[Analogue-commits] r159 - in pkg: R inst

[noreply at r-forge.r-project.org]

Author: gsimpson
Date: 2010-03-02 09:30:14 +0100 (Tue, 02 Mar 2010)
New Revision: 159

Modified:
   pkg/R/roc.R
   pkg/inst/ChangeLog
Log:
fix bug in generation of points on ROC curve where we evaluate other statistics

Modified: pkg/R/roc.R
===================================================================
--- pkg/R/roc.R	2010-01-22 19:02:48 UTC (rev 158)
+++ pkg/R/roc.R	2010-03-02 08:30:14 UTC (rev 159)
@@ -32,7 +32,8 @@
         TPF <- cumsum(tab[, 2])/sum(tab[, 2])
         FPE <- cumsum(tab[, 1])/sum(tab[, 1])
         roc.values <- TPF - FPE
-        roc.points <- rev(sort(c(IN, OUT)))
+        ##roc.points <- rev(sort(c(IN, OUT)))
+        roc.points <- rev(sort(as.numeric(dimnames(tab)[[1]])))
         optimal <- as.numeric(names(max.roc <- which.max(roc.values)))
         names(FPE) <- names(TPF) <- names(roc.values) <- NULL
         wilcox <- wilcox.test(IN, OUT, conf.int = FALSE)

Modified: pkg/inst/ChangeLog
===================================================================
--- pkg/inst/ChangeLog	2010-01-22 19:02:48 UTC (rev 158)
+++ pkg/inst/ChangeLog	2010-03-02 08:30:14 UTC (rev 159)
@@ -14,6 +14,11 @@
 
 	* fixUpTol: was inappropriately matching one of it's arguments.
 
+	* roc: In some circumstances, the generation of the points at which
+	the ROC curve was evaluated resulted in more points than the other
+	statistics. Fixed to use the points established by 'table', used to
+	generate these other statistics. This affected the plot method.
+
 Version 0.6-22
 
 	* tran.formula: bug in my implementation of the standard