api-traceability.md

Scientific API Traceability Matrix

Back to Documentation Index

This matrix assigns every exported scientific type to a technical-reference chapter and identifies the principal implementation and verification evidence. A row marked complete has passed its source audit. A partial row has a documented evidence boundary that prevents a broader claim.

Model Infrastructure and Data

Public API	Technical treatment	Primary implementation	Verification evidence	Status
IModel, ModelBase, ModelParameter, DataComponent, DataComponentType, PriorComponent, PriorComponentType	Models overview; likelihood and prior decomposition	Model support classes	Core infrastructure and pointwise-likelihood tests	Complete
IGMMModel, IQuantilePriors, IUnivariateModel, ISimulatable, ISimulatable<TData>	Models overview; estimation; univariate models	Model support interfaces	Model-estimation and univariate tests	Complete
ParameterPenalty, QuantilePenalty, QuantilePrior, SubscriptFormatter	Priors and parameterization	Model support classes	Support and quantile-prior tests	Complete
Data, ExactData, IntervalData, ThresholdData, UncertainData	Data likelihood	Data-frame data types	Data-frame unit tests and pointwise verification	Complete
DataSeries, ExactSeries, IntervalSeries, ThresholdSeries, UncertainSeries, DataFrame	Data likelihood	Data-frame collections	Data-frame and concurrency tests	Complete
ThresholdDiagnostics, MRLPoint, MeanResidualLifeResult, StabilityPoint, ParameterStabilityResult	Peaks over threshold	Threshold diagnostics	Threshold-diagnostic tests	Complete

Univariate Models and Analyses

Public API	Technical treatment	Primary implementation	Verification evidence	Status
UnivariateDistributionModelBase, UnivariateDistribution	Univariate likelihood and distribution-family chapters	Univariate distribution model	Univariate model and measurement-error tests	Complete
Bulletin17CDistribution, Bulletin17CAnalysis, UncertaintyMethod, CohnConfidenceIntervalResult	Bulletin 17C	B17C model and analysis	Fast moment/covariance/scope tests, seven published examples, three PeakFQ cells, and the current 1,000-refit reliability diagnostic (947 outer-converged, 53 capped)	Complete for published parameter parity and reliability; numerical Cohn values and broad coverage are outside the current claim set
PointProcessModel, PointProcessAnalysis	Point-process models	Point-process model and analysis	Fast contracts plus ten Poisson/GPA, likelihood, calendar/water-year, and recovery cells under production defaults	Complete
MixtureModel, MixtureAnalysis	Mixture models	Full-$K$ public/EM boundary; identified $K-1$ posterior storage	Fast residual-weight, prior, covariance, diagnostic, persistence, and frequency-curve contracts; an independent scikit-learn optimum and retained identified recovery cells	Complete for the declared mixture matrix
CompetingRisksModel, CompetingRiskAnalysis	Competing risks	Competing-risks model and analysis	Fast seed/matrix/initialization contracts, four analytical rank/CDF cells, three identified MLE designs, and two independent Bayesian designs	Complete for the reported matrix; the Bayesian maximum uses bounded priors with the optional Jeffreys scale multiplier disabled. Correlated Bayesian recovery is outside the supported claim
CompositeAnalysis, CompositeType, AverageMethod, WeightedUnivariateAnalysis, CorrelationMatrix	Composite analysis and model averaging	Composite analysis support	Fast criterion/matrix/index/seed/immutability tests, two posterior-resampling oracles, and ten closed-form/published/Cartesian cells	Complete for the declared composite and predictive matrix
UnivariateAnalysis, FittingAnalysis, FittedDistribution	Univariate and distribution-fitting workflows	Analysis classes	Fitting and univariate tests	Complete

Analysis Infrastructure

Public API	Technical treatment	Primary implementation	Verification evidence	Status
IAnalysis, IBayesianAnalysis, IProbabilityOrdinates, IUnivariateAnalysis, AnalysisBase	Analyses overview	Analysis support interfaces and base	Analysis workflow tests	Complete
AnalysisProgress, AnalysisRunCompletedEventArgs, BootstrapDiagnostics	Analyses overview	Analysis support classes	Progress and bootstrap-diagnostic tests	Complete
BatchAnalysisOptions, BatchAnalysisResult, BatchAnalysisRunner	Batch execution	Batch analysis support	Batch analysis tests	Partial

