GitHub user ChengXiangLi opened a pull request:
https://github.com/apache/flink/pull/1161
[FLINK-2549] add topK operator to DataSet.
The topK operator is implemented with `mapPartition()` followed by
`reduceGroup()`, each map tasks select top k elements, and transfer to
singleton reduce task(no `group()` before `reduceGroup()`), and then select top
k elements in reduce task.
The main part of this implementation:
1. An out-of-core self managed memory based PriorityQueue implementation.
2. Use runtime resources to build PriorityQueue for UDF. topK is not real
'native' operator actually, it contains UDF which use self managed memory based
PriorityQueue to select top k elements.
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/ChengXiangLi/flink priorityqueue
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/flink/pull/1161.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #1161
----
commit 7165d5ba78376cec1b3c752afb949e6273d45869
Author: chengxiang li <[email protected]>
Date: 2015-09-16T02:10:05Z
[FLINK-2549] add topK operator to DataSet.
----
---
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.
---
