I’ll Get Those Integrals Right Eventually

So I’ve been coding for a while now on one of my two research projects. One of the critical pieces is an Independent Metropolis Algorithm for getting some Bayes Estimates for the problem at hand.

Anyway, the likelihood is Cauchy over a convex, finite subset of the Real plane. My advisor and I have been going back and forth about what integrals work which way and go where in the transition portion of the algorithm–which we think we finally have straight.

It’s been a fairly interesting process. He’ll say “Okay, here’s what is needed” and I work with it, think about it, do some math, realize something isn’t right in my understanding, talk to him, and then go back and recode some of the math. Today it just happened to be integrals.