Thanks, that can help to generate the RelNode. However it brings a new
issue, not sure it's related.

Briefly I'm testing to replace RelNode with my own RelNode, to build a
pipeline based on the RelNode tree. When running planner.transform(), an
error message is thrown as
https://gist.github.com/XuMingmin/491ff85011d2312acf086c2c0b7343c8.

----
The converter can work without the monotonic column with output:

SQL>:
select stream * from SOJ_EVENT
SQLPlan>
LogicalDelta
  LogicalProject(EVENTTIMESTAMP=[$0], ITEMID=[$1], SITEID=[$2],
PAGEID=[$3], PAGENAME=[$4])
    LogicalTableScan(table=[[SOJ_EVENT]])

beamPlan>
BeamProjectRel(EVENTTIMESTAMP=[$0], ITEMID=[$1], SITEID=[$2], PAGEID=[$3],
PAGENAME=[$4])
  BeamStreamScanRel(table=[[SOJ_EVENT, (STREAM)]])


Also paste my DeltaRule at
https://gist.github.com/XuMingmin/402cb9b3969763ee6b1efc80016c9a3f, Table
schema class here
https://gist.github.com/XuMingmin/0f0efa51be7e5567a7b2d93c206f2275.

Mingmin




On Tue, Feb 28, 2017 at 7:26 PM, 陈江枫 <[email protected]> wrote:

> implement
>
> Statistic getStatistic()  in table interface should work.
>
>
> @Override
> public Statistic getStatistic() {
>     return Statistics.of(100d,
>             ImmutableList.<ImmutableBitSet>of(),
>             RelCollations.createSingleton(0));
> }
>
> is a simple implementation declaring the first field to be ascending
>
>
> 2017-03-01 9:56 GMT+08:00 Mingmin Xu <[email protected]>:
>
> > Hello all,
> >
> > I'm writing some examples to understand the execution plan(more exactly
> > RelNode) of Calcite Streaming, and meet the error 'Streaming aggregation
> > requires at least one monotonic expression in GROUP BY clause' with
> > GROUP-BY.  Can anybody share an Java example on how to declare a column
> > monotonic in a StreamableTable? And, can I disable this check rule?
> >
> > Thank you!
> > --
> > ----
> > Mingmin
> >
>



-- 
----
Mingmin

Reply via email to