kstats-hypothesis

Statistical hypothesis tests returning a unified TestResult with statistic, p-value, degrees of freedom, confidence interval, and isSignificant(alpha) helper.

Getting started

val sample = doubleArrayOf(5.0, 6.0, 7.0, 5.5, 6.5)
val result = tTest(sample, mu = 5.0)
result.statistic          // t-statistic
result.pValue             // p-value
result.confidenceInterval // 95% CI for the mean
result.isSignificant()    // true if p < 0.05

// Normality check before choosing a test
val sw = shapiroWilkTest(sample)
if (!sw.isSignificant()) { /* data is consistent with normality */ }

Available tests

Parametric:

• tTest(sample, mu, alternative, alpha) — one-sample t-test

• tTest(sample1, sample2, alternative, alpha) — two-sample t-test

• pairedTTest(sample1, sample2, alternative, alpha) — paired t-test

• oneWayAnova(vararg groups) — one-way ANOVA (returns AnovaResult)

Non-parametric:

• mannWhitneyUTest(x, y, alternative) — Mann-Whitney U test

• wilcoxonSignedRankTest(sample, mu, alternative) — Wilcoxon signed-rank test

• friedmanTest(vararg groups) — Friedman test

• kolmogorovSmirnovTest(sample, cdf) — one-sample Kolmogorov-Smirnov

• kolmogorovSmirnovTest(sample1, sample2) — two-sample Kolmogorov-Smirnov

Normality:

• shapiroWilkTest(sample) — Shapiro-Wilk test

• andersonDarlingTest(sample) — Anderson-Darling test

• dagostinoPearsonTest(sample) — D'Agostino-Pearson test

• jarqueBeraTest(sample) — Jarque-Bera test

Categorical:

• chiSquaredTest(observed, expected) — goodness-of-fit chi-squared

• chiSquaredIndependenceTest(contingencyTable) — chi-squared independence test

• gTest(observed, expected) — G-test (goodness-of-fit)

• gIndependenceTest(contingencyTable) — G-test (independence)

• fisherExactTest(a, b, c, d) — Fisher exact test (2x2)

• binomialTest(successes, trials, p, alternative) — binomial test

Homogeneity of variance:

• leveneTest(groups) — Levene test

• bartlettTest(vararg groups) — Bartlett test

• flignerKilleenTest(vararg groups) — Fligner-Killeen test

Result types

• TestResult — statistic, pValue, degreesOfFreedom, alternative, confidenceInterval, additionalInfo, isSignificant(alpha).

• AnovaResult — F-statistic, p-value, between/within sum of squares and degrees of freedom.

• Alternative — enum: TWO_SIDED, LESS, GREATER.