Hi,

I have 50 imputed datasets, created by the use of mi ice. I now would like to do Bayesian regression analyses on these data. However, since the official bayes commands are not supported by mi estimate, I thought about doing bayesian analyses on each of the 50 datasets, and then combine them in another way. Could I do this by writing a program similar to the one found at stata.com's FAQ, "How can I combine results other than coefficients in e(b) with multiply imputed data?", by Cañette and Marchenko (https://www.stata.com/support/faqs/s...mputed-data/)? If possible, could anyone give me some advice on how to modify the code below, so it can be applied to the following bayesian model:

Best regards,

Kjell Weyde

I have 50 imputed datasets, created by the use of mi ice. I now would like to do Bayesian regression analyses on these data. However, since the official bayes commands are not supported by mi estimate, I thought about doing bayesian analyses on each of the 50 datasets, and then combine them in another way. Could I do this by writing a program similar to the one found at stata.com's FAQ, "How can I combine results other than coefficients in e(b) with multiply imputed data?", by Cañette and Marchenko (https://www.stata.com/support/faqs/s...mputed-data/)? If possible, could anyone give me some advice on how to modify the code below, so it can be applied to the following bayesian model:

Code:

bayes, gibbs prior({var}, igamma(0.1, 0.1)) mcmcsize(5000) burnin(1000): regress y x1 x2 i.x3

Code:

cap program drop eroctab program eroctab, eclass version 12.0 /* Step 1: perform ROC analysis */ args refvar classvar roctab `refvar' `classvar' /* Step 2: save estimate and its variance in temporary matrices*/ tempname b V mat `b' = r(area) mat `V' = r(se)^2 local N = r(N) /* Step 3: make column names and row names consistent*/ mat colnames `b' = AUC mat colnames `V' = AUC mat rownames `V' = AUC /*Step 4: post results to e()*/ ereturn post `b' `V', obs(`N') ereturn local cmd "eroctab" ereturn local title "ROC area" endmi estimate, cmdok: eroctab attack high_bmi

Kjell Weyde

## Comment