This package was developed to assess the adequacy of continuous trait models. The package includes functions to parse the output of various model-fitting functions in R, rescale phylogenies according to fitted parameters, simulate datasets, calculate summary statistics and plot results. The methods is described in a our paper: "Model adequacy and the macroevolution of angiosperm functional traits" published in The American Naturalist (see link; pdf; and manuscript repo).
To install the package from github, the easiest way is to install using devtools. Install devtools
, then type
devtools::install_github("pennell-lab-ubc/arbutus")
The package depends on the ape
library
The idea behind arbutus
is that models are fit to comparative data using other packages and arbutus
is used to assess the adequacy of the fitted model. There are no native functions for fitting models or performing model selection within this package. Currently, arbutus can handle objects of the following classes:
gfit
objects from fitContinuous
in geiger (>= 2.0)fit.mle
objects from find.mle
in diversitree (>= 0.9.7)mcmcsamples
objects from mcmc
in diversitree (>= 0.9.7)`gls
objects from gls
in nlme
pgls
objects from pgls
in caper
phylolm
objects from phylolm
in phylolm
phylo
objects in ape
format (does not rescale phylogeny)multiPhylo
objects in ape
format (does not rescale phylogeny)OUwie
objects from OUwie
in OUwie (>= 2.6)We plan to add more package interfaces as this project progresses.