[
https://issues.apache.org/jira/browse/FLINK-4017?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15318676#comment-15318676
]
Geoffrey Mon commented on FLINK-4017:
-------------------------------------
At least one of the issues I have come across is that aggregations are picky
about the types of the data they are operating on, which is not explicit for
Python lambdas. For example, in the WordCount example for Python, the return
type of the lambda used in flat_map is assumed by Flink to be raw bytes, which
are not Aggregator-able.
> [py] Add Aggregation support to Python API
> ------------------------------------------
>
> Key: FLINK-4017
> URL: https://issues.apache.org/jira/browse/FLINK-4017
> Project: Flink
> Issue Type: Improvement
> Components: Python API
> Reporter: Geoffrey Mon
> Priority: Minor
>
> Aggregations are not currently supported in the Python API.
> I was getting started with setting up and working with Flink and figured this
> would be a relatively simple task for me to get started with. Currently
> working on this at https://github.com/geofbot/flink
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
