Updated and refactored (now with sklearn 0.23)

Updated to sklearn 0.23 (latest version) rather than using outdated sklearn 0.19.
Added Lipschitz utility rather than just a demo.
Added a static method to evaluate the utility function just using plot x, y coordinates rather than requiring a curve object to be passed in (i.e., LeastMontonicUtility.from_plot_vals(x, y)).  This allows the utilities to be decoupled from the rest of the framework.
Removed / hid older utilities (e.g. TotalVariation) by making them private (i.e., adding _ prefix). Need to test before releasing them in the public module.
Removed monkey patch (since sklearn 0.23 has PDP directly).
Removed deprecated counterfactual module and other deprecated things.
Simplified examples to use helpers.py as much as possible.

Initial code

v0.0.1

minor README updates