Plot P(useful) from Wynants 2018 (doi: 10.1002/sim.7653)

Usage

plot_superiority_prob(
  obj,
  strategies = NULL,
  type = c("best", "useful", "pairwise"),
  min_diff = 0,
  colors = NULL,
  labels = NULL,
  linewidth = 1.5
)

Arguments

obj: BayesDCA or BayesDCASurv object
strategies: Character vector with models or tests to plot. If null, compares either first two in obj$strategies or the first one against Treat all/none (if only one available).
type: One of "best", "useful", or "pairwise".
min_diff: Minimal difference for superiority. Defaults to zero. Used only for type = "pairwise"
colors: Named vector with color for each model or test. If provided for a subset of models or tests, only that subset will be plotted.
labels: Named vector with label for each model or test.
linewidth: Width of plotted lines.
plot_list: If TRUE, returns a list of separate ggplot objects.

Value

A ggplot object.

Examples

data(PredModelData)
fit <- dca(PredModelData)
plot_superiority_prob(fit)