Yves Raimond created SPARK-11496:
------------------------------------
Summary: Parallel implementation of personalized pagerank
Key: SPARK-11496
URL: https://issues.apache.org/jira/browse/SPARK-11496
Project: Spark
Issue Type: Improvement
Components: GraphX
Affects Versions: 1.5.1
Reporter: Yves Raimond
Priority: Minor
The current implementation of personalized pagerank only supports one source
node. Most applications of personalized pagerank require to run the propagation
for multiple source nodes. However code such as:
{code}
sourceVertices.map { sourceVertex =>
graph.staticPersonalizedPageRank(sourceVertex, 10) }
{code}
Will be very slow, as it needs to run 10 iterations * sourceVertices.size
propagation steps.
It would be good to offer an alternative API that runs personalized pagerank
over a list of source vertices in parallel, so that it only needs to run 10
propagation steps in the example above.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
