[R-sig-phylo] converting phylog covariance to phylo object

Steven Kembel skembel at berkeley.edu
Thu Oct 9 00:38:25 CEST 2008


Hi Ben,

The phylosignal function in the picante R package calculates the K  
statistic (via the Kcalc function), and a P-value based on a  
randomization test of the variance of independent contrasts.

It would be very easy to change the Kcalc function to take a  
phylogenetic vcv matrix instead of a phylo object (just change first  
line of code in the function, we can do this for picante 0.5), but off  
the top of my head not sure how to change the PIC randomization test  
to work with a vcv matrix since it currently uses the ape pic function  
to calculate contrasts given a phylo object.

Matt Helmus implemented the methods described by Ives and Godfray 2006  
in the picante package, have a look at the functions pblm and  
pblmpredict.

Cheers,
Steve



More information about the R-sig-phylo mailing list