recent won.txt

`!lg * recent / won` produces 2 queries, and divides one by the other. Here's the numerator:

explain analyze SELECT COUNT(*) AS fieldcount FROM logrecord 
                INNER JOIN l_ktyp ON logrecord.ktyp_id = l_ktyp.id 
                INNER JOIN l_cversion ON logrecord.cv_id = l_cversion.id
                WHERE ((l_ktyp.ktyp = 'winning') AND (l_cversion.cvnum >= '2200099000000'::bigint));

 Aggregate  (cost=446001.87..446001.88 rows=1 width=8) (actual time=9661.391..96
61.391 rows=1 loops=1)
   ->  Hash Join  (cost=4909.80..445813.68 rows=75274 width=0) (actual time=1912
.521..9658.063 rows=8193 loops=1)
         Hash Cond: (logrecord.cv_id = l_cversion.id)
         ->  Nested Loop  (cost=4908.11..445173.50 rows=220446 width=4) (actual 
time=26.322..9616.571 rows=81599 loops=1)
               ->  Seq Scan on l_ktyp  (cost=0.00..1.55 rows=1 width=4) (actual time=1.043..1.086 rows=1 loops=1)
                     Filter: (ktyp = 'winning'::citext)
                     Rows Removed by Filter: 43
               ->  Bitmap Heap Scan on logrecord  (cost=4908.11..442550.43 rows=262152 width=8) (actual time=25.267..9572.177 rows=81599 loops=1)
                     Recheck Cond: (ktyp_id = l_ktyp.id)
                     Rows Removed by Index Recheck: 618597
                     Heap Blocks: exact=37291 lossy=34619
                     ->  Bitmap Index Scan on ind_logrecord_ktyp_id  (cost=0.00..4842.57 rows=262152 width=0) (actual time=18.830..18.830 rows=81599 loops=1)
                           Index Cond: (ktyp_id = l_ktyp.id)
         ->  Hash  (cost=1.51..1.51 rows=14 width=4) (actual time=0.369..0.369 rows=2 loops=1)
               Buckets: 1024  Batches: 1  Memory Usage: 9kB
               ->  Seq Scan on l_cversion  (cost=0.00..1.51 rows=14 width=4) (actual time=0.361..0.365 rows=2 loops=1)
                     Filter: (cvnum >= '2200099000000'::numeric)
                     Rows Removed by Filter: 39
 Planning time: 0.895 ms
 Execution time: 9661.498 ms
(20 rows)

recent.txt

`!lg * recent / won` produces 2 queries, and divides one by the other. Here's the denominator:

explain analyze SELECT COUNT(*) FROM logrecord 
                INNER JOIN l_cversion ON logrecord.cv_id = l_cversion.id 
                WHERE (l_cversion.cvnum >= '2200099000000'::bigint);

Finalize Aggregate  (cost=567080.75..567080.76 rows=1 width=8) (actual time=112
6.771..1126.771 rows=1 loops=1)
   ->  Gather  (cost=567080.54..567080.75 rows=2 width=8) (actual time=1126.701..1131.216 rows=3 loops=1)
         Workers Planned: 2
         Workers Launched: 2
         ->  Partial Aggregate  (cost=566080.54..566080.55 rows=1 width=8) (actual time=1122.313..1122.313 rows=1 loops=3)
               ->  Hash Join  (cost=1.69..562630.46 rows=1380030 width=0) (actual time=244.437..1109.496 rows=228415 loops=3)
                     Hash Cond: (logrecord.cv_id = l_cversion.id)
                     ->  Parallel Seq Scan on logrecord  (cost=0.00..550923.16 rows=4041516 width=4) (actual time=0.042..783.569 rows=3233213 loops=3)
                     ->  Hash  (cost=1.51..1.51 rows=14 width=4) (actual time=0.049..0.049 rows=2 loops=3)
                           Buckets: 1024  Batches: 1  Memory Usage: 9kB
                           ->  Seq Scan on l_cversion  (cost=0.00..1.51 rows=14 width=4) (actual time=0.035..0.040 rows=2 loops=3)
                                 Filter: (cvnum >= '2200099000000'::numeric)
                                 Rows Removed by Filter: 39
 Planning time: 0.878 ms
 Execution time: 1131.304 ms