Handle Massive Permutations

[wbuchanan]

I, and a few users, have noticed that as the size of variables increases the internal calls to tuples will tend to throw an error for multivariate functions. Remember to check your email for the info from the tuples package authors.

It would be useful to build a mechanism that would check/test the total number of permutations and use a returned value from that function to specify any necessary options in subsequent calls to tuples. Possibly ask StataCorps/Bill Gould if it would be possible to include a copy of the Mata function that would handle this computation from the Mata Book.