Estimation and Diagnostics

Public API	Technical treatment	Primary implementation	Verification evidence	Status
OptimizationMethod, CovarianceComputationStatus, MaximumLikelihood, MaximumAPosteriori	MLE and MAP chapters	Estimation classes and explicit covariance status	R bbmle profile, analytical posterior profile, covariance-status tests, and compiled workflows	Complete
GeneralizedMethodOfMoments, GMMIdentificationStatus, GMMEstimationStrategy, MomentConditionFunction, PointwiseMomentConditionFunction, JacobianFunction, PenaltyFunction	GMM chapter	GMM estimator	R gmm fit/specification/covariance and analytical sandwich reconstruction; compiled workflow	Complete for fixed-weight and efficient fit/covariance
BayesianAnalysis, SamplerType, PointEstimateType	Bayesian MCMC and model comparison	Bayesian analysis and pinned Numerics samplers	R loo and posterior parity; Bayesian/MCMC test sources; compiled workflow	PSIS, ARWMH adaptation, NUTS acceptance/gradient routes, live-sampler diagnostics, and rank-normalized R-hat/bulk-tail ESS verified
NumericalDiff	MLE, MAP, GMM, and leverage diagnostics	Numerical differentiation helper	Numerical-difference tests	Complete
InfluenceDiagnostics, ObservationInfluence, ParetoKCategory	Influence diagnostics	Influence diagnostics	R loo threshold parity, serialization tests, compiled workflow	Complete
LeverageDiagnostics, ObservationLeverage, PriorComponentLeverage	Influence diagnostics	Leverage diagnostics	Leverage unit tests; compiled workflow	Complete
PriorInfluenceDiagnostics, PriorComponentSummary	Influence diagnostics	Prior-influence diagnostics	Prior-influence tests; compiled workflow	Complete
PosteriorPredictiveCheck, PriorPredictiveCheck, PredictiveCheckResults, PredictiveSummary	Predictive checks	Predictive diagnostics	Predictive-check test sources; compiled workflow	Complete with the documented joint-prior limitation

Time Series and Rating Curves

Public API	Technical treatment	Primary implementation	Verification evidence	Status
AutoRegressive, ARAnalysis	Autoregressive chapter	AR model and analysis	Transform, likelihood, generation, recovery, and compiled workflow evidence	Complete
MovingAverage, MAAnalysis	Moving-average chapter	MA model and analysis	Transform, likelihood, generation, recovery, and compiled workflow evidence	Complete
ARIMA, ARIMAAnalysis, Transform	ARIMA and time-series chapters	ARIMA model and analysis	Independent recurrence, forecast, transform, MLE/Bayesian recovery, and compiled workflow evidence	Complete
ARIMAX, ARIMAXAnalysis, Trend, CovariateExtensionMethod	ARIMAX chapter	ARIMAX model and analysis	Date-indexed likelihood, forecast, generation, recovery, and compiled workflow evidence	Complete
RatingCurve, RatingCurveAnalysis	Rating-curve chapter	Rating-curve model and analysis	SciPy likelihood/optimum, analytical continuity, MLE/Bayesian recovery, and compiled workflow evidence	Complete

Bivariate and Spatial Models

Public API	Technical treatment	Primary implementation	Verification evidence	Status
BivariateDistribution, BivariateAnalysis	Bivariate distributions and analysis	Bivariate model and analysis	Independent copula optima, seven recovery cells, unit/integration sources, and compiled workflow	Complete; marginals are conditioned on rather than jointly estimated
CoincidentFrequencyAnalysis	Coincident frequency	Coincident-frequency analysis	Numerical-integration test sources; compiled workflow	Complete
SpatialGEV, SpatialGEVAnalysis, SpatialGEVUncertaintyMethod, SpatialDistanceMetric	Spatial GEV hierarchy, likelihood, fitting, distance metric, uncertainty methods, and results	Spatial model and analysis	Likelihood, kriging, and geodesic oracles; criteria, cross-validation, prediction, simulation, dispatch, and eight ten-site by 100-row recovery cells	Complete for the tested network sizes and dependence structures
SpatialGEVSiteResults, SpatialGEVCrossValidationResults, SpatialGEVCrossValidationFoldStatus, SpatialGEVBootstrapResults	Spatial result construction and leave-one-site-out validation	Spatial result types	Result DTO, fold-accounting, reduced-model, cross-validation, prediction, regional-posterior, and bootstrap sources	Complete
GaussianCopula, CachedMultivariateNormal, SpatialRegressionErrors	Spatial copula, Gaussian-process errors, and kriging	Spatial dependence support	Gaussian-copula, cached-MVN, R mvtnorm, and conditional-GP tests	Complete
ICorrelationModel, CorrelationFunctionType, BasicExponential, PoweredExponential, Spherical	Spatial correlation functions and bounds	Correlation implementations	Fixed-value correlation, haversine, and Cartesian distance tests	Complete

Trend and Link Functions

Public API	Technical treatment	Primary implementation	Verification evidence	Status
ITrendModel, TrendModelBase, TrendModelType	Trend functions	Trend support	Trend base tests	Complete
ConstantTrend, LinearTrend, QuadraticTrend, CubicTrend, ExponentialTrend, LogisticTrend, PowerTrend, ReciprocalTrend, SinusoidalTrend, StepFunction, GeneralLinearFunction	Trend functions	Trend implementations	Per-function unit tests	Complete
SESLink, ASinHLink, CenteredLink, LogASinHLink, LogSESLink, BestFitLinkFunctionFactory	Link functions	Link implementations	Link-function and serialization tests	Complete

Coverage Rule

The documentation test checks every exported scientific type against this matrix. Constructors and behavior-affecting members belong to their type's assigned chapter and are reviewed through its API table and compiled examples. Grouped result properties must state their units, nullability, and interpretation.

---

Back to Documentation Index