www.cmm.bristol.ac.uk/forum

If you are running the model with MCMC ( EstM=1 ) and do not specify starting values then it runs IGLS first and uses the result from this as starting values. Otherwise it uses the values that you provide. You can see this in lines 1496-1520 of https://github.com/r-forge/r2mlwin/blob/master/R2MLwiN/...

Do you get any error messages when this happens, in particular anything in the command output window that runs in the background?

I suggest that you also double check the Installation instructions to check that you haven't accidently missed a step.

Bill gives some reasons why the variance can increase in his first comment in the following thread. It might be worth seeing whether that helps to explain what you are seeing too:

https://www.cmm.bristol.ac.uk/forum/vie ... f=1&t=3624

You can find an example of setting starting values for parameters in section 5.5 of the MCMC manual replication materials. If you do not specify these then the model will be run with (R)IGLS (up to the specified maximum number of iterations) to generate starting values before the MCMC run begins. Fo...

This may be a bug in the software. Can you please confirm which version of MLwiN you are using, as well as provide more details regarding exactly what steps you were taking when it happened?

R2MLwiN does not currently provide an implementation of the VarCorr() function, however it does return these values within the output of coef() . The following shows the same model being fitted with glmer , brms , MLwiN(IGLS) and MLwiN(MCMC) : data(mmmec, package = "R2MLwiN") library(lme4...

Unfortunately we don't have many resources covering negative binomial models in MLwiN. You can find some information on https://www.bristol.ac.uk/cmm/software/mlwin/mlwin-resources.html#discrete however I don't believe that this includes worked examples. Setting up these models in runmlwin should ta...

When you specify more than one chain in R2MLwiN each estimated chain is sent to a child R process so that they can be run in parallel. This can result in status messages not being sent back to the main R instance for display. You can check whether this is the cause by temporarily changing nchains ba...

To get the right answers the data will need to be sorted (or you could potentially turn on the xc option as you are using MCMC). This should be as simple as: dataset <- dataset[order(dataset$region, dataset$household), ] Setting sort.force to TRUE is supposed to do this for you within the function. ...

If you look at the code for the runMLwiN function at https://github.com/r-forge/r2mlwin/blob/master/R2MLwiN/R/runMLwiN.R you will see that the first message comes from line 601 and is only triggered if the levID option has been set. By default this is NULL and your syntax doesn't appear to set the o...