www.cmm.bristol.ac.uk/forum

Thanks for reporting this, it appears to be a bug. The contents of the global macro and the command option should be stored in a local macro, and then this used within the command. It appears that in some places, such as discrete models, we are using the global macro directly. I will fix this for th...

MLwiN works out that it is a level 2 variable from the fact that its value is constant within each level 2 identifier. This can be demostrated by modifying the data prior to adding the variable to the model so that this is no longer the case. If you do this the subscript will then be ij instead of j.

Since your initial post we have done some testing and as a result have updated the FAQ on running MLwiN on non-Windows systems ( http://www.bristol.ac.uk/cmm/software/mlwin/features/sysreq.html#unix ). Once we have a package that is confirmed to work we also intend to provide an FAQ on using runmlwi...

You can prevent Stata giving a 'more' prompt with the command: set more off For more information see http://www.stata.com/help.cgi?more . You can turn off the output from runmlwin by either prefixing it with quietly (see http://www.stata.com/help.cgi?quietly ), or you can turn off individual parts w...

We have not had any reports of MLwiN working under Wine before, which is why we haven't created versions of the plugin for Linux or Mac. I would be very interested in hearing how well it runs in this environment. If runmlwin is unable to find a particular plugin it should be able to fall back to eit...

One possible reason for this behaviour is that the two models are converging from different directions and are considered to have converged before reaching exactly the same point. The default tolerance in MLwiN (and runmlwin) is 10e-2, which corresponds to a 1% change in the value of the estimate fr...

The negative binomial distribution is currently not implemented in MLwiN when using MCMC estimation. We will try to provide a better error message to indicate this in a later release of runmlwin.

Looking at your data it appears that none of your identifier variables contain a value of zero, so that bug isn't the cause of the difference that you are seeing. A more likely cause of this issue is the change in starting values for the residuals in cross-classified models that was made in version ...

If you are experiencing this bug then adding one to the relevant classification columns should have fixed it. One other possibility is that because cross-classified models can't get starting residuals from a previously specified model that it is taking longer for the estimates to get into the right ...

There are some examples in section 4.2 of the User Guide (http://www.bristol.ac.uk/cmm/software/m ... man-09.pdf) and section 2 of the Manual Supplement (http://www.bristol.ac.uk/cmm/software/m ... 0-supp.pdf) that you might find useful.