bhatt.test: Bhattacharryya test for comparing two samples.

In alR: Arc Lengths of Statistical Functions

Description

Use the multinomial distribution to test the hypothesis that two samples come from the same distribution.

Usage

bhatt.test(y, x, k)

Arguments

y, x	Samples to be compared.
k	Number of proportions.

Details

It is assumed that the two samples come from the same kernel density distribution. The support of the KDE of the first sample is divided into k equally spaced quantiles, and then compared to the corresponding proportions of the second.

Value

bat.test: A list with the following components:

• df=2*k: where k is the number of proportions used.

• y.prop, x.prop: Vectors of proportions.

• D2: Measure of divergence between samples.

• test.stat: The test statistic for the Bhattacharrayya test.

• p.value: The p-value of the test.

Examples

y <- rnorm(1000)
x <- rnorm(1000)
bhatt.test(y,x,10)

alR documentation built on Dec. 7, 2017, 5:03 p.m.