perm.test {broman}

Permutation t-test

Description

Calculates a p-value for a t-test via permutations.

Usage

perm.test(x, y, n.perm = NULL, var.equal = TRUE, pval = TRUE)

Arguments

x

A numeric vector.

y

A second numeric vector.

n.perm

Number of permutations to perform. If NULL, all possible permutations are considered, and an exact p-value is calculated.

var.equal

A logical variable indicating whether to treat the two population variances as being equal.

pval

If TRUE, return just the p-value. If FALSE, return the actual permutation results (with the observed statistic as an attribute, "tobs").

Details

This calls the function stats::t.test() to calculate a t-statistic comparing the vectors x and y. Permutations are perfomed to give an exact or approximate conditional p-value.

Value

If pval=TRUE, the output is a single number: the P-value testing for a difference in the distributions of the populations from which x and y were drawn. If pval=FALSE, the output is a vector of the t statistics from the permutations. An attributed "tobs" contains the t statistic with the observed data.

See Also

stats::t.test(), paired.perm.test()

Examples

x <- c(43.3, 57.1, 35.0, 50.0, 38.2, 61.2)
y <- c(51.9, 95.1, 90.0, 49.7, 101.5, 74.1)
perm.test(x,y)