We should implement the ability to calculate the relative entropy using ensemble averages from Equation 4 of the relentless paper.
This method will give a path to designing bonded interactions without use of bonded distributions functions and should be more robust without sensitivity to histogramming in g(r) calculations. Long term we could probably end support for the g(r) method.