API Reference¶
The SeroEpi API is organized into three primary sections:
seroepi.dataset & variables— Immutable isolate container (Dataset), persistent reference assets (LoadedData,CountryData,KlebnetNeonatalSepsis), semantic variable hierarchy (Variable,Trait,Spatial,Temporal,Target,Criterion), sparse pairwise genomic distances (GenomicDistances), asynchronous API clients (PathogenwatchClient,WorldBankClient), and tabular/distance parsers (seroepi.io).seroepi.aggregate & estimators— Stratified count aggregation (Aggregator), estimator base contracts (BaseEstimator,StratifiedResultMixin), analytical and bootstrap estimators (UnpooledEstimator, [BootstrapEstimator][seroepi.estimators.unpooled.BootstrapEstimator]), JAX/NumPyro hierarchical Bayesian partial-pooling models (HierarchicalMCMCEstimator,HierarchicalSVIEstimator), clonal outbreak deduplication (TransmissionEstimator), and LOSO-validated multi-valent formulation optimization & differential prevalence testing (Formulation,FormulationComparison).seroepi.plotting & app— Theme-aware Altair visualizers (BasePlotter,BarPlotter,ChoroplethPlotter,HeatmapPlotter,PyramidPlotter,UpSetPlotter,FormulationCoveragePlotter,FormulationStabilityPlotter,DifferentialPrevalencePlotter,RankShiftPlotter) and the modular OOP PyShiny web application (seroepi.app).