Plot diagnostics for a hef object

plot method for class "hef".

# S3 method for hef
plot(
  x,
  y,
  ...,
  params = c("hyper", "ru", "pop"),
  which_pop = NULL,
  plot_type = NULL,
  one_plot = FALSE,
  add_legend = FALSE,
  legend_position = "topright",
  legend_text = NULL,
  num = 100
)

Arguments

x

an object of class "hef", a result of a call to ru.

y

Not used.

...

Additional arguments passed to plot.ru, hist or pairs. In particular, ru_scale = TRUE produces a plot using the parameterization used for ratio-of-uniforms sampling.

params

A character scalar that determines to which parameters the plots relate.

"hyper": the posterior sample of all hyperparameter values in \(\phi\) is plotted using plot.ru.
"ru": only the posterior sample generated using ru is plotted using plot.ru. This produces a different plot to params = "hyper" if ru is used only on a subset of \(\phi\). For example, this may be the case if x is the result of a call to hanova1. See vignette("bang-c-anova-vignette", package = "bang") for information.
"pop": posterior samples and/or densities of the population-specific parameter \(\theta\) are plotted. The population(s) included are determined by which_pop and the type of plot is determined by plot_type. If plot_type is not supplied then it is set to "dens".

which_pop

An integer vector or character scalar. If params = "pop" then which_pop indicates which populations to include in the plot. If which_pop is supplied then params is set to "pop". If which_pop = "all" then all populations are included. If there are many populations then this may fail if plot_type = "pairs" and/or one_plot = FALSE.

plot_type

A character scalar that determines the type of plot produced when params = "pop". If plot_type is supplied then params is set automatically to "pop".

"sim": histograms of the posterior samples of \(\theta\) for the populations in which_pop.
"dens": estimates of the marginal posterior densities of \(\theta\) for the populations in which_pop.
"both": both the histograms and estimated posterior densities.
"pairs": pairwise scatter plots of the posterior samples of \(\theta\) for the populations in which_pop, which must have length greater than one.

one_plot, add_legend, legend_position, legend_text

Only relevant if plot_type = "dens". If one_plot = TRUE then the estimated marginal posterior densities are plotted in the same graph and if add_legend = TRUE then a legend is added to this plot using legend in the position indicated by the character scalar legend_position. A character vector legend_text may be used to override the default legend text.

num

A numeric scalar. If plot_type == "dens" or plot_type == "both" then num gives the number of points at which the marginal densities are evaluated to produce plots.

Examples

See the examples in hef and hanova1.

Plot diagnostics for a hef object

Arguments

Examples

See also