Building the Plot for nextBest-NextBestNCRMLoss Method.

Helper function which creates the plot for nextBest-NextBestNCRMLoss() method.

Usage

h_next_best_ncrm_loss_plot(
  prob_mat,
  posterior_loss,
  max_overdose_prob,
  dose_grid,
  max_eligible_dose_level,
  doselimit,
  next_dose,
  is_unacceptable_specified,
  prob_plot_type = c("lollipop", "bar"),
  dose_scale = c("linear", "log", "factor"),
  axis_ticks = c("dosegrid", "regular"),
  axis_text_angle = ifelse(match.arg(axis_ticks) == "dosegrid", 45, 0),
  safe_dose_marker = c("line", "triangle")
)

Arguments

prob_mat

(numeric)
matrix with probabilities of a grid doses to be in a given interval. If is_unacceptable_specified is TRUE, there must be 4 intervals (columns) in prob_mat: underdosing, target, excessive, unacceptable. Otherwise, there must be 3 intervals (columns): underdosing, target, overdose. Number of rows must be equal to number of doses in a grid.

posterior_loss

(numeric)
posterior losses.

max_overdose_prob

(number)
maximum overdose posterior probability that is allowed.

dose_grid

(numeric)
dose grid.

max_eligible_dose_level

(number)
maximum eligible dose level in the dose_grid.

doselimit

(number)
the maximum allowed next dose.

next_dose

(number)
next best dose.

is_unacceptable_specified

(flag)
is unacceptable interval specified?

prob_plot_type

(string)
probability geometry, either "lollipop" or "bar".

dose_scale

(string)
dose-axis scale: "linear", "log", or "factor" for equally spaced dose levels. The log scale requires all doses to be strictly positive.

axis_ticks

(string)
x-axis tick positions, either at each dose-grid value ("dosegrid", the default) or at regular positions selected by ggplot2 ("regular").

axis_text_angle

(number)
rotation angle for x-axis tick labels. Defaults to 45 degrees for axis_ticks = "dosegrid" and 0 degrees for axis_ticks = "regular".

safe_dose_marker

(string)
mark the maximum dose passing the overdose criterion with a red dashed "line" (default) or a red "triangle" above its probability geometry.