sweet spot analysis to CSV files in the specified results directory.