Dear all,
I am trying to optimize the tree transformation parameters for a character
with 6 states using fitDiscrete in package geiger. However, I get the
following error:
Error in getQ(exp(out$par), nRateCats, model) :
You must supply the correct number of rate categories.
This has been reported on this list before and I have played around with
the different settings to try and isolate the problem: It seems to come up
with any character with more than 2 states, but only if model="ARD", and
irrespective of whether the tree transformation option is used or not.
e.g.
data(geospiza)
attach(geospiza)
gb <- ifelse(geospiza.data[, 1] < 4, 0, ifelse(geospiza.data[, 1] > 4.2, 2,
1))
names(gb) <- rownames(geospiza.data)
fitDiscrete(geospiza.tree, gb, model="ARD")
I wanted to check if I was doing something wrong before considering this a
bug in fitDiscrete. Has anyone had this problem and managed to solve it?
Thanks,
Jack Viljoen
[[alternative HTML version deleted]]
_______________________________________________
R-sig-phylo mailing list
[email protected]
https://stat.ethz.ch/mailman/listinfo/r-sig-phylo
