uf3.representation.angles.generate_triplets

generate_triplets(i_where: numpy.ndarray, j_where: numpy.ndarray, sup_composition: numpy.ndarray, hashes: numpy.ndarray, distance_matrix: numpy.ndarray, knot_sets: List[List[numpy.ndarray]]) → List[Tuple]

Identify unique “i-j-j’” tuples by combining provided i-j pairs, then compute i-j, i-k, and j-k pair distances from i-j-k tuples,

distance matrix, and knot sequence for cutoffs.

TODO: refactor to break up into smaller, reusable functions

Parameters

• i_where (np.ndarray) – sorted “i” indices

• j_where (np.ndarray) – sorted “j” indices

• sup_composition (np.ndarray) – composition given by atomic numbers.

• hashes (np.ndarray) – array of unique integer hashes for interactions.

• distance_matrix (np.ndarray) – pair distance matrix.

• knot_sets (np.ndarray) – list of lists of knot sequences per interaction

Returns

array of shape (n_triangles, 3)

Return type

tuples_idx (np.ndarray)