An embedded and charset-unspecified text was scrubbed... Name: not available URL: <https://stat.ethz.ch/pipermail/r-help/attachments/20121219/46ff4a86/attachment.pl>
pROC and ROCR give different values for AUC
6 messages · Max Kuhn, David Winsemius, Ivana Cace +2 more
An embedded and charset-unspecified text was scrubbed... Name: not available URL: <https://stat.ethz.ch/pipermail/r-help/attachments/20121219/58d641d5/attachment.pl>
On Dec 19, 2012, at 4:47 AM, Ivana Cace wrote:
Packages pROC and ROCR both calculate/approximate the Area Under (Receiver Operator) Curve. However the results are different. I am computing a new variable as a predictor for a label. The new variable is a (non-linear) function of a set of input values, and I'm checking how different parameter settings contribute to prediction. All my settings are predictive, but some are better. The AUC i got with pROC was much lower then expected, so i tried ROCR. Here are some comparisons: AUC from pROC AUC from ROCR 0.49465 0.79311 0.49465 0.79349 0.49701 0.79446 0.49701 0.79764 When i draw the ROC (with pROC) i get the curve i expect. But why is the AUC according to pROC so different?
Why are you sending this to the Rhelp mailing list? You should instead be reporting it to the package maintainers.
maintainer('ROCR')
[1] "Tobias Sing <tobias.sing at mpi-sb.mpg.de>"
maintainer('pROC')
[1] "Xavier Robin <Xavier.Robin at unige.ch>"
David Winsemius Alameda, CA, USA
I considered that but on the list more people are likely to see it. So if they ran into the same thing they may already have figured out what is going on and have answers. Or if not, I'll make a reproducible example, ask the maintainers and post back here where others may find it. (but not till the end of the week)
Why are you sending this to the Rhelp mailing list? You should instead be reporting it to the package maintainers.
On Wed, Dec 19, 2012 at 12:36 PM, Ivana Cace <I.Cace at ati-a.nl> wrote:
I considered that but on the list more people are likely to see it. So if they ran into the same thing they may already have figured out what is going on and have answers. Or if not, I'll make a reproducible example, ask the maintainers and post back here where others may find it.
Even a reproducible example posted here, along w/ cc-ing the maintainers, would probably go a long way to get your more help. People are too busy to rig up toy examples to test, but given a toy example, it might be easy for someone to see where things are going south. -steve
Steve Lianoglou Graduate Student: Computational Systems Biology | Memorial Sloan-Kettering Cancer Center | Weill Medical College of Cornell University Contact Info: http://cbio.mskcc.org/~lianos/contact
An embedded and charset-unspecified text was scrubbed... Name: not available URL: <https://stat.ethz.ch/pipermail/r-help/attachments/20121220/7adb43c4/attachment.pl>