Thanks for the response, Liam. I just tried out OUwie and it works. However, it appears to only fit 1 variable at a time (I only get one sigma squared value). When running some tests with OUCH, I found that sigma squared is different when one variable is fit to an OU model, versus when multiple variables are fit together. Is that expected?
-Pascal On Mon, Mar 12, 2012 at 11:39 AM, Liam J. Revell <liam.rev...@umb.edu>wrote: > You might want to try the OUWie package by Beaulieu & O'Meara ( > http://cran.r-project.org/**web/packages/OUwie/index.html<http://cran.r-project.org/web/packages/OUwie/index.html> > )**. I have not tried it yet, but it promises to do multi-optimum OU > model fitting as well. > > All the best, Liam > > -- > Liam J. Revell > University of Massachusetts Boston > web: > http://faculty.umb.edu/liam.**revell/<http://faculty.umb.edu/liam.revell/> > email: liam.rev...@umb.edu > blog: http://phytools.blogspot.com > > > On 3/12/2012 2:31 PM, Pascal Title wrote: > >> Hi all >> >> I have a dataset of 5 PC axes for a phylogeny with 52 tips and I would >> like >> to get the variance-covariance matrix under a global OU model of >> evolution. I find that OU is strongly favored over BM, based on >> fitContinuous in GEIGER. >> However, I am having issues with convergence when trying to fit a hansen >> model to my data. What can I do to get around this problem? Alternatively, >> is there another way to get at the multivariate VCV matrix other than with >> the OUCH package? >> >> The error is: >> *unsuccessful convergence, code 1, see documentation for `optim'* >> *Warning message:* >> *In hansen(tree = ot, data = otd[varnames], regimes = otd["regimes"], :* >> * unsuccessful convergence* >> >> >> >> I have posted a small file that contains the following R script along with >> the data >> here<http://dl.dropbox.com/u/**34644229/OUhansen.zip<http://dl.dropbox.com/u/34644229/OUhansen.zip> >> >. >> >> >> Any advice would be greatly appreciated! Thanks! >> >> require(ouch) >> require(ape) >> >> tree<-read.nexus('tree.nex') >> data<-read.csv('data.csv') >> rownames(data)<-data$X >> data$X<-NULL >> >> tree >> head(data) >> >> colnames(data)->varnames #for hansen() >> >> ot<-ape2ouch(tree) >> otd<-as(ot,"data.frame") >> data$labels<-rownames(data) >> otd<-merge(otd,data,by="**labels",all=TRUE) >> rownames(otd)<-otd$nodes >> ot<-with(otd,ouchtree(nodes=**nodes,ancestors=ancestors,** >> times=times,labels=labels)) >> otd$regimes<-as.factor("**global") >> h1<-hansen(tree=ot,data=otd[**varnames],regimes=otd["** >> regimes"],sqrt.alpha=c(1,0,0,**0,0,1,0,0,0,1,0,0,1,0,1),** >> sigma=c(1,0,0,0,0,1,0,0,0,1,0,**0,1,0,1)) >> >> >> >> -- Pascal Title Graduate Student Burns Lab http://kevinburnslab.com/ Department of Evolutionary Biology San Diego State University 5500 Campanile Drive San Diego, CA 92182-4614 [[alternative HTML version deleted]] _______________________________________________ R-sig-phylo mailing list R-sig-phylo@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo
