Utilities

Utility functions.

action_effective	Test for the effect of the action on the state in the data H(state|prev_action) based on their conditional entropies.
action_reward	Test for the effect of the action on the reward in the data H(reward|prev_action).
backtest	Utility to perform backtesting on simulator.
better_than_random	Tests if the information gain of the input series given a conditioning series H(Y|X) is better than random.
conditional_entropy	The conditional entropy of the input series given a conditioning series H(Y|X).
data_generator_simple	Generate different types of data for your testing.
entropy	The entropy of the input series.
force_assert	Raise AssertionError when condition is False.
information_gain	Calculate the information gain of the input series given a conditioning series H(Y|X).
is_markovian	Test for the Markov property in the data H(state|prev_state, prev_action) based on their conditional entropies.
markovian_matrix	Test for the key MDP properties based on their conditional entropies.
normalized_markovian_matrix	Test for the key MDP properties based on their conditional entropies.
plot_information	Plot the results of MDP checks on dataframe data.
reward_function	Test for a reward function in the data H(r|state,action) based on their conditional entropies.
set_seed	Minimalistic implementation to fix random seeds in python.random, numpy, and pytorch.
stationary_policy	Test for a stationary policy in the data H(action|state) based on their conditional entropies.
tokenize	Concats all columns to one.