metadata_matrix

FeatureSystem.metadata_matrix(feature: tuple[int, int], data_list: list, normal_form: str = 'disjunctive', _known_feature_matrices: dict[int, tuple] = None) → ndarray

Explain the meaning of a feature, generated by repeatedly taking corners of features, by calculating a simplified logical term which explains the feature.

Parameters

featureFeature

The feature to explain.

data_listlist

The possible values of info_objects for separations. The rows of the matrix are also in this order.

normal_form{‘disjunctive’, ‘conjunctive’}

Disjunctive Normal Form (DNF) means that the result is a union of intersections. Conjunctive Normal Form (CNF) means that the result is an intersection of unions.

_known_feature_matricesdict

Cache already computed results. This avoids computing the same stuff twice.

Returns

np.ndarray

Matrix in CNF or DNF explaining the feature.