Hi Everyone,

I have the following problem:

I have dataset A and B.
I am running several (like 5 variations) of a first stage regressions on Dataset A and then I plan on using the predict command to create a variable and use that variable in several second stage regressions on Dataset B. If I had only one variation of the first stage - I could simply do:

- run regression on dataset A
- clear _all then load dataset B
- predict, f1
- include f1 in the new regression run on dataset B

I can do this for each variation of the first stage iteratively. However, I am wondering if there are some more efficient methods out there where I can run all the first stage regressions on dataset A, all at once and perhaps save those estimates and then use them to make predictions on dataset B. This way I have to load dataset A and B for once regardless of the number of variations I am experimenting with.

Thanks for reading.
- Ougni