[
https://issues.apache.org/jira/browse/FLINK-3770?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15245554#comment-15245554
]
ASF GitHub Bot commented on FLINK-3770:
---------------------------------------
Github user vasia commented on a diff in the pull request:
https://github.com/apache/flink/pull/1899#discussion_r60047518
--- Diff:
flink-libraries/flink-gelly-examples/src/test/java/org/apache/flink/graph/library/TriangleEnumeratorITCase.java
---
@@ -47,10 +47,11 @@ public void testTriangleEnumerator() throws Exception
{
env);
List<Tuple3<Long,Long,Long>> actualOutput = graph.run(new
TriangleEnumerator<Long, NullValue, NullValue>()).collect();
- List<Tuple3<Long,Long,Long>> expectedResult =
TriangleCountData.getListOfTriangles();
+ List<Tuple3<Long,Long,Long>> expectedResult =
TriangleCountData.getListOfTriangles();
- Assert.assertEquals(actualOutput.size(), expectedResult.size());
+ Assert.assertEquals(expectedResult.size(), actualOutput.size());
for(Tuple3<Long,Long,Long> resultTriangle:actualOutput) {
+ System.out.println(resultTriangle);
--- End diff --
Do we need to print the result? We try to avoid printing from tests unless
it's absolutely necessary.
> Fix TriangleEnumerator performance
> ----------------------------------
>
> Key: FLINK-3770
> URL: https://issues.apache.org/jira/browse/FLINK-3770
> Project: Flink
> Issue Type: Improvement
> Components: Gelly
> Affects Versions: 1.1.0
> Reporter: Greg Hogan
> Assignee: Greg Hogan
> Fix For: 1.1.0
>
>
> Triangle enumeration is optimized by grouping edges by the vertex with lesser
> degree. Currently {{TriangleEnumerator}} thinks about this and then ignores
> the reordered edges. Also, since it is known that the vertex list will be
> much smaller than the edge list we can provide a {{JoinHint}}.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
