Comment by integraleq

You are correct about the representation, but I went for the most low-scale approach since the search space is so small. The next version will learn how to build its own rules, but that is for another day. Thanks for all of your help though. I really appreciate it.