Definition of AUC?

[ch2ohch2oh]

When calculating AUC of ROC, most people use false positve and true positve as their axises. From my understanding, efficiency corresponds to true positve but purity does not match to 1 - false positive.

purity = true signals that passed the cut / events that passed the cut
1 - false positive = true backgrounds that failed the cut / true backgrounds

The consequence is that the ROC curve of efficiency and purity does not start and end at the diagonal points. Is my understanding correct?

[ch2ohch2oh]

OK I figured out you were using the integral of precision-recall curve. It would be nice if you can point this out in the comment of the source code. However, since precision-recall curve is more sensitive to imbalanced data, is there any particular reason not to use the AUC of false positive vs true positive?