Posit is a machine learning algorithm. It chooses the most advantageous output at each point in time based on observed data. Use at your own risk!
Note that currently this library is immature/nonfunctional. Check back later.

TO DO LIST: -Revise model.cc to work with the move to pattern instead of occurrence
-Add a pointer back to the model in model_node and revise the get_supers() and get_subs() functions to have the same return format
-Smooth out the members of model_node so that we can make assurances about reference counting, iterator validity, and leave no dangling pointers or leak any memory.
-configuration space code does not have a way to truncate the current set of givens before the fulcrum node; we might want to have that.
-write a partial delink function so that we can delink parts of the training set without delinking the whole training set.
-write the link_node() function
-write a distribute_counts() function
-write the optimize() function