MolecularDiffusion.modules.models.syncogen.data.utils

Functions

encode_no_edge(→ torch.Tensor)
to_dense_bonds(bonds, bonds_len[, max_num_bonds])	Densify variable-length per-graph bonds lists into a padded batch.
to_dense_coords(coordinates, coords_mask, batch, ...)	Densify coordinates and coordinate masks if present.
to_dense_graph(x, edge_index, edge_attr, batch, ...)	Densify node and edge tensors and produce padding masks. Returns indices.
to_dense_pharmacophores(pharm_types, pharm_pos, pharm_len)	Densify variable-length per-graph pharmacophores into padded batches.

Module Contents

MolecularDiffusion.modules.models.syncogen.data.utils.encode_no_edge(edges_onehot: torch.Tensor) → torch.Tensor

MolecularDiffusion.modules.models.syncogen.data.utils.to_dense_bonds(bonds, bonds_len, max_num_bonds: int = None)

Densify variable-length per-graph bonds lists into a padded batch.

MolecularDiffusion.modules.models.syncogen.data.utils.to_dense_coords(coordinates, coords_mask, batch, max_num_nodes)

Densify coordinates and coordinate masks if present.

MolecularDiffusion.modules.models.syncogen.data.utils.to_dense_graph(x, edge_index, edge_attr, batch, max_num_nodes)

Densify node and edge tensors and produce padding masks. Returns indices.

MolecularDiffusion.modules.models.syncogen.data.utils.to_dense_pharmacophores(pharm_types, pharm_pos, pharm_len, max_num_pharm: int = None)

Densify variable-length per-graph pharmacophores into padded batches.