API Reference

Main Classes

High-level experiment runners that compose the lower-level building blocks.

`AATest`	A class for conducting A/A tests with configurable parameters.
`ABTest`	A class for conducting A/B tests with configurable statistical tests and multiple testing correction.
`HomogeneityTest`	A class for conducting homogeneity tests between the groups.
`matching.Matching`	A class for performing matching analysis with configurable distance metrics and quality tests.

All comparators live in hypex.comparators.

Backend-adaptive tests — automatically select the best implementation for the active dataset backend (pandas vs. Spark).

`comparators.TTest`	Two-sample t-test for numeric targets.
`comparators.Chi2Test`	Chi-square test of independence for categorical targets.
`comparators.KSTest`	Two-sample Kolmogorov-Smirnov test for numeric targets.
`comparators.UTest`	Mann-Whitney U test (Wilcoxon rank-sum) for numeric targets.

`comparators.GroupDifference`
`comparators.GroupSizes`
`comparators.PSI`
`comparators.MahalanobisDistance`

`comparators.PowerTesting`
`comparators.MDEBySize`

Extend these to implement custom comparators.

`comparators.Comparator`
`comparators.StatHypothesisTesting`

`splitters.AASplitter`
`splitters.AASplitterWithStratification`

Pre-processing steps applied to Dataset objects.

`transformers.NaFiller`
`transformers.NanFilter`
`transformers.OutliersFilter`
`transformers.ConstFilter`
`transformers.CorrFilter`
`transformers.CVFilter`
`transformers.DummyEncoder`
`transformers.TypeCaster`
`transformers.CategoryAggregator`
`transformers.CUPEDTransformer`
`transformers.Shuffle`

Pipeline runners that chain executors over ExperimentData.

`experiments.Experiment`
`experiments.GroupExperiment`
`experiments.OnRoleExperiment`
`experiments.CycledExperiment`

`reporters.Reporter`
`reporters.DatasetReporter`
`reporters.DictReporter`
`reporters.HomoDatasetReporter`
`reporters.HomoDictReporter`

A calculator for estimating the minimum required sample size for multi-group comparisons.

`dataset.Dataset`
`dataset.ExperimentData`

Roles tag columns with semantic meaning so executors can locate them without hard-coding column names.

`dataset.TargetRole`
`dataset.TreatmentRole`
`dataset.GroupingRole`
`dataset.FeatureRole`
`dataset.InfoRole`
`dataset.PreTargetRole`
`dataset.StatisticRole`
`dataset.StratificationRole`
`dataset.FilterRole`
`dataset.AdditionalTargetRole`
`dataset.AdditionalFeatureRole`
`dataset.AdditionalGroupingRole`
`dataset.AdditionalMatchingRole`
`dataset.AdditionalTreatmentRole`