FittingRunner

Synchronous driver that turns one fit configuration into a serializable FitResultData by orchestrating the appropriate engine: PDFModeler for the continuous path and PMFModeler for the discrete path. It owns import and the estimate/score/evaluate sequence, and hands the live results to FitResultExtractor for translation into the DTO. The live engine objects are used only transiently and are never exposed — presentation is driven entirely from the returned DTO (plus the caller's own raw data for plots).

This runner does no event emission, no validation, and no async control: the async session wraps it in a coroutine and converts thrown exceptions into a typed failure result.

Functions

Link copied to clipboard
fun familyFrequencyBootstrap(dataset: NamedDataset, config: FamilyBootstrapConfig, estimatorIds: Set<String> = emptySet(), scoringModelIds: Set<String> = emptySet(), evaluationMethod: EvaluationMethod = EvaluationMethod.SCORING, automaticShifting: Boolean = true, catalog: FittingCatalog = FittingCatalog): FamilyFrequencyResult

Runs the family-frequency bootstrap (Kind 2) — a standalone, continuous analysis, separate from fit. It resamples dataset config.numSamples times, re-runs the full fit + evaluation on each resample, and tallies how often each family is the recommended fit. Estimators/scoring default to the continuous catalog defaults unless provided.

Link copied to clipboard
fun fit(config: FitConfiguration, importer: DatasetImporter = DatasetImporter.default, catalog: FittingCatalog = FittingCatalog): FitResultData

Imports the configured data and runs the fit on the path selected by config.kind.