likTraitPhylo: Log-likelihood estimation for traits and phylogenies
In motmot: Models of Trait Macroevolution on Trees

Description Usage Arguments Details Value Author(s) References Examples

This function calculates the log-likelihood and Brownian (co)variance for a trait(s) and a phylogeny using phylogenetically independent contrasts Note that as.rateMatrix calls the CAIC function vcv.array multiple times and this can be slow for large phylogenies (though faster than using the "ape" equivalent vcv.phylo).

1	likTraitPhylo(y, phy, covPIC = TRUE, brCov = NULL)

`y`	A matrix of trait data. Rownames must be included and match the taxon names in the phylogeny. Can accept single traits (calculates variance) or multiple traits (calculates variance-covariance matrix).
`phy`	An object of class "phylo" (see ape).
`covPIC`	Logical - allow for covariance between multivariate traits (`TRUE`), or assume not covariance (`FALSE`). Only applicable to multivariate traits
`brCov`	If `NULL` (the default), Brownian covariance is analytically estimated. If a user-supplied numerical value is suppied the likelihood is calculate given this value

The phylo object must be rooted and fully dichotomous

brownianVariance Brownian variance (or covariance for multiple traits) given the data and phylogeny

logLikelihood The log-likelihood of the model and data

Gavin Thomas, Rob Freckleton

Felsenstein J. 1973. Maximum-likelihood estimation of evolutionary trees from continuous characters. Am. J. Hum. Genet. 25, 471-492.

Felsenstein J. 1985. Phylogenies and the comparative method. American Naturalist 125, 1-15.

Freckleton RP & Jetz W. 2009. Space versus phylogeny: disentangling phylogenetic and spatial signals in comparative data. Proc. Roy. Soc. B 276, 21-30.

data(anolis.tree)
data(anolis.data)
## calculate Brownian variance log-likelihood of female SVL
female.svl <- matrix(anolis.data[,"Female_SVL"], 
dimnames=list(rownames(anolis.data)))
input.data <- sortTraitData(phy=anolis.tree, y=female.svl, log.trait=TRUE)
likTraitPhylo(phy = input.data$phy, y=input.data$trait)