Hi David Annat,
Have you tried method = GLS in ace()? To decide which corStruct is
better you can fit intercept-only models with gls(), eg:
gls(x ~ 1, correlation = corBrownian(1, tr))
gls(x ~ 1, correlation = corMartins(1, tr))
The AIC values will suggest you which corStruct to choose. In
Hello all,
I have been using ace() to reconstruct ancestral states for a number
of continuous morphological traits. I need confidence intervals for my
analysis. However, for some sets of traits, I get the following error:
Warning message:
In sqrt(diag(solve(out$hessian))) : NaNs produced
The
Hi Dave and everyone,
From the responses I got to my email a few months ago, it seems to be
related to having a lot of polytomies or zero-length branches. adding a
small number to the branch length doesn't help if there are many of them
because the matrix is still very near singular and the
