图神经网络GNN：给图多个 node features和edge features

\'\'\'
摘自https://docs.dgl.ai/en/0.6.x/guide_cn/graph-feature.html
\'\'\'

import dgl
import torch as th

# ========================= 无权图 ======================================

g = dgl.graph(([0, 0, 1, 5], [1, 2, 2, 0])) # 6个节点，4条边

# each graph can have many \'node features\'
g.ndata[\'x\'] = th.ones(g.num_nodes(), 3)   # 节点特征x, 特征长度为3
g.ndata[\'y\'] = th.randn(g.num_nodes(), 5)  # 节点特征y，特征长度为5

# similarly, each graph can have many \'edge features\'
g.edata[\'x\'] = th.ones(g.num_edges(), dtype=th.int32)  # 标量整型边特征x
g.edata[\'z\'] = th.ones(g.num_edges(), dtype=th.float32)  # 浮点型型边特征z

print(\'g:\n\', g)

print(g.ndata[\'x\'][1])     # 获取节点特征x的节点1特征
print(g.ndata[\'y\'][1])     # 获取节点特征y的节点1特征

print(g.edata[\'x\'][th.tensor([0, 3])])  # 获取边特征x下的0和3节点特征
print()


# ========================= 有权图 ======================================

# edges 0->1, 0->2, 0->3, 1->3
edges = th.tensor([0, 0, 0, 1]), th.tensor([1, 2, 3, 3])
weights = th.tensor([0.1, 0.6, 0.9, 0.7])  # weight of each edge
g = dgl.graph(edges)
g.edata[\'w\'] = weights  # give it a name \'w\'
print(\'weighted graph:\n\', g)