Thanks, Harihar. But this is slightly more complicate than just using subgraph(filter()).
See the transitive reduction. http://en.wikipedia.org/wiki/Transitive_reduction My case has one more additional requirement to think about weight (like a maximum spanning tree). Using a linear transitive reduction algorithm (and get some hints from "TriangleCount.scale" in GraphX), it might have some steps as 1. Compute the set of neighbors for each vertex. 2. For each edge, compute the intersection of the sets and send the weight to both vertices. 3. For each vertex, mark an edge as "False" if it has the intersection, but lower weight. 4. Remove all "False" edges using subgraph. But I am sure GraphX developer might have a better scalable and succinct idea for this problem. Thanks, Ted -- View this message in context: http://apache-spark-user-list.1001560.n3.nabble.com/Spark-GraphX-question-tp20768p20777.html Sent from the Apache Spark User List mailing list archive at Nabble.com. --------------------------------------------------------------------- To unsubscribe, e-mail: user-unsubscr...@spark.apache.org For additional commands, e-mail: user-h...@spark.apache.org
