Github user ankurdave commented on the pull request:
https://github.com/apache/spark/pull/10539#issuecomment-170731872
@jkbradley I'm not completely sure of the cause, but one thing I noticed
was that PowerIterationClustering calls `GraphImpl.fromExistingRDDs` in several
places without preparing the vertices properly. That might be resulting in a
corrupted routing table and causing confusion here.
Each VertexRDD contains routing tables corresponding to a particular
EdgeRDD, and `GraphImpl.fromExistingRDDs` assumes the routing tables match the
provided EdgeRDD. If that's not true, the returned graph will give incorrect
results on triplet-based operations like `mapTriplets` and `aggregateMessages`.
I fixed this in PowerIterationClustering by replacing calls to
`GraphImpl.fromExistingRDDs(vertices, edges)` with `Graph(vertices, edges)` and
the two tests now fail in identical ways.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
