hypothesis

const hypothesis: object

Defined in: ds/src/stats/index.js:60

Type Declaration

chiSquareTest

chiSquareTest: (observed, expected) => any = chiSquareTestFn

Chi-square goodness of fit test

Parameters

observed

number[]

Observed frequencies

expected

number[]

Expected frequencies

Returns

any

{statistic, pValue, df}

kruskalWallis

kruskalWallis: (groups) => any = kruskalWallisFn

Kruskal-Wallis H test Non-parametric alternative to one-way ANOVA

Parameters

groups

number[][]

Array of group samples

Returns

any

{statistic (H), pValue, df}

mannWhitneyU

mannWhitneyU: (sample1, sample2, options) => any = mannWhitneyUFn

Mann-Whitney U test (Wilcoxon rank-sum test) Non-parametric alternative to two-sample t-test

Parameters

sample1

number[]

First sample

sample2

number[]

Second sample

options?

any = {}

{alternative: ‘two-sided’|‘less’|‘greater’}

Returns

any

{statistic (U), pValue, alternative}

oneSampleTTest

oneSampleTTest: (sample, options) => any = oneSampleTTestFn

One-sample t-test

Parameters

sample

number[]

Sample data

options?

any = {}

{mu: hypothesized mean, alternative: ‘two-sided’|‘less’|‘greater’}

Returns

any

{statistic, pValue, df, mean, se}

oneWayAnova

oneWayAnova: (groups) => any = oneWayAnovaFn

One-way ANOVA

Parameters

groups

number[][]

Array of group samples

Returns

any

{statistic, pValue, dfBetween, dfWithin, MSbetween, MSwithin}

pairedTTest

pairedTTest: (sample1, sample2, options) => any = pairedTTestFn

Paired t-test for dependent samples

Parameters

sample1

number[]

First sample (before)

sample2

number[]

Second sample (after)

options?

any = {}

{mu: hypothesized mean difference (default 0), alternative: ‘two-sided’|‘less’|‘greater’}

Returns

any

{statistic, pValue, df, meanDiff, se}

tukeyHSD

tukeyHSD: (groups, options) => any = tukeyHSDFn

Tukey’s Honestly Significant Difference (HSD) test Post-hoc test for pairwise comparisons after ANOVA

Parameters

groups

number[][]

Array of group samples

options?

any = {}

{alpha: significance level (default 0.05), anovaResult: optional precomputed ANOVA result}

Returns

any

{ comparisons: Array of {groups: [i,j], diff, lowerCI, upperCI, pValue, significant}, groupMeans: Array of group means, groupNames: Array of group indices, MSwithin: Mean square within groups, dfWithin: Degrees of freedom, alpha: Significance level }

twoSampleTTest

twoSampleTTest: (sample1, sample2, options) => any = twoSampleTTestFn

Two-sample t-test (assuming equal variances)

Parameters

sample1

number[]

First sample

sample2

number[]

Second sample

options?

any = {}

{alternative: ‘two-sided’|‘less’|‘greater’}

Returns

any

{statistic, pValue, df, mean1, mean2, pooledSE}