(Numba) Explore Numba Implementation

[rappdw]

In speaking with a couple of the folks at NVIDIA GTC, a path forward in this exploration would be:

1. Get basics working with Numba
2. Get basics working with Numba in "No Python" mode
3. will this work with @vectorize rather than @jit
4. If this all works, look at targeting GPU