Have you considered curve smoothing or apply a small buffer around the orginal data/
Sandro Santilli wrote: > It will be _very_ slow. > So much that you may prefer to make the topology persistent. > > Another approach I'd like to see implemented would be to expose a > tolerance > to the GEOS implemented functions. Basically exposing the GEOS > PrecisionModel. By doing that you could specify a tolerance in the overlay > operations and the result should never have gaps bigger than the given > tolerance. > > --strk; > > On Thu, Oct 25, 2012 at 10:28:19AM -0700, pcr...@pcreso.com wrote: >> My impression is that working topologies is slow, especially with large >> datasets - lots of computation involved. >> >> How effective would this be with hundreds of thousands or millions or >> small polygons? >> >> Just running ST_Union() takes hours to days. >> >> Thanks >> >> Brent Wood >> >> --- On Fri, 10/26/12, Sandro Santilli <s...@keybit.net> wrote: >> >> From: Sandro Santilli <s...@keybit.net> >> Subject: Re: [postgis-users] How to eliminate small gaps produced by >> ST_union? >> To: "PostGIS Users Discussion" <postgis-users@postgis.refractions.net> >> Date: Friday, October 26, 2012, 6:04 AM >> >> You may try to construct a topology within PostGIS-2.0, using >> a small tolerance, and then get the geometries back. >> >> --strk; >> >> On Thu, Oct 25, 2012 at 05:54:05PM +0100, Alexandre Neto wrote: >> > I'm trying to Aggregate\Dissolve a set of adjacent, but not touching >> > polygons in Postgis. The geometries were created editing the Postgis >> table >> > in QGIS created, with snapping and avoid overlap in the layer. >> > >> > I used ST_Union like this: >> > >> > SELECT ST_Union(the_geom) >> > FROM table; >> > >> > Small gaps similar to lines appear in the result. This happens when >> the >> > adjacent polygons do not share the exact same vertices. >> > >> > [image: Inline image 2] >> > Is there a way to eliminate this "gaps" or "small angles" within >> Postgis? >> > >> > I know that v.clean in GRASS can eliminate small angles, and correct >> > topology, but that would mean I needed to copy the features to GRASS, >> > correct them and copy it back to Postgis. >> > >> > Thanks in advance, >> > >> > Alexandre Neto >> >> >> _______________________________________________ >> postgis-users mailing list >> postgis-users@postgis.refractions.net >> http://postgis.refractions.net/mailman/listinfo/postgis-users > >> _______________________________________________ >> postgis-users mailing list >> postgis-users@postgis.refractions.net >> http://postgis.refractions.net/mailman/listinfo/postgis-users > > > -- > > http://www.cartodb.com - Map, analyze and build applications with your > data > > ~~ http://strk.keybit.net > > _______________________________________________ > postgis-users mailing list > postgis-users@postgis.refractions.net > http://postgis.refractions.net/mailman/listinfo/postgis-users > -- _______________________________________________ postgis-users mailing list postgis-users@postgis.refractions.net http://postgis.refractions.net/mailman/listinfo/postgis-users