##Available Estimators
- Maximum Likelihood
MaximumLikelihoodEntropyEstimator
MLEntropyEstimator
MLEE
- Miller-Maddow Bias Correction
MillerMaddowBiasCorrectedEntropyEstimator
MillerMaddowBiasCorrected
MMBCEE
- WORK IN PROGRESS
##Basic Usage
The most general syntax is
H = entropy(estimator, observations, base)
The default base is e
, i.e. the default unit is nats.
Using the maximum likelihood estimator, the entropy in nats can be calculated like this:
H = entropy(MLEE(), Frequencies([1, 1, 2]))
Using bits instead of nats:
H = entropy_ml(Frequencies([1, 1, 2]), 2) # bits
print(H) # -> 1.5
Instead of frequencies, the samples can also be provided directly:
H = entropy(MLEE(), Samples([1, 2, 3, 3]))
Some estimators provide convenience functions:
H = entropy_ml(Frequencies([1, 1, 2]))