MetricsWeighted (version 0.3.0)

AUC: Area under the ROC

Description

Function copied from glmnet package (modified to ensure deterministic results). Calculates weighted AUC, i.e. the area under the receiver operating curve. The larger, the better.

Usage

AUC(actual, predicted, w = NULL, ...)

Arguments

actual

Observed values (0 or 1).

predicted

Predicted values of any value (not necessarly between 0 and 1).

w

Optional case weights.

...

Further arguments passed by other methods.

Value

A numeric vector of length one.

Details

The unweighted version can be different from the weighted one with unit weights due to ties in predicted.

See Also

gini_coefficient.

Examples

# NOT RUN {
AUC(c(0, 0, 1, 1), c(0.1, 0.1, 0.9, 0.8))
AUC(c(1, 0, 0, 1), c(0.1, 0.1, 0.9, 0.8))
AUC(c(1, 0, 0, 1), 2 * c(0.1, 0.1, 0.9, 0.8))
AUC(c(1, 0, 0, 1), c(0.1, 0.1, 0.9, 0.8), w = rep(1, 4)) # different from last due to ties
AUC(c(1, 0, 0, 1), c(0.1, 0.2, 0.9, 0.8))
AUC(c(1, 0, 0, 1), c(0.1, 0.2, 0.9, 0.8), w = rep(1, 4)) # same as last (no ties)
AUC(c(0, 0, 1, 1), c(0.1, 0.1, 0.9, 0.8), w = 1:4)
# }

Run the code above in your browser using DataCamp Workspace