option to increase precision of floating point output

I love using `debug()` but its failing me right now with my current problem.  I'm trying to isolate a very small numeric divergence between a c++ and python implementation of two methods and I basically can't use `debug()` because it very much rounds the output precision for floats