From: | Adam Ruth <adamruth(at)mac(dot)com> |
---|---|
To: | Grzegorz Jaśkiewicz <gryzman(at)gmail(dot)com> |
Cc: | Postgres User <postgres(dot)developer(at)gmail(dot)com>, pgsql-general <pgsql-general(at)postgresql(dot)org> |
Subject: | Re: Converting each item in array to a query result row |
Date: | 2009-05-29 12:00:31 |
Message-ID: | DE7C3696-FCDA-4198-8095-12645869A9A7@mac.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Always test your performance assumptions. The plpgsql function is
faster than the sql function, a lot faster on smaller arrays.
unnest - 10 element array - 100,000 times: 6701.746 ms
unnest - 100 element array - 100,000 times: 11847.933 ms
unnest - 1000 element array - 100,000 times: 59472.691 ms
explode - 10 element array - 100,000 times: 1941.942 ms
explode - 100 element array - 100,000 times: 8521.289 ms
explode - 1000 element array - 100,000 times: 44980.048 ms
On 29/05/2009, at 8:55 PM, Grzegorz Jaśkiewicz wrote:
> why complicate so much ? this is a single sql query. It really shocks
> me, how people easily lean on plpgsql and for/loops - which are times
> slower than simple sql query:
>
> CREATE FUNCTION unnest(anyarray) RETURNS SETOF anyelement AS
> $_$
> SELECT ($1)[i] FROM
> generate_series(array_lower($1,1),array_upper($1,1)) i;
> $_$
> LANGUAGE sql IMMUTABLE;
From | Date | Subject | |
---|---|---|---|
Next Message | Grzegorz Jaśkiewicz | 2009-05-29 12:08:23 | Re: Converting each item in array to a query result row |
Previous Message | Thomas Pundt | 2009-05-29 11:22:38 | Re: Please remove me from the list! |