[Vegan-commits] r1551 - in pkg/vegan: R inst

noreply at r-forge.r-project.org noreply at r-forge.r-project.org
Wed Mar 23 06:39:38 CET 2011


Author: jarioksa
Date: 2011-03-23 06:39:33 +0100 (Wed, 23 Mar 2011)
New Revision: 1551

Modified:
   pkg/vegan/R/anova.ccabyaxis.R
   pkg/vegan/inst/ChangeLog
Log:
prevent running anova.ccabyaxis in partial models because there is A BUG THAT NEEDS AN URGENT FIX

Modified: pkg/vegan/R/anova.ccabyaxis.R
===================================================================
--- pkg/vegan/R/anova.ccabyaxis.R	2011-03-22 13:54:07 UTC (rev 1550)
+++ pkg/vegan/R/anova.ccabyaxis.R	2011-03-23 05:39:33 UTC (rev 1551)
@@ -1,6 +1,10 @@
 "anova.ccabyaxis" <-
 function (object, cutoff = 1,  ...) 
 {
+    ## FIXME: wrong results in partial model
+    ## BUG -- NEEDS AN URGENT FIX
+    if (!is.null(object$pCCA))
+        stop("anova by 'axis' does not work in partial models: call the developers")
     cutoff <- cutoff + sqrt(.Machine$double.eps)
     rnk <- object$CCA$rank
     if (!max(rnk, 0)) 

Modified: pkg/vegan/inst/ChangeLog
===================================================================
--- pkg/vegan/inst/ChangeLog	2011-03-22 13:54:07 UTC (rev 1550)
+++ pkg/vegan/inst/ChangeLog	2011-03-23 05:39:33 UTC (rev 1551)
@@ -4,6 +4,12 @@
 
 Version 1.18-25 (opened March 10, 2011)
 
+	* anova.ccabyaxis: I (JO) noticed in an ORDNEWS message that
+	anova(..., by = "axis") drops the original conditions in partial
+	models and so gives wrong results. The first thing here prevents
+	the analysis, and a final fix brings along a new release (this
+	also breaks the tests).
+
 	* ordilabel: gained argument 'xpd' to draw labels outside the plot
 	region. 
 



More information about the Vegan-commits mailing list