Tom Lane <> wrote:

> Hm, that means there's only one grouping column (and it's the second
> tlist entry of the child plan node).  So that seems conclusive that
> the unique-ification is being done wrong.

Further confirmation using the EXPLAIN patch with Antonin's v2
patch against the table before any EXPLAIN or ANALYZE:

 Hash Join  (cost=37.12..80.40 rows=442 width=12)
   Hash Cond: (((upper.f2)::double precision = lower.f3) AND (upper.f1 = 
   ->  Seq Scan on subselect_tbl upper  (cost=0.00..27.70 rows=1770 width=16)
   ->  Hash  (cost=34.12..34.12 rows=200 width=12)
         ->  HashAggregate  (cost=32.12..34.12 rows=200 width=12)
               Group Key: lower.f2
               ->  Seq Scan on subselect_tbl lower  (cost=0.00..27.70 rows=1770 

The additional information is so useful, I'm all for committing
that patch.

Kevin Grittner
The Enterprise PostgreSQL Company

Sent via pgsql-hackers mailing list (
To make changes to your subscription:

Reply via email to