[R-sig-phylo] yule.cov

Santiago Claramunt sclara1 at tigers.lsu.edu
Wed Jun 3 19:46:14 CEST 2009


Hi all,

I'm using the yule.cov function to investigate the relationship  
between a morphological variable and diversifficaiton using a dataset  
of ca. 300 species. The morphological variable was log-transformed and  
scaled, and ancestral reconstruction ran smoothly.

When I run the yule.cov function I get the results with all the  
information excepto for the StdErr, and I do get a warning message:

"The likelihood gradient seems flat in at least one dimension (null  
gradient):
cannot compute the standard-errors of the transition rates."

I wonder if anybody has some insights on why this is happening.


The oder issue is that the Log-likelihoods I obtained are smaller than  
for simple yule model fitted using yule(). Is ape reporting -LogLiks?


Thanks in advance,

Santiago



Santiago Claramunt
Museum of Natural Science,
119 Foster Hall,
Louisiana State University,
Baton Rouge, LA70803
sclara1 at tigers.lsu.edu
http://www.museum.lsu.edu/Claramunt/Home.html



More information about the R-sig-phylo mailing list