Score Functions - Categorical Kernels

Categorical score kernels used by the score module. They were extracted into this sibling module so that causaliq_data.score stays within the module size limits described in the development policy. Every name is re-exported from causaliq_data.score, so import paths are unchanged.

Bayesian Scores

bayesian_score

bayesian_score(N_ijk: ndarray, q_i: int, type: str, params: Dict[str, Any]) -> float

Return Bayesian based scores for marginal counts (for a node).

Variable names follow standard names used for the relevant formula in the literature.

Parameters:

• N_ijk

  (ndarray) –

  2-D array of instance counts for node i, parental combo j and node value k.

• q_i

  (int) –

  Maximum possible number of parental value combinations (even if they don't all necessarily occur in the data).

• type

  (str) –

  Bayesian score type required e.g. bde, k2.

• params

  (dict) –

  Parameters used in score computation: 'base' is logarithm base.

Returns:

• dict ( float ) –

  Of requested scores {score: value}.

Raises:

• TypeError –

  If any argument has invalid type.

• ValueError –

  If any argument has invalid value.

Entropy Scores

entropy_scores

entropy_scores(
    Nijk: ndarray,
    types: Union[str, List[str]],
    params: Dict[str, Any],
    N: int,
    free_params: int,
) -> Dict[str, float]

Return entropy based scores for marginal counts (for a node).

Parameters:

• Nijk

  (ndarray) –

  2-D array of instance counts for node i, parental combo j and node value k.

• types

  (str / list) –

  Entropy based score(s) required.

• params

  (dict) –

  Parameters used in score computation: 'base' is logarithm base.

• N

  (int) –

  Number of cases (instances).

• free_params

  (int) –

  Number of free parameters.

Returns:

• dict ( Dict[str, float] ) –

  Of requested scores {score: value}.

Raises:

• TypeError –

  If any argument has invalid type.

• ValueError –

  If any argument has invalid value.