Plot Profile-Likelihood Curves — plot.profile

Draw a one-dimensional profile-likelihood curve or a two-dimensional profile-likelihood surface.

Usage

# S3 method for class 'profile_likelihood'
plot(
  x,
  parameter = NULL,
  type = c("profile", "contour", "image"),
  line_col = "#C0392B",
  line_lwd = 2,
  cutoff_col = "#7F8C8D",
  mle_col = "#1F618D",
  point_col = "#C0392B",
  surface_cols = grDevices::hcl.colors(12, "YlOrRd", rev = TRUE),
  xlab = NULL,
  ylab = "Objective",
  main = NULL,
  ...
)

Arguments

x: An object returned by profile_likelihood.
parameter: Optional parameter name to plot. Defaults to the first available profile.
type: Plot type. Use "profile" for a one-dimensional curve, "contour" for a two-dimensional contour plot, or "image" for a two-dimensional image plot.
line_col: Line color for the profile curve.
line_lwd: Line width for the profile curve.
cutoff_col: Color used for the profile-likelihood cutoff line.
mle_col: Color used for the MLE reference line.
point_col: Color used for profiled grid points.
surface_cols: Colors used for two-dimensional image plots.
xlab: Optional x-axis label.
ylab: Optional y-axis label.
main: Plot title. When NULL, a default title is used.
...: Additional graphical arguments passed to plot.

Value

Invisibly returns the plotted profile data frame.

Details

For type = "profile", the x-axis is the fixed parameter value and the y-axis is the deviance-scale objective after re-optimizing all remaining parameters. The vertical reference line marks the fitted value and the horizontal reference line marks the profile-likelihood cutoff.

For type = "contour" or type = "image", the x- and y-axes are the two fixed parameter values and the contours show the same deviance-scale objective after re-optimizing all remaining parameters. The emphasized contour marks the approximate joint confidence-region cutoff for the pair.