mindspore_gl.graph.BatchMeta

class mindspore_gl.graph.BatchMeta(graph_nodes, graph_edges)[source]

BatchMeta, meta information for a batched graph.

Parameters

  • graph_nodes (numpy.array) – array of accumulated node sum for graphs in batched graph (first element is 0).

  • graph_edges (numpy.array) – array of accumulated edge sum for graphs in batched graph (first element is 0).

Supported Platforms:

Ascend GPU

Examples

>>> from mindspore_gl.graph import BatchMeta
>>> graph_nodes = np.array([0, 20, 40, 60, 80])
>>> graph_edges = np.array([0, 100, 200, 300, 400])
>>> graph = BatchMeta(graph_nodes, graph_edges)
>>> print(graph[1])
(20, 100)
property edge_map_idx

Index of edge list.

Returns

  • numpy.array, array indicate graph index for each edge

property graph_count

Graph numbers.

Returns

  • int, total graph count in this batched graph

property graph_edges

Edges array of graph.

Returns

  • numpy.array, accumulated edge sum for graphs in batched graph(first element is 0)

property graph_nodes

Nodes array of graph.

Returns

  • numpy.array, accumulated node sum for graphs in batched graph(first element is 0)

property node_map_idx

Index of node list.

Returns

  • numpy.array, array indicate graph index for each node