www.cmm.bristol.ac.uk/forum

Hello,

using this
sqrt(coef(modelstim)[c("RP2_var_Intercept")])

I'm able to get the variance for level two, but wondered is there a way to do this for the credible interval? and if not, how is it possible to calculate this myself? As this returns the variance directly, does the standard error ...

Thanks so much Chris thats much clearer to me thank you

Hi Chris sorry to bother you again!

I am getting myself a little confused about the IGLS starting values. From everything I have read, it seems to indicate that when running the model if you have EstM=1, this generates starting values from IGLS, is this right? But I have seen some comments about ...

Hello,

When I run a null or basic adjusted model by age and sex only the variance is low, but when I fully adjust the model including adding area level variables, the variance (random intercept) increases? Does this incidate a problem somewhere as I thought that it should decrease when adjusting ...

Thats great thanks so much!

Hi Chris,

Thank you so much thats so helpful!

I have a couple other questions too. Firstly, how do you build up a model from previous values in R? And also if you run the model initially in IGLS do you have to specificy something when running then in MCMC about using those starting values or is it ...

Hello,

I am trying to calculate the median incidence rate ratio for my poisson three level model. I have a calculation and code to do this, however its based on either the glmer or brm package in R. I was wondering if someone might help me apply this to the functions/terms in MLwiN?

this is the ...

Hi Chris,

You are right thank you, i didn't notice that it had finished, I had used multiple chains so that makes sense thank you. The < was a mistake in writing out the code I used. Thanks for your help and quick reply!

Hello,

So my code worked with running this

model<-log(death)~1+gender+education+parentdeath+offset(log(personyears))+(1|region)+(1|household)
model1<-runMLwiN(Formula=model, D="Poisson", data=dataset, estoptions=list(EstM=1))

however, when I add the iterations and burn in, it says 'MLwiN is ...

Sorting the data with that code worked! My model is now running, thank you!