dgl.contrib.graph_store.SharedMemoryDGLGraph.update_all

SharedMemoryDGLGraph.update_all(message_func='default', reduce_func='default', apply_node_func='default')[source]

Distribute the computation in update_all among all pre-defined workers.

update_all requires that all workers invoke this method and will return only when all workers finish their own portion of computation. The number of workers are pre-defined. If one of them doesn’t invoke the method, it won’t return because some portion of computation isn’t finished.

Parameters:
  • message_func (callable, optional) – Message function on the edges. The function should be an Edge UDF.
  • reduce_func (callable, optional) – Reduce function on the node. The function should be a Node UDF.
  • apply_node_func (callable, optional) – Apply function on the nodes. The function should be a Node UDF.