Hello all.
I am a researcher interested in performing PGLS using Martin's correlation
structure. I am using a current version of both R and APE, but am having
some trouble with this analysis. If I use a starting alpha value of 1 with
the following code:
BR.diet<-corMartins(1,lemtre, fixed=F)
BR.gls<-gls(PC1~Fruit + Leaves + Seeds, correlation=BR.diet,
data=diet.shape.d)
I get the following error message:
Error in recalc.corStruct(object[[i]], conLin) :
NA/NaN/Inf in foreign function call (arg 4)
If I use corBrownian or corGrafen I do not get this error message. I tried
varying the starting alpha and setting "fixed" to "TRUE" for corMartins. All
of these permutations give me results, but they are all identical to one
another, regardless of staring alpha and regardless of whether fixed is T or
F. This does not seem correct to me and I am wondering whether this is a
known error with a correction?
I am not a subscriber to the r-sig-phylog listserve, so if anyone is able to
provide assistance, please email me at karen.b...@stonybrook.edu.
Thank you very much for your assistance.
Karen Baab
--
Karen L. Baab
Department of Anatomical Sciences
Stony Brook University Medical Center
Health Sciences Center
Stony Brook, NY 11794
Office: 631.444.9462
[[alternative HTML version deleted]]
_______________________________________________
R-sig-phylo mailing list
R-sig-phylo@r-project.org
https://stat.ethz.ch/mailman/listinfo/r-sig-phylo
