This function estimates the best fit for an evolutionary parameter (α, λ, or g) in a phylogenetic generalised least squares (PGLS) model, using log-likelihood.
fitEvolPar(dat, tre, mod=c("OU", "lambda","EB"))
dat a data frame containing the two variables (predictive and response) to include in a PGLS analysis
tre a phylogenetic tree (object of class "phylo"); tre needs not be ultrametric
mod the phylogenetic correlation structure (covariance matrix) for the evolutionary model of the PGLS
This function is designed to provide the best estimate of the parameter of choice if optim fails to return its value when compiling it in a gls object. In such a case, the value for it needs to be fixed (fixed = TRUE), which requires an input for the parameter's best fit.
Function will stop if the data contains missing values and/or if the tips of the tree do not match the row names in the data frame. Data frame dat needs to have two columns: the predictive variable (x) on the left, and the response one (y) on the right.
Correlations structures defined by the user with mod can be: "OU"(Ornstein-Uhlenbeck – Martins and Hansen, 1997; Butler and King, 2004), "lambda" (Brownian Motion with a maximum likelihood estimate of Pagel's λ – Freckleton et al., 2002), or "EB" (Early Burst – Blomberg et al., 2003). PGLS models are fitted using gls in nlme with correlation structures from ape (see ?corClasses).
A parameter estimate for the model of choice (α for "OU", Pagel's λ for "lambda", or g for "EB"). Value is comprised between 0 and 1 (0.1 and 1 for g), with one significant figure.
Lucas Legendre lucasjlegendre@gmail.com
This function was adapted from a piece of code written by Emmanuel Paradis on the R-sig-phylo mailing list (https://stat.ethz.ch/mailman/listinfo/r-sig-phylo).
Blomberg, S. P., Garland, T. and Ives, A. R. (2003) Testing for phylogenetic signal in comparative data: Behavioral traits are more labile. Evolution, 57, 717–745.
Butler, M. A. and King, A. A. (2004) Phylogenetic comparative analysis: A modeling approach for adaptive evolution. The American Naturalist, 164, 683–695.
Freckleton, R. P., Harvey, P. H. and Pagel, M. (2002) Phylogenetic analysis and comparative data: A test and review of evidence. The American Naturalist, 160, 712–726.
Martins, E. P. and Hansen, T. F. (1997) Phylogenies and the comparative method: a general approach to incorporating phylogenetic information into the analysis of interspecific data. The American Naturalist, 149, 646–667.
library(nlme)
library(ape)
library(evobiR)
library(phytools)
# Load the data
data(anole.data)
data(anoletree)
# Reorganize the data into 'dat'
dat<-cbind.data.frame(anole.data$SVL, anole.data$HL)
colnames(dat)<-c("SVL", "HL"); rownames(dat)<-rownames(anole.data)
# Estimate individual parameter for a lambda model
fitEvolPar(dat, anoletree, "lambda")
# Use in a 'gls' object with an OU model
reg<-gls(HL~SVL, anole.data, correlation=corMartins(fitEvolPar(dat, anoletree, "OU"), phy = anoletree, fixed = TRUE, form=~spp))
summary(reg)