calcite-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Julian Hyde <jh...@apache.org>
Subject Re: Distinct aggregates
Date Tue, 04 Apr 2017 21:08:36 GMT
Calcite doesn’t currently support distinct aggregates in window functions. See https://issues.apache.org/jira/browse/CALCITE-476
<https://issues.apache.org/jira/browse/CALCITE-476>. 

That said, it shouldn’t silently swallow the DISTINCT keyword. It should give an error.
I suspect that some rule is ignoring the distinct flag. Can you log a JIRA case please?

Julian


> On Apr 4, 2017, at 9:46 AM, Radu Tudoran <radu.tudoran@huawei.com> wrote:
> 
> Hi,
> 
> I am trying to parse the logical query plan that results after executing queries such
as:
> 
> SELECT B1, SUM(DISTINCT B2) OVER  (ORDER BY  B4 RANGE BETWEEN INTERVAL '10' SECOND PRECEDING
AND CURRENT ROW) FROM T
> 
> However, when I look in the list of the aggregates from the LogicalWindow, the Aggregate
does not have any marker of being distinct. I also checked the isDistinct() and it returns
false for the aggregate.
> 
> .. I get as LogicalWindow object
> LogicalWindow(window#0=[window(partition {} order by [2] range between $3 PRECEDING and
CURRENT ROW aggs [COUNT($1), $SUM0($1)])])
> 
> Moreover if I run a query that has a count and a count distinct over a window it will
be simplified to a single count aggregate.
> 
> Am I missing something or is there a bug?
> 
> Best regards,
> 
> Radu
> 


Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message