Method for Profiling cauphyfit Objects

Source: R/fitCauchy.R

Investigates the profile log-likelihood function for a fitted model of class cauphyfit.

# S3 method for class 'cauphyfit'
profile(fitted, which = 1:npar, level = 0.8, npoints = 100, ...)

Arguments

fitted

the cauphyfit fitted model object.

which

the original model parameters which should be profiled. This can be a numeric or character vector. By default, all parameters are profiled.

level

highest confidence level for parameters intervals, computed using the approximated Hessian (see compute_vcov).

npoints

number of points to profile the likelihood for each parameter.

...

further arguments passed to or from other methods.

Value

An object of class profile.cauphyfit, which is a list with an element for each parameter being profiled. The elements are data-frames with two variables:

par.vals:

a matrix of parameter values for each fitted model.

profLogLik:

the profile log likelihood.

Details

This function computes a confidence interval for the parameters using confint.cauphyfit, and then computes the likelihood function on a grid with npoints values evenly spaced between the bounds of the interval, for each parameter one by one, all other parameters being fixed.

Examples

phy <- ape::rphylo(5, 0.1, 0)
dat <- rTraitCauchy(n = 1, phy = phy, model = "cauchy", parameters = list(root.value = 0, disp = 1))
fit <- fitCauchy(phy, dat, model = "cauchy", method = "reml")
pr <- profile(fit)
plot(pr)