Hi, I am trying to make forest plots using the ipdforest command in a dataset with and one without multiple imputed data. I got two different errors:
- in the non mi dataset, Stata reports:
'Warning: Binary exposure variable arm_original used
cluster variable trialnumf cannot be a string variable
use 'encode' and re-run regression model
not possible with string variable'
the variable is actually numeric, not string. (If it was a string variable I wouldn't have been able to run the xtmixed command as well).

-in the mi dataset, Stata requires the esample() option in the xtmixed estimation to be used, I have no idea how to add this option in the mi estimate command.


The exmixed command is defined as: xtmixed diff_pain arm_original pain0100 i.trialnum, || trialnumf:arm_original, nocons

the ipdforest is defined as: arm_original as expoure variable, fe: pain0100 i.trialnum, fets: trialnumf

The only instructions that I found on how to use the ipdforest command were from the paper from E. Kontopantelis and I couldn't find any solution there.
If anyone could help me I would be very thankful.