plot_class_roc: Receiver operating characteristic plots for outcome classes.

In PiotrTymoszuk/caretExtra: Extra functionality for predictions and quality control of caret models

Receiver operating characteristic plots for outcome classes.

Description

Plots receiver-operating characteristic (ROC) plots for outcome classes. ROC metrics for particular classes are obtained in an one versus rest comparison.

Usage

plot_class_roc(
  predx_object,
  one_plot = FALSE,
  palette = NULL,
  line_size = 0.5,
  point_size = 0.3,
  plot_title = NULL,
  plot_subtitle = NULL,
  plot_tag = NULL,
  show_annotation = TRUE,
  annotation_size = 2.75,
  annotation_x = 0.6,
  annotation_y = 0.3,
  annotation_offset = 0.1,
  annotation_hjust = 0,
  annotation_vjust = 1,
  cust_theme = NULL,
  ...
)

Arguments

predx_object	a predx class object.
one_plot	logical: should all ROC curves be presented in one plot?
palette	a character vector with color names of hex codes. If NULL, the default ggplot2 palette is used.
line_size	width of the ROC curve lines.
point_size	size of the plot point.
plot_title	plot title.
plot_subtitle	plot subtitle.
plot_tag	plot tag.
show_annotation	logical: should sensitivity and specificity values be displayed in the plot?
annotation_size	size of the annotation text and of the cutoff label.
annotation_x	annotation x position.
annotation_y	annotation y position.
annotation_offset	line spacing in the annotation text. Applies only if one_plot = TRUE.
annotation_hjust	horizontal justification of the annotation text.
annotation_vjust	horizontal justification of the annotation text.
cust_theme	custom ggplot theme. If NULL, the default ROC theme of the plotROC is used.
...	extra arguments passed to geom_roc.

Details

For regression, NULL is returned with a warning.

Value

if one_plot = TRUE a single ggplot object. Otherwise, a list of ggplot objects.

PiotrTymoszuk/caretExtra documentation built on Oct. 15, 2023, 10:03 p.m.