https://github.com/cran/ape
Tip revision: 6f7033d8327bcce6c8a279fe7fffe84bfb1252ca authored by Emmanuel Paradis on 14 February 2017, 18:15:21 UTC
version 4.1
version 4.1
Tip revision: 6f7033d
diversity.contrast.test.Rd
\name{diversity.contrast.test}
\alias{diversity.contrast.test}
\title{Diversity Contrast Test}
\description{
This function performs the diversity contrast test comparing pairs of
sister-clades.
}
\usage{
diversity.contrast.test(x, method = "ratiolog",
alternative = "two.sided", nrep = 0, ...)
}
\arguments{
\item{x}{a matrix or a data frame with at least two columns: the first
one gives the number of species in clades with a trait supposed to
increase or decrease diversification rate, and the second one the
number of species in the sister-clades without the trait. Each row
represents a pair of sister-clades.}
\item{method}{a character string specifying the kind of test:
\code{"ratiolog"} (default), \code{"proportion"},
\code{"difference"}, \code{"logratio"}, or any unambiguous
abbreviation of these.}
\item{alternative}{a character string defining the alternative
hypothesis: \code{"two.sided"} (default), \code{"less"},
\code{"greater"}, or any unambiguous abbreviation of these.}
\item{nrep}{the number of replications of the randomization test; by
default, a Wilcoxon test is done.}
\item{\dots}{arguments passed to the function
\code{\link[stats]{wilcox.test}}.}
}
\details{
If \code{method = "ratiolog"}, the test described in Barraclough et
al. (1996) is performed. If \code{method = "proportion"}, the version
in Barraclough et al. (1995) is used. If \code{method = "difference"},
the signed difference is used (Sargent 2004). If \code{method = "logratio"},
then this is Wiegmann et al.'s (1993) version. These
four tests are essentially different versions of the same test (Vamosi
and Vamosi 2005, Vamosi 2007). See Paradis (2012) for a comparison of
their statistical performance with other tests.
If \code{nrep = 0}, a Wilcoxon test is done on the species diversity
contrasts with the null hypothesis is that they are distributed around
zero. If \code{nrep > 0}, a randomization procedure is done where the
signs of the diversity contrasts are randomly chosen. This is used to
create a distribution of the test statistic which is compared with the
observed value (the sum of the diversity contrasts).
}
\value{
a single numeric value with the \emph{P}-value.
}
\references{
Barraclough, T. G., Harvey, P. H. and Nee, S. (1995) Sexual
selection and taxonomic diversity in passerine birds.
\emph{Proceedings of the Royal Society of London. Series B. Biological
Sciences}, \bold{259}, 211--215.
Barraclough, T. G., Harvey, P. H., and Nee, S. (1996) Rate of
\emph{rbc}L gene sequence evolution and species diversification in
flowering plants (angiosperms). \emph{Proceedings of the Royal Society
of London. Series B. Biological Sciences}, \bold{263}, 589--591.
Paradis, E. (2012) Shift in diversification in sister-clade
comparisons: a more powerful test. \emph{Evolution}, \bold{66},
288--295.
Sargent, R. D. (2004) Floral symmetry affects speciation rates in
angiosperms. \emph{Proceedings of the Royal Society of London. Series
B. Biological Sciences}, \bold{271}, 603--608.
Vamosi, S. M. (2007) Endless tests: guidelines for analysing non-nested
sister-group comparisons. An addendum. \emph{Evolutionary Ecology
Research}, \bold{9}, 717.
Vamosi, S. M. and Vamosi, J. C. (2005) Endless tests: guidelines for
analysing non-nested sister-group comparisons. \emph{Evolutionary
Ecology Research}, \bold{7}, 567--579.
Wiegmann, B., Mitter, C. and Farrell, B. 1993. Diversification of
carnivorous parasitic insects: extraordinary radiation or specialized
dead end? \emph{American Naturalist}, \bold{142}, 737--754.
}
\author{Emmanuel Paradis}
\seealso{
\code{\link{slowinskiguyer.test}}, \code{\link{mcconwaysims.test}}
\code{\link{richness.yule.test}}
}
\examples{
### data from Vamosi & Vamosi (2005):
fleshy <- c(1, 1, 1, 1, 1, 3, 3, 5, 9, 16, 33, 40, 50, 100, 216, 393, 850, 947,1700)
dry <- c(2, 64, 300, 89, 67, 4, 34, 10, 150, 35, 2, 60, 81, 1, 3, 1, 11, 1, 18)
x <- cbind(fleshy, dry)
diversity.contrast.test(x)
diversity.contrast.test(x, alt = "g")
diversity.contrast.test(x, alt = "g", nrep = 1e4)
slowinskiguyer.test(x)
mcconwaysims.test(x)
}
\keyword{htest}