Source code for torch_geometric.utils.convert

import torch
import scipy.sparse
import networkx as nx


    import trimesh
except ImportError:
    trimesh = None

from .num_nodes import maybe_num_nodes

[docs]def to_scipy_sparse_matrix(edge_index, edge_attr=None, num_nodes=None): r"""Converts a graph given by edge indices and edge attributes to a scipy sparse matrix. Args: edge_index (LongTensor): The edge indices. edge_attr (Tensor, optional): Edge weights or multi-dimensional edge features. (default: :obj:`None`) num_nodes (int, optional): The number of nodes, *i.e.* :obj:`max_val + 1` of :attr:`index`. (default: :obj:`None`) """ row, col = edge_index.cpu() if edge_attr is None: edge_attr = torch.ones(row.size(0)) else: edge_attr = edge_attr.view(-1).cpu() assert edge_attr.size(0) == row.size(0) N = maybe_num_nodes(edge_index, num_nodes) out = scipy.sparse.coo_matrix((edge_attr, (row, col)), (N, N)) return out
[docs]def from_scipy_sparse_matrix(A): r"""Converts a scipy sparse matrix to edge indices and edge attributes. Args: A (scipy.sparse): A sparse matrix. """ A = A.tocoo() row = torch.from_numpy(A.row).to(torch.long) col = torch.from_numpy(A.col).to(torch.long) edge_index = torch.stack([row, col], dim=0) edge_weight = torch.from_numpy( return edge_index, edge_weight
[docs]def to_networkx(data, node_attrs=None, edge_attrs=None, to_undirected=False, remove_self_loops=False): r"""Converts a :class:`` instance to a :obj:`networkx.DiGraph` if :attr:`to_undirected` is set to :obj:`True`, or an undirected :obj:`networkx.Graph` otherwise. Args: data ( The data object. node_attrs (iterable of str, optional): The node attributes to be copied. (default: :obj:`None`) edge_attrs (iterable of str, optional): The edge attributes to be copied. (default: :obj:`None`) to_undirected (bool, optional): If set to :obj:`True`, will return a a :obj:`networkx.Graph` instead of a :obj:`networkx.DiGraph`. The undirected graph will correspond to the upper triangle of the corresponding adjacency matrix. (default: :obj:`False`) remove_self_loops (bool, optional): If set to :obj:`True`, will not include self loops in the resulting graph. (default: :obj:`False`) """ if to_undirected: G = nx.Graph() else: G = nx.DiGraph() G.add_nodes_from(range(data.num_nodes)) values = {} for key, item in data: if torch.is_tensor(item): values[key] = item.squeeze().tolist() else: values[key] = item if isinstance(values[key], (list, tuple)) and len(values[key]) == 1: values[key] = item[0] for i, (u, v) in enumerate(data.edge_index.t().tolist()): if to_undirected and v > u: continue if remove_self_loops and u == v: continue G.add_edge(u, v) for key in edge_attrs if edge_attrs is not None else []: G[u][v][key] = values[key][i] for key in node_attrs if node_attrs is not None else []: for i, feat_dict in G.nodes(data=True): feat_dict.update({key: values[key][i]}) return G
[docs]def from_networkx(G): r"""Converts a :obj:`networkx.Graph` or :obj:`networkx.DiGraph` to a :class:`` instance. Args: G (networkx.Graph or networkx.DiGraph): A networkx graph. """ G = nx.convert_node_labels_to_integers(G) G = G.to_directed() if not nx.is_directed(G) else G edge_index = torch.tensor(list(G.edges)).t().contiguous() data = {} for i, (_, feat_dict) in enumerate(G.nodes(data=True)): for key, value in feat_dict.items(): data[key] = [value] if i == 0 else data[key] + [value] for i, (_, _, feat_dict) in enumerate(G.edges(data=True)): for key, value in feat_dict.items(): data[key] = [value] if i == 0 else data[key] + [value] for key, item in data.items(): try: data[key] = torch.tensor(item) except ValueError: pass data['edge_index'] = edge_index.view(2, -1) data = data.num_nodes = G.number_of_nodes() return data
[docs]def to_trimesh(data): r"""Converts a :class:`` instance to a :obj:`trimesh.Trimesh`. Args: data ( The data object. """ if trimesh is None: raise ImportError('Package `trimesh` could not be found.') return trimesh.Trimesh(vertices=data.pos.detach().cpu().numpy(), faces=data.face.detach().t().cpu().numpy(), process=False)
[docs]def from_trimesh(mesh): r"""Converts a :obj:`trimesh.Trimesh` to a :class:`` instance. Args: mesh (trimesh.Trimesh): A :obj:`trimesh` mesh. """ if trimesh is None: raise ImportError('Package `trimesh` could not be found.') pos = torch.from_numpy(mesh.vertices).to(torch.float) face = torch.from_numpy(mesh.faces).t().contiguous() return, face=face)