Chris Jackson

`sim.fmsm` simulates the entire pathway through a multistate model though. Do I understand correctly that you want to simulate data where each individual spends a time period of at least...

Can't see anything wrong with the transformation procedure there

Looks like something has gone wrong with the default method used to calculate matrix exponentials, which comes from the `expm` package. You can supply arguments to `expm` through `pmatrix.msm`. See...

Looking at that `qmatrix` and `ematrix`, that model will not be identifiable from the data. You are allowing misclassification between all states, and instantaneous transitions between all states. There will...

Can I explain in the simplest case with a 2-state model? If you have, say `qmatrix = rbind(c(0, 0.01), c(1, 0.02))`, and `ematrix = rbind(c(0, 0.2), c(0.1, 0))`, there are...

Sorry, the code for `pearson.msm` is messy! It's difficult for me to understand this issue though without a reproducible example - could you email me a minimal one, with simulated...

Again, a reproducible example might help (PB isn't in the data you sent me). Have you tried all the tricks in the "convergence failure" section of the PDF manual? Perhaps...

Try setting `control=list(fnscale=4000)` in the call to `msm`. Since the -2 log likelihood is around 4000 in this example, this normalises the -2 log-likelihood during optimisation to a scale of...

I have experimented with doing this rescaling automatically for all model fits, but it seemed to cause worse convergence for as many examples as it improved.

Could you give a reproducible example please @hbg26