Computes the state-specific absolute pointwise mutual information between
two distributions.
This is the same as the MI, but with the absolute value.
- Parameters:
p (np.ndarray[float]) – The first probability distribution.
q (np.ndarray[float]) – The second probability distribution.
state (int | tuple[int, ...])
- Returns:
float – The maximum absolute pointwise mutual information.
- Return type:
float
