tag:blogger.com,1999:blog-8499895524521663926.post2367225968005183363..comments2024-03-27T07:13:39.236-04:00Comments on Phylogenetic Tools for Comparative Biology: Fitting a multivariate BM process in which the matrices in different subtrees different by a constantLiam Revellhttp://www.blogger.com/profile/04314686830842384151noreply@blogger.comBlogger3125tag:blogger.com,1999:blog-8499895524521663926.post-11648498143121116892016-04-14T09:52:13.051-04:002016-04-14T09:52:13.051-04:00Yes I obtains the same results when fitting multip...Yes I obtains the same results when fitting multiple independent rates matrices:<br /><br /> ## Simulate the data<br /> set.seed(14)<br /> tree<-pbtree(n=26,tip.label=LETTERS,scale=2)<br /> Q<-matrix(c(-1,1,1,-1),2,2)<br /> rownames(Q)<-colnames(Q)<-letters[1:2]<br /> tree<-sim.history(tree,Q)<br /> V<-list(a=matrix(c(1,0.5,0.5,1),2,2),b=matrix(c(10,5,5,10),2,2))<br /> X<-sim.corrs(tree,V)<br /> <br /> ## Fit:<br /> mvBM(tree,X)<br /> evol.vcv(tree,X)<br /><br />There is still the possibility to add various "constraints" (shared eigen vectors, shared variances, shared correlations, proportional matrices...) through the "param" list. I will update the package soon... and try to make a clearer vignette.<br />I heavily rely on your simmap format! Thank again,<br /><br />All the best,<br /><br />JulienAnonymoushttps://www.blogger.com/profile/05202832458676683014noreply@blogger.comtag:blogger.com,1999:blog-8499895524521663926.post-87782648446347077292016-04-10T09:59:21.259-04:002016-04-10T09:59:21.259-04:00Neat. Have you compared it to evol.vcv for real da...Neat. Have you compared it to evol.vcv for real datasets? I should try mvMORPH in a future workshop.<br />Thanks Julien. All the best, LiamLiam Revellhttps://www.blogger.com/profile/04314686830842384151noreply@blogger.comtag:blogger.com,1999:blog-8499895524521663926.post-69057100370340111572016-04-06T16:14:14.110-04:002016-04-06T16:14:14.110-04:00Hi Liam,
I also implemented such an approach in m...Hi Liam,<br /><br />I also implemented such an approach in mvMORPH which is maybe slightly more efficient.<br /><br />Just have to use the following code:<br /><br />mvBM(tree, X, param=list(constraint="proportional"))<br /><br />All the best,<br /><br />JulienAnonymoushttps://www.blogger.com/profile/05202832458676683014noreply@blogger.com