plot.roc_band: Receiver Operating Characteristic (ROC) band

In sharp: Stability-enHanced Approaches using Resampling Procedures

Receiver Operating Characteristic (ROC) band

Description

Plots the True Positive Rate (TPR) as a function of the False Positive Rate (FPR) for different thresholds in predicted probabilities. If the results from multiple ROC analyses are provided (e.g. output of ExplanatoryPerformance with large K), the point-wise median is represented and flanked by a transparent band defined by point-wise quantiles. See examples in ROC and ExplanatoryPerformance.

Usage

## S3 method for class 'roc_band'
plot(
  x,
  col_band = NULL,
  alpha = 0.5,
  quantiles = c(0.05, 0.95),
  add = FALSE,
  ...
)

Arguments

x	output of ROC or ExplanatoryPerformance.
col_band	colour of the band defined by point-wise quantiles.
alpha	level of opacity for the band.
quantiles	point-wise quantiles of the performances defining the band.
add	logical indicating if the curve should be added to the current plot.
...	additional plotting arguments (see par).

Value

A base plot.

See Also

ROC, ExplanatoryPerformance

sharp documentation built on April 11, 2025, 5:44 p.m.