Predictability.py contains functions that evaluate Random, Shannon, and time-correlated entropy, and calculate the value of Maximum Predictability as described in the paper by K Zhao, D Khryashchev, H Vo "Predicting Taxi and Uber Demand in Cities: Approaching the Limit of Predictability" in IEEE transactions on knowledge and data engineering (TKDE), 2019.
Examples of usage are shown in PredictabilityAndEntropy.ipynb notebook.