Lopez, Dan
2013-Dec-05 02:10 UTC
[R] pROC plot.roc - Plotting more than two curves - Is it possible
Hi R Experts, The info in this link suggests that plot in the pROC package can only compare two ROC curves at one time. https://stat.ethz.ch/pipermail/r-packages/2011/001220.html But this is from a couple of years ago. Curious if that has changed and if anyone knows how to plot more than one ROC curve using this package? If not can you please point me to good link/example on how to plot multiple ROC curves on one plot using another package? Thanks Dan Lopez Workforce Analyst LLNL [[alternative HTML version deleted]]
Murat Tasan
2013-Dec-05 02:18 UTC
[R] pROC plot.roc - Plotting more than two curves - Is it possible
have you looked at the ROCR package? there are at least a few ways to plot multiple ROC curves with the ROCR functions that come to mind. e.g. if you pass lists of 'scores' and 'labels' in to the prediction(...) method, the default plot(...) method will display all of the ROC (or precision-recall, or whichever measure you like) curves. another method is to run the prediction(...) and performance(...) functions separately for your cases, then use the plot(..., add TRUE) command. or you could even extract the x,y coordinates from the performance objects themselves and use lines(...) and/or points(...). cheers, -murat On Wed, Dec 4, 2013 at 9:10 PM, Lopez, Dan <lopez235 at llnl.gov> wrote:> Hi R Experts, > > The info in this link suggests that plot in the pROC package can only compare two ROC curves at one time. > https://stat.ethz.ch/pipermail/r-packages/2011/001220.html > > But this is from a couple of years ago. Curious if that has changed and if anyone knows how to plot more than one ROC curve using this package? > If not can you please point me to good link/example on how to plot multiple ROC curves on one plot using another package? > > Thanks > Dan Lopez > Workforce Analyst > LLNL > > > [[alternative HTML version deleted]] > > ______________________________________________ > R-help at r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code.