Is there a way to minimize the number of missing obs. for a set of regressors?

[Adam Sadi]

Hello. I am trying to predict a continuous variable of income that is declared in interval in my survey data by using the command -intreg-

I have a set of variables defined for the prediction.
Is there a command I can use to tell me what is the most optimal combination of variables, for a given number of predictors, to use in a regression to minimize missing values?

Something like "optimize depvar varlist, n(5)" if I have 5 predictors?

Since my purpose is to simulate as best as I can, I don't think there is a problem with doing this right?

Any advice is appreciated. Enjoy your weekend

[Carlo Lazzaro]

Adam:
not that I know.
Stata applies listwise deletion and the -e(sample)- is the array of the observations with all observed values (aka, complete case amalysis).

[daniel klein]

Since my purpose is to simulate as best as I can, I don't think there is a problem with doing this right?

I do not understand what you are implying here. What do you want to simulate? Do you mean predict? If so, as opposed to what? Inference? What are the implicit alternatives that you seem to have thought about but do not want for whatever reasons?