uf3.representation.distances.compute_direction_cosines¶
- compute_direction_cosines(sup_positions: numpy.ndarray, distance_matrix: numpy.ndarray, i_where: numpy.ndarray, j_where: numpy.ndarray, n_atoms: int) numpy.ndarray[source]¶
Intermediate function for computing derivatives for forces.
- Parameters
sup_positions – atom positions in supercell.
distance_matrix – output of spatial.distance.cdist(sup_positions, sup_positions).
i_where – indices of i-th atom based on distance mask.
j_where – indices of j-th atom based on distance mask.
n_atoms – number of atoms in original unit cell.
- Returns
- np.ndarray of shape (n_atoms, 3, n_distances)
and the second dimension corresponds to x, y, and z directions. Used to evaluate forces.
- Return type
drij_dR