Hi Russell , It can happen that PGLS models have difficulties to fit to the data when there is a free parameter in the correlation structure. This is the case of corMartins and corPagel. A way around is to set 'fixed = TRUE' when creating these correlation structures which has the effect of taking the value given to the argument of the same name as fixed. You may then try different values and avoid the error you observed.
HTH Best, Emmanuel ----- Le 14 Mai 21, à 2:56, Russell Engelman <neovenatori...@gmail.com> a écrit : > Dear R-Sig-Phylo > I am trying to re-submit a paper for publication and one of the things the > reviewers asked of me was to run the data under an OU model in addition to a > Brownian one. The problem is when I try to run an OU model I get the following > error: > "Error in corFactor.corStruct(object) : NA/NaN/Inf in foreign function call > (arg > 1)" > I looked up the error message but I couldn't really find what could be driving > it. My data doesn't have any missing values or tips without data, and I > checked > the branch lengths and none of the branches have a length of zero. Even when > running a force.ultrametric() parameter on the phylogeny the function still > fails to run. > The strange thing is the same code fits a Brownian model just fine but throws > back an error message when I try fitting an OU model. Attached is some code > that shows what I mean. I was wondering if anyone might know what was going on > with this? > Sincerely, > Russell > _______________________________________________ > R-sig-phylo mailing list - R-sig-phylo@r-project.org > https://stat.ethz.ch/mailman/listinfo/r-sig-phylo > Searchable archive at http://www.mail-archive.com/r-sig-phylo@r-project.org/ [[alternative HTML version deleted]] _______________________________________________ R-sig-phylo mailing list - R-sig-phylo@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo Searchable archive at http://www.mail-archive.com/r-sig-phylo@r-project.org/
