www.cmm.bristol.ac.uk/forum

What I would like to do is take an existing MCMC model and add more iterations to it.

If I have
(mymodel7 <- runMLwiN(normexam ~ 1 + standlrt + (1 | school) + (1 | student), estoptions = list(EstM = 1, mcmcMeth = list(burnin = 0,
iterations = 500),
startval = startval), data = tutorial))



Can I update mymodel7 by adding (updating) the model with 5000 iterations?

I know that there is an update() in R2mlwin, but I am just not getting the specifications correctly.

Thanks

This option is currently not available, although you can get close by using the parameter estimates from the initial run as starting values for the second, and then merging the output chains. There is a saveworksheet() option that you can provide to R2MLwiN will allows you to specify a file used to store the MLwiN worksheet after a model run, so you could manually load this and do more iterations, however you will not be able to automatically bring these back into R. Once you know how long the model needs to be run you could then run it again in one go for the full length.