[R-sig-phylo] "error in in vcv[1, k] <<- var" in compar.gee

Emmanuel Paradis Emmanuel.Paradis at mpl.ird.fr
Thu Feb 11 09:33:34 CET 2010


Adrian,

(writing privately for the moment; I'll update on the mailinglist later)

Adrian Jaeggi wrote on 10/02/2010 15:49:
> Dear all,
> 
> I'm trying to run an analysis on the correlated evolution of two 
> discrete (binary) traits. I've read the tree, and the data file, putting 
> the species names as row names and re-ordering them to match the tree:
> 
> tree<- read.nexus(tree)
> data<- read.table("data.txt", stringsAsFactors=FALSE)
> rownames(data)<- data$Genus_species
> data<- data[match(tree$tip.label,rownames(data)),]
> 
> However, when I then enter the formula I get an error message (original 
> in German, own translation):
> 
> gee<- compar.gee(y~x, data=data, family=binomial, phy=tree) ## y and x 
> are binary variables
> Fehler in vcv[l, k] <<- var : Indizierung außerhalb der Grenzen
> Error in vcv[1, k] <<- var : Subscript out of limits

The error happens when computing the VCV matrix from the tree. You may 
try vcv(tree) which should result in the same error. You can send your 
tree so I can see what's wrong: this may be a bug in vcv.phylo() or your 
tree could be wrongly built.

Best,

Emmanuel

> I also get the same error message when using family=Gaussian (which of 
> course you shouldn't when analyzing binary data) or when trying to do 
> the analysis using gls.
> 
> I couldn't figure out what the error message means, could anyone tell me 
> how to fix this?
> Thanks very much for your time and help!
> 
> Best,
> Adrian
> 
> 
> **************************************************************************
> 
> Adrian Jaeggi
> PhD Student
> dipl. anthropol./MSc. UZH
> Anthropological Institute & Museum
> University of Zurich, Irchel
> Winterthurerstr. 190
> 8057 Zurich, Switzerland
> Tel: ++41 (0) 44 635 54 33
> www.aim.uzh.ch
> 
> _______________________________________________
> R-sig-phylo mailing list
> R-sig-phylo at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-phylo
> 

-- 
Emmanuel Paradis
IRD, Montpellier, France
   ph: +33 (0)4 67 16 64 47
  fax: +33 (0)4 67 16 64 40
http://ape.mpl.ird.fr/



More information about the R-sig-phylo mailing list