perm_dist: Permutation distribution

In floatofmath/adaperm: Permutation tests for adaptive designs

Description

Compute the permutation distribution given observations and a test statistic.

Usage

perm_dist(x1, x2, g1, g2, stat, B, x3 = NULL, g3 = NULL,
  restricted = TRUE, stratified = TRUE)

Arguments

x1	First stage data
x2	Second stage data
g1	(Dummy) first stage treatment assignments
g2	(Dummy) second stage treatment assignments (see Details)
stat	Function that computes the test statistic (signature function(x,g))
B	Number of permutations
x3	(Dummy) third stage data (e.g. sample size increase)
g3	(Dummy) third stage treatment assignments (see Details)
restricted	Should group sizes be considered fixed
stratified	Should permutation be stratified by stage

Details

First, second and third stage treatment assignments are only passed to define the second stage sample size and the sizes of the corresponding (treatment) groups. If the number of permutations is less than the number of all possible permutations the test observed test statistic will be added to the result such that $B+1$ values will be returned. If the number of requested permutations B is larger than the number of all possible permutations, only the latter will be used.

Value

numeric vector

Author(s)

Florian Klinglmueller

floatofmath/adaperm documentation built on May 16, 2019, 1:18 p.m.