R/SlowinskiGuyer.R

Defines functions diversity.contrast.test richness.yule.test mcconwaysims.test slowinskiguyer.test

Documented in diversity.contrast.test mcconwaysims.test richness.yule.test slowinskiguyer.test

## SlowinskiGuyer.R (2016-10-23)

##   Tests of Diversification Shifts with Sister-Clades

## Copyright 2011-2016 Emmanuel Paradis

## This file is part of the R-package `ape'.
## See the file ../COPYING for licensing issues.

slowinskiguyer.test <- function(x, detail = FALSE)
{
    r <- x[, 1]
    n <- x[, 1] + x[, 2]
    pp <- (n - r)/(n - 1)
    chi <- -2 * sum(log(pp))
    df <- as.integer(2 * length(pp))
    pval <- pchisq(chi, df, lower.tail = FALSE)
    res <- data.frame("chisq" = chi, "df" = df, "P.val" = pval, row.names = "")
    if (detail)
        res <- list(res, individual_Pvalues = pp)
    res
}

mcconwaysims.test <- function(x)
{
    LRTp <- function(x) {
        f <- function(x) ifelse(x == 0, 0, x * log(x))
        n1 <- x[1]
        n2 <- x[2]
        1.629*(f(n1 - 1) - f(n1) + f(n2 - 1) - f(n2) - f(2) -
               f(n1 + n2 - 2) + f(n1 + n2))
    }
    chi <- sum(apply(x, 1, LRTp))
    pval <- pchisq(chi, df <- nrow(x), lower.tail = FALSE)
    data.frame("chisq" = chi, "df" = df, "P.val" = pval, row.names = "")
}

richness.yule.test <- function(x, t)
{
    n1 <- x[, 1]
    n2 <- x[, 2]
    n <- c(n1, n2)
    tb <- c(t, t)

    .PrNt.Yule <- function(N, age, birth) {
        tmp <- -birth * age
        tmp + (N - 1) * log(1 - exp(tmp)) # on a log-scale
    }

    ## the functions to minimize:
    minusloglik0 <- function(l) -sum(.PrNt.Yule(n, tb, l))
    minusloglika <- function(l) -sum(.PrNt.Yule(n1, t, l[1])) - sum(.PrNt.Yule(n2, t, l[2]))

    ## initial values (moment estimators):
    ipa <- c(mean(log(n1)/t), mean(log(n2)/t))
    ip0 <- mean(ipa)

    out0 <- nlminb(ip0, minusloglik0, lower = 0, upper = 1)
    outa <- nlminb(ipa, minusloglika, lower = c(0, 0), upper = c(1, 1))
    chi <- 2 * (out0$objective - outa$objective)
    pval <- pchisq(chi, 1, lower.tail = FALSE)
    data.frame(chisq = chi, df = 1, P.val = pval, row.names = "")
}

diversity.contrast.test <-
    function(x, method = "ratiolog", alternative = "two.sided", nrep = 0, ...)
{
    method <- match.arg(method, c("ratiolog", "proportion",
                                  "difference", "logratio"))
    alternative <- match.arg(alternative, c("two.sided", "less", "greater"))

    minmax <- t(apply(x, 1, sort)) # sort all rows
    DIFF <- x[, 1] - x[, 2]
    SIGN <- sign(DIFF)

    CONTRAST <- switch(method,
                       "ratiolog" = {
                           if (any(minmax == 1))
                               minmax <- minmax + 1 # prevent division by 0
                           ## Note: if min = max, no need to set the contrast
                           ## to zero since this is done with sign()
                           log(minmax[, 2]) / log(minmax[, 1])
                       },
                       "proportion" = minmax[, 2] / (minmax[, 2] + minmax[, 1]),
                       "difference" = abs(DIFF),
                       "logratio" = log(minmax[, 1] / minmax[, 2]))

    y <- SIGN * CONTRAST # the signed contrasts

    if (nrep) {
        n <- length(SIGN)
        RND <-
            replicate(nrep,
                      sum(sample(c(-1, 1), size = n, replace = TRUE) * CONTRAST))
        cases <- switch(alternative,
                        "two.sided" = sum(abs(RND) > sum(y)),
                        "less" = sum(RND < sum(y)),
                        "greater" = sum(RND > sum(y)))
        cases/nrep
    } else wilcox.test(x = y, alternative = alternative, ...)$p.value
}

Try the ape package in your browser

Any scripts or data that you put into this service are public.

ape documentation built on March 31, 2023, 6:56 p.m.