plot.rocdata: plot.rocdata

In AndrewLawrence/dCVnet: doubly cross-validated elastic-net regularised generalised linear models

plot.rocdata

Description

acts on a rocdata object to provide a ROC plot with the following:

• y-axis : Sensitivity (i.e. True Positive Rate)

• x-axis : 1 - Specificity (i.e. False Positive Rate)

• curve : shows c.d.f. as classification threshold varies

Usage

## S3 method for class 'rocdata'
plot(
  x,
  legend = TRUE,
  alphalabel = c(0.25, 0.5, 0.75),
  guide_labels = c(group = "Model", threshold = expression(P[Threshold]), refline =
    "Chance\nPerformance"),
  ...
)

Arguments

x	rocdata object
legend	logical. Display a legend?
alphalabel	should certain alpha values (probability thresholds) on the curve be highlighted with symbols indicating threshold?
guide_labels	a named list of 3 labels used in the legend: group = label for the Grouping factor threshold = label for the Threshold factor refline = label for the Reference line
...	additional arguments (unused)

Details

requires ggplot2.

Value

a ROC plot, as above.

AndrewLawrence/dCVnet documentation built on Sept. 24, 2024, 5:24 a.m.