KSLCore/ksl.app.dist.catalog

Package-level declarations

Types

data class DistributionFamilyDescriptor(val id: String, val displayName: String, val kind: DistributionKind, val rvType: RVParametersTypeIfc)

Self-describing record for one distribution family (Normal, Weibull, Poisson, ...). Derived from the set of registered estimators rather than hand-curated, so adding an estimator for a new family automatically surfaces a family descriptor without a parallel edit.

EstimatorDescriptor

data class EstimatorDescriptor(val id: String, val displayName: String, val kind: DistributionKind, val rvType: RVParametersTypeIfc, val familyId: String, val checksRange: Boolean, val factory: () -> ParameterEstimatorIfc)

Self-describing record for one parameter-estimation algorithm registered with the fitting catalog. The id is the stable, wire-safe handle that configurations and DTOs use; the factory materializes a fresh KSL estimator instance on demand.

FittingCatalog

object FittingCatalog

Stable-ID registry of every parameter estimator, scoring model, and distribution family currently reachable from the fitting subsystem.

ScoringModelDescriptor

data class ScoringModelDescriptor(val id: String, val displayName: String, val factory: () -> PDFScoringModel)

Self-describing record for one PDF scoring model registered with the fitting catalog. Scoring models are always created fresh per fit, since PDFModeler already copies them internally during scoring and a fresh instance at the catalog boundary avoids any chance of cross-fit state.