plot.ll.prof: plot likelihood profile for a pair of parameters

In dfeehan/mortfit: Fit and evaluate mortality models

Description

given a fitted mortalityModel, plot the profile of the log-likelihood function for a given pair of parameters. this can take a while if the likelihood takes a while to evaluate and/or the resolution is chosen to be large

Usage

plot.ll.prof(fit, theta.idx = c(1:2), delta = 0.1, res = 200,
  heatmap = FALSE)

Arguments

fit	a mortalityFit object
theta.idx	vector with the indexes of the two parameters to plot against each other; for example, in a four-parameter model, to examine the profile of the likelihood as parameters one and three vary, pass in theta.idx=c(1,3)
delta	window around the fitted max to profile. if delta is 0.1, then plot values for (0.9*max, 1.1*max), where max is the fitted maximum parameter value
res	the resolution, or size of the grid on which the likelihood is evaluated. large grids may take a long time
heatmap	if TRUE, return a heatmap for each pair of parameters; otherwise, return a contour plot

Value

a ggplot2 plots of the profile

dfeehan/mortfit documentation built on Nov. 14, 2020, 9:04 p.m.