Update the clipping function so it's representative of the final published functions

[kperrynrel]

Hey all, in putting together the data prize notebooks this week, I've noticed that the original clipping function in PVAnalytics that is based off of the paper Matt Muller and I put together (https://ieeexplore.ieee.org/abstract/document/9518733) is a little out-of-date, with the final RDtools functions from this paper giving more accurate clipping results. For example, in analyzing a power stream for system 2107 from the data prize, PVAnalytics clipping returns this:

And rdtools logic-based (which the PVAnalytics function is based on) returns this:

The XGBoost ML variation from Rdtools (which I also cover in the paper) returns this:

Should we just import the rdtools functions directly into PVAnalytics? Trying to think of a good way to handle the fact that we'd be looking at identical functions across two packages with the same citation.