Plot likelihood surface — plot_likelihood

Create a 2D contour plot of the log-likelihood surface for two-parameter models. Useful for visualizing parameter identifiability and correlation.

Usage

plot_likelihood_surface(
  spec,
  y,
  par,
  which_par = NULL,
  n_grid = 50,
  range_mult = 2,
  type = c("contour", "filled", "both"),
  show_mle = TRUE,
  ...
)

Arguments

spec: A model_spec object with exactly 2 parameters
y: Numeric vector of observed data
par: Named numeric vector of parameter values (used as center point)
which_par: Optional character vector of length 2 specifying which parameters to plot (for models with >2 parameters)
n_grid: Number of grid points per dimension (default 50)
range_mult: Multiplier for parameter range around center (default 2)
type: Plot type: "contour" (default), "filled", or "both"
show_mle: Logical; show MLE point (default TRUE)
...: Additional arguments passed to plotting functions

Value

Invisible list with grid coordinates and likelihood values

Examples

spec <- model_spec_normal()
y <- rnorm(100, mean = 5, sd = 2)
plot_likelihood_surface(spec, y, par = c(mu = 5, sigma = 2))
plot_likelihood_surface(spec, y, par = c(mu = 5, sigma = 2), type = "filled")