ROC curves after multiple imputation

Hello all!

Just wanted to drop a lifeline in here to see if there any Stata experts with more knowledge than me on the subject.

I have run multiple imputation on covariates before putting them through mi estimate : logistic, on the way to building a prognostic model.
I've identified the relevant covariates, and I would like to build an ROC curve for my model. I have been attempting to use the code Steve Samuels came up with for combining survival curves, and I thought that perhaps this may be helpful.

What I have so far is listed below - but I keep coming up with error - "d1 invalid name r(198);

Perhaps it is not in fact possible to do this...in which case I would be grateful for any advice or alternative methods of "combining" imputed data sets to build an ROC curve.


__________________
save results, emptyok replace
use "my data", clear
local isets = r(M)
qui mi estimate: logistic severe_2015 cr between_onset_abx dyspnea alt
return list

tempfile d0
save "d0", replace
local isets = 20
qui{
forvalues i = 1/1{
tempfile d1
use "d0", clear
mi extract 1
logistic severe_2015 cr between_onset_abx dyspnea alt
lroc
outfile ("d1", replace)
use results, clear
append using "d1"
save, replace
}
}
__________________________