Hi all,



I want to find out how PG gets rows when it explains a query.

For example, when I issue:



explain select count(*) from pg_class where relname='pg_class';

QUERY PLAN

--------------------------------------------------------------

Aggregate (cost=4.55..4.55 rows=1 width=0)

-> Seq Scan on pg_class (cost=0.00..4.55 rows=1 width=0)

Filter: (relname = 'pg_class'::name)

(3 rows)



Where and how PG get "rows=1"?



Thank you!

On Fri, Jul 29, 2005 at 11:54:18AM -0600, Lee Wu wrote:
> explain select count(*) from pg_class where relname='pg_class';
> QUERY PLAN
> --------------------------------------------------------------
> Aggregate (cost=4.55..4.55 rows=1 width=0)
> -> Seq Scan on pg_class (cost=0.00..4.55 rows=1 width=0)
> Filter: (relname = 'pg_class'::name)
> (3 rows)
>
> Where and how PG get "rows=1"?

The planner uses statistics to estimate the number of rows that a
query will return. Here are a couple of links that provide more
information:

http://www.postgresql.org/docs/8.0/s...ner-stats.html
http://www.postgresql.org/docs/8.0/s...-pg-stats.html
http://www.postgresql.org/docs/8.0/s...l-analyze.html
http://developer.postgresql.org/docs...s-details.html

--
Michael Fuhr
http://www.fuhr.org/~mfuhr/

---------------------------(end of broadcast)---------------------------
TIP 9: In versions below 8.0, the planner will ignore your desire to
choose an index scan if your joining column's datatypes do not
match