fpr: Calculates the False Positive Rate (FPR)

In ksboxer/CDIPATools: CDI Predictive Analytics Tools

Description

This function takes the predictions of a model, (can be either binary 0 or 1, or continous numeric [0,1]) and calculates the False Positive Rate. Given that predictions need to be binary for the FPR to be calculated you need to pass in a threshold to cut the predictions off. If the predictions are already binary, then pass in .5 FPR = FP / (FP + TN)

Usage

fpr(predictions, outcomes, threshold)

Arguments

predictions	list of numerics, predicted values
outcomes	list of numerics, actual values/outcomes
threshold	numeric, value between 0 - 1 to cut predictions that are continous within binary 0s and 1s

Value

numeric, false positive rate

Examples

fpr(predictions = FakePredictionResults$est.risk.score,
outcomes = FakePredictionResults$true.risk.bin, threshold = .5)

ksboxer/CDIPATools documentation built on June 5, 2019, 8:29 a.m.