## caracal for
## http://stats.stackexchange.com/questions/6127/which-permutation-test-implementation-in-r-to-use-instead-of-t-tests-paired-and
## paired data (two dependent groups)
x1 <- c(99, 99.5, 65, 100, 99, 99.5, 99, 99.5, 99.5, 57, 100, 99.5,
        99.5, 99, 99, 99.5, 89.5, 99.5, 100, 99.5)
y1 <- c(99, 99.5, 99.5, 0, 50, 100, 99.5, 99.5, 0, 99.5, 99.5, 90,
        80, 0, 99, 0, 74.5, 0, 100, 49.5)

## shorten data to something where we can do a complete permutation test
Nexact <- 12
x1 <- x1[1:Nexact]
y1 <- y1[1:Nexact]
DV <- c(x1, y1)
IV <- factor(rep(c("A", "B"), c(length(x1), length(y1))))
id <- factor(rep(1:length(x1), 2))

library(lmPerm)                        # for aovp()
summary(aovp(DV ~ IV + Error(id/IV)))  # p-value is far off from correct, see below

library(coin)                          # for oneway_test()
oneway_test(DV ~ IV | id, alternative="two.sided", distribution=approximate(B=9999))  # close to correct

## exact test (two-sided) for correct permutation p-value
DVd    <- y1-x1
sgnLst <- lapply(numeric(Nexact), function(x) { c(-1, 1) } )
sgnMat <- as.matrix(expand.grid(sgnLst))
getMD  <- function(x) { mean(abs(DVd) * x) }
resMD  <- apply(sgnMat, 1, getMD)
(pVal  <- sum(abs(resMD) >= abs(mean(DVd))) / length(resMD))