Optional Homework Exercises, Set #7, for Bock & Jones class #2:

1. (Hypothetically, here) If you were to set out to port the DA-Gibbs R 
procedures (DrawZ, DrawAlphaBeta, and the iteration loop) to C++, what 
additional functions would you need to implement? (Don't forget we already
have everything that's available in Davies' NewMat library, 
and you have uniform and normal random number generators in the
IRT Scale Score C++ project, and NormalProb from the ML estimation C++ 
from last week. So the question is, what else do you need besides
those things?) (Hint: The answer should be two routines, and you can explain 
either where you would obtain those routines or how you would code them.)

2. (This one would be very challenging.) Actually port the DA-Gibbs R
procedures (DrawZ, DrawAlphaBeta, and the iteration loop) to C++. This would
most easily be done as a modification of the ML C++ code from last week.
Indeed, it might be a good idea to simply add the MCMC code to the ML code, 
so that the ML computations could be used to obtain initial values of
alpha and beta for the MCMC iterations.