[R-sig-phylo] Problems with fitDiscrete

2010-08-06 Thread Alejandro Gonzalez
Hello, I am trying to estimate transition rate parameters for simulated data using fitDiscrete in geiger. The data was simulated on an ultrametric phylogeny using sim.char. I've set the q matrix based on estimated transition parameters from a binary trait, on the same phylogeny where I now aim

Re: [R-sig-phylo] Problems with fitDiscrete

2010-08-06 Thread Carl Boettiger
Hi Alejandro, Probably your simulation involved no changes from the root state; take a look at fcare values. If no transitions are observed in the data, the routine cannot estimate a rate at which transitions occur. Another way to interpret the error message is branch lengths are too short