ROC: Fast Computation of the Receiver Operating Characteristic...

In wush978/FastROC: Fast computation of ROC

Description

Fast Computation of the Receiver Operating Characteristic Curve (ROC)

Usage

  ROC(y.true, y.estimate, resize = NULL)

Arguments

y.true	logical vector. The answer of the classification problem.
y.estimate	numeric vector. The ranking of the predicted result corresponding to Ranswer.
resize	integer vector or NULL(default). If is.null(resize), the complete ROC curve is returned. Otherwise, the number of returned points is down sampled to resize.

Value

list(x, y). The x-coordinate and y-coordinate of the ROC curve.

Examples

n <- 10^5
pred.x <- rnorm(n)
true.x <- runif(n) < 1 / (1 + exp(-pred.x))
roc <- ROC(true.x, pred.x)
auc <- AUC(roc$x, roc$y)

wush978/FastROC documentation built on May 4, 2019, 12:01 p.m.