roc_score_vec {SeaVal}R Documentation

ROC score (AUC)

Description

Calculates the area under curve (AUC) or ROC-score from a vector of probabilities and corresponding observations. Formula (1a) from Mason&2018 is used in the calculation, corresponding to trapezoidal interpolation. This is mostly an auxiliary function for the ROCS function, but also used in the ROC-diagram function, where the AUC is added to the diagrams.

Usage

roc_score_vec(probs, obs)

Arguments

probs

vector with probabilities (between 0 and 1)

obs

vector with categorical observations (as TRUE/FALSE)

Value

numeric. The ROC score.

Examples

roc_score_vec(probs = c(0.1,0.6,0.3,0.4),
             obs = c(FALSE,TRUE,TRUE,FALSE))

[Package SeaVal version 1.1.1 Index]