ROCplot: Simple ROC curve with optimal threshold and AUC approximation

Description Usage Arguments Details

Description

ROCplot(predictions, truth, N = 100, nulline = TRUE,...)

Usage

1
ROCplot(predictions, truth, N = 100, nullline = TRUE, ...)

Arguments

predictions

is prediction values between 0 and 1

truth

the actual 0 and 1 values

N

number of candidate thresholds; default is 100

...

additional arguments passed to plot

nulline

whether to plot a line with slope=1

Details

Plots a ROC curve. Also gives optimal cut point (several if they are equal in effectiveness) and area under the curve by the trapezoid method, so accuracy improves with greater N.


helophilus/ColsTools documentation built on May 30, 2019, 4:03 p.m.