Skip to content

a question about the r package Epi

1 message · Yanwen Chen

#
Hi,
I have a question about the r package Epi.  I used the following code:
rc <- ROC(form = data$event ~ data$continuous, plot="ROC")
I used this code and want to find out the optimum cutoff of my continuous
variable.  It seemed it works.  I also got a ROC curve indicating the
corresponding sensitivity,  specificity, PV+ and PV-.  However, I found
that the sensitivity and specificity are correct.  But PV+ and PV- are
actually 1-PV+ and 1-PV-.
Is that what designed by the package?

Thank you.

Best,
Yanwen