core.graph.radius_graph_pbc

Copyright (c) Meta Platforms, Inc. and affiliates.

This source code is licensed under the MIT license found in the LICENSE file in the root directory of this source tree.

Functions

sum_partitions(→ torch.Tensor)
get_counts(x, length)
compute_neighbors(data, edge_index)
get_max_neighbors_mask(natoms, index, atom_distance, ...)	Give a mask that filters out edges so that each atom has at most
radius_graph_pbc(data, radius, max_num_neighbors_threshold)
canonical_pbc(data, pbc)
radius_graph_pbc_v2(data, radius, ...[, ...])

Module Contents

core.graph.radius_graph_pbc.sum_partitions(x: torch.Tensor, partition_idxs: torch.Tensor) → torch.Tensor

core.graph.radius_graph_pbc.get_counts(x: torch.Tensor, length: int)

core.graph.radius_graph_pbc.compute_neighbors(data, edge_index)

core.graph.radius_graph_pbc.get_max_neighbors_mask(natoms, index, atom_distance, max_num_neighbors_threshold, degeneracy_tolerance: float = 0.01, enforce_max_strictly: bool = False)

Give a mask that filters out edges so that each atom has at most max_num_neighbors_threshold neighbors. Assumes that index is sorted.

Enforcing the max strictly can force the arbitrary choice between degenerate edges. This can lead to undesired behaviors; for example, bulk formation energies which are not invariant to unit cell choice.

A degeneracy tolerance can help prevent sudden changes in edge existence from small changes in atom position, for example, rounding errors, slab relaxation, temperature, etc.

core.graph.radius_graph_pbc.radius_graph_pbc(data, radius, max_num_neighbors_threshold, enforce_max_neighbors_strictly: bool = False, pbc: torch.Tensor | None = None)

core.graph.radius_graph_pbc.canonical_pbc(data, pbc: torch.Tensor | None)

core.graph.radius_graph_pbc.radius_graph_pbc_v2(data, radius, max_num_neighbors_threshold, enforce_max_neighbors_strictly: bool = False, pbc: torch.Tensor | None = None)