fit_OU: Fits an OU model based on a given configuration

In khabbazian/l1ou: Tools for detecting past changes in the expected mean trait values and studying trait evolution from comparative data

Fits an OU model based on a given configuration

Description

Fits an OU model based on a given configuration

Usage

fit_OU(tree, Y, shift.configuration, criterion = c("pBIC", "pBICess", "mBIC",
  "BIC", "AICc"), root.model = c("OUfixedRoot", "OUrandomRoot"),
  cr.regimes = NULL, alpha.starting.value = NA,
  alpha.upper = alpha_upper_bound(tree), alpha.lower = NA,
  l1ou.options = NA)

Arguments

tree	ultrametric tree of class phylo, with branch lengths, and edges in postorder.
Y	trait vector/matrix without missing entries. The row names of the data must be in the same order as the tip labels.
shift.configuration	shift positions, i.e. vector of indices of the edges where the shifts occur.
criterion	an information criterion (see Details).
root.model	model for the ancestral state at the root.
alpha.starting.value	optional starting value for the optimization of the phylogenetic adaptation rate.
alpha.upper	optional upper bound for the phylogenetic adaptation rate. The default value is log(2) over the minimum length of external branches, corresponding to a half life greater or equal to the minimum external branch length.
alpha.lower	optional lower bound for the phylogenetic adaptation rate.
l1ou.options	if provided, all the default values will be ignored.

Details

AICc gives the usual small-sample size modification of AIC. BIC gives the usual Bayesian information criterion, here penalizing each shift as 2 parameters. mBIC is the modified BIC proposed by Ho and Ané (2014). pBIC is the phylogenetic BIC for shifts proposed by Khabbazian et al. pBICess is a version of pBIC where the determinant term is replaced by a sum of the log of effective sample sizes (ESS), similar to the ESS proposed by Ané (2008).

Value

an object of class l1ou similar to estimate_shift_configuration.

References

Cécile Ané, 2008. "Analysis of comparative data with hierarchical autocorrelation". Annals of Applied Statistics 2(3):1078-1102.

Ho, L. S. T. and Ané, C. 2014. "Intrinsic inference difficulties for trait evolution with Ornstein-Uhlenbeck models". Methods in Ecology and Evolution. 5(11):1133-1146.

Mohammad Khabbazian, Ricardo Kriebel, Karl Rohe, and Cécile Ané (2016). "Fast and accurate detection of evolutionary shifts in Ornstein-Uhlenbeck models". Methods in Ecology and Evolution. doi:10.1111/2041-210X.12534

See Also

estimate_shift_configuration adjust_data

Examples

data(lizard.tree, lizard.traits)
lizard <- adjust_data(lizard.tree, lizard.traits[,1])
eModel <- estimate_shift_configuration(lizard$tree, lizard$Y)

### building l1ou object out of the second best score 
eModel2 = fit_OU(eModel$tree, eModel$Y, eModel$profile$configurations[[2]], 
                          l1ou.options=eModel$l1ou.options)
plot(eModel2)

### hypothesis testing

data("lizard.traits", "lizard.tree")
Y <- lizard.traits[,1:1]
tr  <- lizard.tree

tr <- multi2di(tr)
tr <- reorder(tr, "postorder")

### visualizing the tree with the edge indeces 
plot(tr)
edgelabels()

## place the shift position based on the hypothesis
shift.config <- c(116, 77)

hModel <- fit_OU(tr, Y, shift.config, criterion="AICc")
plot(hModel)
print(hModel)

khabbazian/l1ou documentation built on Aug. 10, 2022, 7:41 p.m.