[postgis-users] Trying to figure out how statistics for theplannerwork

Mark Cave-Ayland m.cave-ayland at webbased.co.uk
Mon Mar 13 06:58:49 PST 2006

> -----Original Message-----
> From: postgis-users-bounces at postgis.refractions.net [mailto:postgis-users-
> bounces at postgis.refractions.net] On Behalf Of strk at refractions.net
> Sent: 11 March 2006 18:46
> To: PostGIS Users Discussion <postgis-users at postgis.refractions.net>
> Subject: Re: [postgis-users] Trying to figure out how statistics for
> theplannerwork
> Happy to hear making it a define helped :)
> Note that there should also be a define for fine-tuning
> standard deviation computing. The standard deviation
> computation has been introduced to avoid problems when
> hard deviant are present (possibly bogus entries completely
> outside of the *real*  spatial domain, like geoms with
> wrong projection and similar).
> I wonder if it would be worth an additional final check
> to abort the standard-deviation computed bbox when
> the number of "hard-deviant" is above a given threshold...
> --strk;

Hi strk/Steve,

If Steve could do some tests with USE_STANDARD_DEVIATION enabled while
increasing the SDFACTOR constant (probably setting it to 3.5 - 4.0 would
make things work), I'd be happy to consider increasing the default. This was
really a result of my testing on our datasets here, so it is by no means
exhaustive. The idea, of course, was to catch a problem we had here where
someone had got an import script wrong (so most geometries were a factor of
100 out, including some infinite points!) which led to query plans being
created which ground the server to a halt.... :(

Kind regards,


WebBased Ltd
17 Research Way

T: +44 (0)1752 797131
F: +44 (0)1752 791023


This email and any attachments are confidential to the intended recipient
and may also be privileged. If you are not the intended recipient please
delete it from your system and notify the sender. You should not copy it or
use it for any purpose nor disclose or distribute its contents to any other

More information about the postgis-users mailing list