MCMC adding iterations and specifying diagonal matrix

Welcome to the forum for R2MLwiN users. Feel free to post your question about R2MLwiN here. The Centre for Multilevel Modelling take no responsibility for the accuracy of these posts, we are unable to monitor them closely. Do go ahead and post your question and thank you in advance if you find the time to post any answers!

Go to R2MLwiN: Running MLwiN from within R >> http://www.bris.ac.uk/cmm/software/r2mlwin/
Post Reply
tjsduq64
Posts: 31
Joined: Mon Jul 15, 2019 10:04 pm

MCMC adding iterations and specifying diagonal matrix

Post by tjsduq64 »

Hi,

1) I know in MLwiN, I can run additional interations after running some. For example, I run 500 burnins and 5000 iterations. But, if I find that for some parameters, I need to run more, I can just add more iterations. Can I do this in R using R2MLwiN?

I see that in R examples for MCMC manual, in Chapter 13.5, it just runs another model, rather than adding iterations using "More".

2) when I use separate coding to include three variables in random parts, I also get the covariances between these terms. I know in IGLS, I can specify diagonal matrix by using "clre" or "smat". How can I do this in MCMC?

Sun

ChrisCharlton
Posts: 1193
Joined: Mon Oct 19, 2009 10:34 am

Re: MCMC adding iterations and specifying diagonal matrix

Post by ChrisCharlton »

1) Due to the difficulty of coordinating running extra iterations between R and MLwiN we have not currently implemented this an option from R2MLwiN. What you could do however is to use the saveworksheet option in R2MLwiN to save a copy of the MLwiN worksheet after the initial model fit. You could then load this into MLwiN and run extra iterations there, although you would not be able to get the results easily back into R.

2) The clre and smat options that you have used with IGLS should also work the same way with MCMC models.

tjsduq64
Posts: 31
Joined: Mon Jul 15, 2019 10:04 pm

Re: MCMC adding iterations and specifying diagonal matrix

Post by tjsduq64 »

That is a good alternative, thanks!

Post Reply