Run validation jobs with checkpointing and deterministic seeds

Part of the research-scale validation, advanced-model, interoperability, storage, adapter, or reproducibility programme. Experimental model functions remain subject to declared evidence gates.

Usage

run_validation_jobs(plan, simulator, fitter, extractor, truth_extractor, output_dir,
  workers = 1L, backend = c("auto", "sequential", "future"), isolation = c("auto",
  "in_process", "callr"), timeout_seconds = Inf, memory_limit_mb = Inf,
  stale_lock_seconds = 3600, overwrite = FALSE, fail_fast = FALSE,
  progress = interactive(), job_ids = NULL, chunks = NULL, simulation_args = list(),
  fit_args = list(), diagnostics_extractor = NULL, draws_extractor = NULL,
  predictions_extractor = NULL, confidence = 0.95, run_metadata = list())

Arguments

plan: Validation plan or manifest directory.
simulator: Simulation function.
fitter: Fitting function receiving the simulated object first.
extractor: Function extracting parameter estimates.
truth_extractor: Function extracting named true parameter values.
output_dir: Validation output directory.
workers: Number of workers.
backend: Sequential or optional `future` backend.
isolation: In-process execution or optional `callr` isolation.
timeout_seconds: Per-job timeout. Enforced only with `callr` isolation.
memory_limit_mb: Best-effort per-job memory limit.
stale_lock_seconds: Age after which an abandoned job lock may be reclaimed.
overwrite: Re-run completed checkpoints.
fail_fast: Stop after the first failed/nonconverged job.
progress: Display progress in sequential mode.
job_ids: Optional subset of job identifiers.
chunks: Optional subset of chunk identifiers.
simulation_args: Additional simulator arguments.
fit_args: Additional fitter arguments.
diagnostics_extractor: Optional diagnostics extractor.
draws_extractor: Optional posterior-draw extractor.
predictions_extractor: Optional prediction extractor.
confidence: Confidence level used when standard errors are supplied.
run_metadata: Named metadata included in the runner fingerprint; use it to record code, prior, or engine variants captured outside function bodies.

Value

The documented eyeprocess object, data frame, plot, report path, or adapter result.