def from_networkx(G):
r"""Converts a :obj:`networkx.Graph` or :obj:`networkx.DiGraph` to a
:class:`torch_geometric.data.Data` 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 = torch_geometric.data.Data.from_dict(data)
data.num_nodes = G.number_of_nodes()
return data
def from_networkx(G):
r""""Converts a networkx graph to a data object graph.
Args:
G (networkx.Graph): A networkx graph.
"""
edge_index = torch.tensor(list(G.edges)).t().contiguous()
keys = []
keys += list(G.nodes(data=True)[0].keys())
keys += list(list(G.edges(data=True))[0][2].keys())
data = {key: [] for key in keys}
for _, feat_dict in G.nodes(data=True):
for key, value in feat_dict.items():
data[key].append(value)
for _, _, feat_dict in G.edges(data=True):
for key, value in feat_dict.items():
data[key].append(value)
for key, item in data.items():
data[key] = torch.tensor(item)
data['edge_index'] = edge_index
return torch_geometric.data.Data.from_dict(data)
def from_networkx(G):
r"""Converts a :obj:`networkx.Graph` or :obj:`networkx.DiGraph` to a
:class:`torch_geometric.data.Data` instance.
Args:
G (networkx.Graph or networkx.DiGraph): A networkx graph.
"""
G = G.to_directed() if not nx.is_directed(G) else G
edge_index = torch.tensor(list(G.edges)).t().contiguous()
keys = []
keys += list(list(G.nodes(data=True))[0][1].keys())
keys += list(list(G.edges(data=True))[0][2].keys())
data = {key: [] for key in keys}
for _, feat_dict in G.nodes(data=True):
for key, value in feat_dict.items():
data[key].append(value)
for _, _, feat_dict in G.edges(data=True):
for key, value in feat_dict.items():
data[key].append(value)
for key, item in data.items():
data[key] = torch.tensor(item)
data['edge_index'] = edge_index
data = torch_geometric.data.Data.from_dict(data)
data.num_nodes = G.number_of_nodes()
return data
def from_networkx(G,
group_node_attrs: Optional[Union[List[str], all]] = None,
group_edge_attrs: Optional[Union[List[str], all]] = None):
r"""Converts a :obj:`networkx.Graph` or :obj:`networkx.DiGraph` to a
:class:`torch_geometric.data.Data` instance.
Args:
G (networkx.Graph or networkx.DiGraph): A networkx graph.
group_node_attrs (List[str] or all, optional): The node attributes to
be concatenated and added to :obj:`data.x`. (default: :obj:`None`)
group_edge_attrs (List[str] or all, optional): The edge attributes to
be concatenated and added to :obj:`data.edge_attr`.
(default: :obj:`None`)
.. note::
All :attr:`group_node_attrs` and :attr:`group_edge_attrs` values must
be numeric.
"""
import networkx as nx
G = nx.convert_node_labels_to_integers(G)
G = G.to_directed() if not nx.is_directed(G) else G
if isinstance(G, (nx.MultiGraph, nx.MultiDiGraph)):
edges = list(G.edges(keys=False))
else:
edges = list(G.edges)
edge_index = torch.tensor(edges, dtype=torch.long).t().contiguous()
data = defaultdict(list)
if G.number_of_nodes() > 0:
node_attrs = list(next(iter(G.nodes(data=True)))[-1].keys())
else:
node_attrs = {}
if G.number_of_edges() > 0:
edge_attrs = list(next(iter(G.edges(data=True)))[-1].keys())
else:
edge_attrs = {}
for i, (_, feat_dict) in enumerate(G.nodes(data=True)):
if set(feat_dict.keys()) != set(node_attrs):
raise ValueError('Not all nodes contain the same attributes')
for key, value in feat_dict.items():
data[str(key)].append(value)
for i, (_, _, feat_dict) in enumerate(G.edges(data=True)):
if set(feat_dict.keys()) != set(edge_attrs):
raise ValueError('Not all edges contain the same attributes')
for key, value in feat_dict.items():
key = f'edge_{key}' if key in node_attrs else key
data[str(key)].append(value)
for key, value in data.items():
try:
data[key] = torch.tensor(value)
except ValueError:
pass
data['edge_index'] = edge_index.view(2, -1)
data = torch_geometric.data.Data.from_dict(data)
if group_node_attrs is all:
group_node_attrs = list(node_attrs)
if group_node_attrs is not None:
xs = []
for key in group_node_attrs:
x = data[key]
x = x.view(-1, 1) if x.dim() <= 1 else x
xs.append(x)
del data[key]
data.x = torch.cat(xs, dim=-1)
if group_edge_attrs is all:
group_edge_attrs = list(edge_attrs)
if group_edge_attrs is not None:
xs = []
for key in group_edge_attrs:
key = f'edge_{key}' if key in node_attrs else key
x = data[key]
x = x.view(-1, 1) if x.dim() <= 1 else x
xs.append(x)
del data[key]
data.edge_attr = torch.cat(xs, dim=-1)
if data.x is None and data.pos is None:
data.num_nodes = G.number_of_nodes()
return data
