Date: Tue, 26 Oct 1999 08:06:57 +1000
Reply-To: Peter Baade <Peter_Baade@HEALTH.QLD.GOV.AU>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: Peter Baade <Peter_Baade@HEALTH.QLD.GOV.AU>
Subject: ROC curves -Reply
Content-Type: text/plain; charset=US-ASCII
Use the option outroc= in proc logistic. This gives the data necessary to generate a graph. The actual value of the area under the ROC curve is given by the "c" statistic in the standard logistic regression output (under "Association of predicted probabilities and observed responses). Not sure if this was what you were after, but I hope it helps.
Peter.
>>> wcc <wcc@HA.MC.NTU.EDU.TW> 26/October/1999 02:26am >>>
How to plot and evaluate ROC curve by SAS?
|