Get the Area under the ROC curve to assess classifier performance using pROC
auc(prob, labels, method = c("auc_pairs", "pROC", "ROCR"),
verbose = FALSE)Float, Vector: Probabilities or model scores (e.g. c(.32, .75, .63), etc)
True labels of outcomes (e.g. c(0, 1, 1))
String: "pROC", "auc_pairs", or "ROCR": Method to use. Will use pROC::roc, auc_pairs,
ROCR::performance, respectively. They should all give the same result, they are included for peace of mind.
See Details
Logical: If TRUE, print messages to output
Consider looking at Balanced Accuracy and F1 as well
Important Note: We always assume that true labels are a factor where the first level is the "positive" case, aka the event. All methods used here, "pROC", "auc_pairs", "ROCR", have been setup to expect this. This goes against the default sertting for both "pROC" and "ROCR", which will not give an AUC less than .5 because they will reorder levels. We don't want this because we believe you CAN make a classifier perform worse than chance (for research or whatnot). It can be very confusing if levels are reordered automatically and different functions give you different AUC. Also, AUC has been very popular, but I strongly recommend reporting Balanced Accuracy instead.