sum_nodes(graph, feat[, weight]) Sums all the values of node field feat in graph, optionally multiplies the field by a scalar node field weight. sum_edges(graph, feat[, weight]) Sums all the values of edge field feat in graph, optionally multiplies the field by a scalar edge field weight. mean_nodes(graph, feat[, weight]) Averages all the values of node field feat in graph, optionally multiplies the field by a scalar node field weight. mean_edges(graph, feat[, weight]) Averages all the values of edge field feat in graph, optionally multiplies the field by a scalar edge field weight. max_nodes(graph, feat) Take elementwise maximum over all the values of node field feat in graph max_edges(graph, feat) Take elementwise maximum over all the values of edge field feat in graph topk_nodes(graph, feat, k[, descending, idx]) Return graph-wise top-k node features of field feat in graph ranked by keys at given index idx. topk_edges(graph, feat, k[, descending, idx]) Return graph-wise top-k edge features of field feat in graph ranked by keys at given index idx. softmax_nodes(graph, feat) Apply batch-wise graph-level softmax over all the values of node field feat in graph. softmax_edges(graph, feat) Apply batch-wise graph-level softmax over all the values of edge field feat in graph. broadcast_nodes(graph, feat_data) Broadcast feat_data to all nodes in graph, and return a tensor of node features. broadcast_edges(graph, feat_data) Broadcast feat_data to all edges in graph, and return a tensor of edge features.