ScoreCalculator

class rectools.models.vector.ScoreCalculator(distance: Distance, subjects_factors: ndarray, objects_factors: ndarray)

Bases: object

Calculate proximity scores between one subject (e.g. user) and all objects (e.g. items) according to given distance metric.

Parameters

• distance (Distance) – Distance metric.

• subjects_factors (np.ndarray) – Array of subject embeddings, shape (n_subjects, n_factors).

• objects_factors (np.ndarray) – Array with embeddings of all objects, shape (n_objects, n_factors).

Inherited-members

Methods

calc(subject_id)

Calculate proximity scores between one subject and all objects according to given distance metric.

calc(subject_id: int) → ndarray

Calculate proximity scores between one subject and all objects according to given distance metric.

Parameters

subject_id (int) – Subject index.

Returns

Array of scores, shape (n_objects,).

Return type

np.ndarray