Hi all, My understanding of lambda is that it is a scalar multiplier for the off-diagonal elements of the variance covariance matrix computed from your tree. When lambda = 1, then the trait is evolving according to Brownian motion, when lambda = 0 there is no phylogenetic signal in the data.
When I do a PGLS using packages ape and nlme I get a single estimate of lambda for the following model. myCor<-corPagel(value=1,phy=myTree,fixed=F) gls(y~x, data=myDF, corr=myCor). I thought this single lambda value represented the estimate of lambda for variable y on myTree, but it is a different value than what I get when I use estimate lambda for variable y using package geiger fitDiscrete(phy=myTree,data=myDF$y,treeTransform="lambda"). Using figDiscrete I get a value very close to zero, whereas the lambda estimated from the gls is around .65 and is significantly different from zero. Have I made an error, or do I not understand how lambda is being calculated? Thanks! Andrew Barr PhD Student Dept. of Anthropology University of Texas Austin _______________________________________________ R-sig-phylo mailing list R-sig-phylo@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo
