bayou.checkModel: Function for checking parameter lists, prior and models are...

In uyedaj/bayou: Bayesian Fitting of Ornstein-Uhlenbeck Models to Phylogenies

Function for checking parameter lists, prior and models are consistent and error-free

Description

Function for checking parameter lists, prior and models are consistent and error-free

Usage

bayou.checkModel(
  pars = NULL,
  tree,
  dat,
  pred = NULL,
  SE = 0,
  prior,
  model = "OU",
  autofix = TRUE
)

Arguments

pars	A list of parameters that will be specified as starting parameter
tree	An object of class “phylo”
dat	A named data vector that matches the tip lables in the provided tree
pred	A matrix or data frame with named columns with predictor data represented in the specified formula
SE	The standard error of the data. Either a single value applied to all the data, or a vector of length(dat).
prior	A prior function made using make.prior
model	Either one of c("OU", "QG" or "OUrepar") or a list specifying the model to be used.
autofix	A logical that indicates whether certain errors should be automatically fixed.

Details

A series of checks are performed, run internally within bayou.makeMCMC, but can also be run on provided inputs prior to this. Errors are reported.

If autofix == TRUE, then the following errors will be automatically corrected:

Branch lengths == 0; any branches of length 0 will be given length .Machine$double.eps is.binary(tree) == FALSE; runs multi2di pars do not match prior$fixed; parameters are resimulated from prior

Value

A list of results of the checks and if 'autofix==TRUE', then ..$autofixed returns a list of all the input elements, with corrections.

uyedaj/bayou documentation built on Feb. 22, 2025, 8:31 p.m.