Statistics Package

daspi.statistics

Statistics package — confidence intervals, hypothesis tests, estimators, and Monte Carlo simulation.

This package bundles four complementary submodules that cover the full statistical workflow from raw samples to process-capability reporting:

confidence Two-sided confidence interval functions for mean, median, variance, standard deviation, proportions, Cp/Cpk, and regression predictions. All functions return (point_estimate, lower, upper) tuples.

hypothesis Hypothesis tests for normality (Anderson-Darling, KS), variance equality (F-test, Levene), location (t-test, Mann-Whitney U), proportions, and distribution shape (skewness, kurtosis). Every function returns (p_value, statistic, ...).

estimation High-level estimator classes (LocationDispersionEstimator, DistributionEstimator, ProcessEstimator, GageEstimator) and standalone helpers for kernel density estimation, LOESS/LOWESS smoothing, and GUM measurement-uncertainty propagation.

montecarlo Data structures for encoding engineering specifications (SpecLimits, Specification) and classes for Monte Carlo process simulation (RandomProcessValue, Binning).

All public names from each submodule are re-exported at the package level, so from daspi.statistics import mean_ci works without knowing which submodule contains it.

Module reference

Module	Contents
Hypothesis	Normality, variance, location, proportion, and shape tests
Confidence	CIs for mean, variance, Cp/Cpk, proportions, regression
Estimation	LocationDispersionEstimator, ProcessEstimator, GageEstimator, LOESS
Monte Carlo	SpecLimits, Specification, RandomProcessValue, Binning