roc_curve: Receiver Operating Characteristic (ROC)

In nikitagusarov/performancer: Performance Metrics for Discrete Choice Analysis

Receiver Operating Characteristic (ROC)

Description

Get Receiver Operating Characteristic (ROC) curve construction. An ROC curve is a plot in which the horizontal axis (the x axis) denotes the false-positive rate. The vertical axis (the y axis) denotes the true-positive rate of a classifier. ROC analysis studies the relationship between the sensitivity and the specificity of the classifier. It makes sence in the context of binary classification.

Usage

roc_curve(
  y_real,
  y_predicted,
  threshold_range = c(0, 1),
  threshold_step = 0.01
)

Arguments

y_real	Observed values to compare with in matrix format.
y_predicted	Predicte values (probabiblities by class).
threshold_range	Decision boundary threshold range. Defaults to 'c(0, 1)' for basic probabilistic classifier, but other values may be used.
threshold_step	Step to use over the boundary threshold range. Defaults to 0.01 for binary classification.

Value

matrix of ROC curve coordinates for given probabilistic classifier output.

nikitagusarov/performancer documentation built on Jan. 12, 2023, 12:19 a.m.