<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7652.24">
<TITLE>RE: [postgis-devel] More Cascade Union Adventures</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/plain format -->
<P><FONT SIZE=2>Martin,<BR>
Well there is plpgsql too which is very procedural so doesn't have to be expressed in sql per se. I use sql because it is generally speedwise faster for the planner to integrate and easier to prototype with (at least for me since I can get my thoughts down more succinctly granted probably harder to read for others) than plpgsql.<BR>
<BR>
If we replace the garray accum with a custom append - it can also look at the geometries as they are coming in and perhaps do a collect instead of an append and do some other conditional stuff.<BR>
<BR>
Where would I download this code to see what it looks like (JTS site only seems to go to JTS 1.8). I'm having a hard time picturing what you are saying. But if I see it in Java I can probably figure out the most efficient translation in sql/plpgsql.<BR>
<BR>
Thanks,<BR>
Regina<BR>
<BR>
-----Original Message-----<BR>
From: postgis-devel-bounces@postgis.refractions.net on behalf of Martin Davis<BR>
Sent: Wed 8/13/2008 8:08 PM<BR>
To: PostGIS Development Discussion<BR>
Subject: Re: [postgis-devel] More Cascade Union Adventures<BR>
<BR>
While pondering OJ's failure to complete the union on this dataset, I<BR>
realized that one trick that CascadedUnion uses is that when unioning<BR>
two MultiPolygons, it *avoids* unioning polygon components which lie<BR>
outside the intersection of the envelopes of the two input geometries. <BR>
Those polygon components will be left unchanged by the union, and so can<BR>
just be added in afterwards. This can make a big difference when large<BR>
far-flung polygons are being unioned - as is the case with this dataset.<BR>
<BR>
Not sure if you can use this technique in your algorithm - seems like it<BR>
might be tricky to express in SQL.<BR>
<BR>
Obe, Regina wrote:<BR>
><BR>
> Try this set. - this has 2895 records<BR>
><BR>
> My timings are<BR>
> ---time 161,047 ms = SELECT 161047/1000.0/60 = 2.68 minutes<BR>
> SELECT ST_CascadeUnion(the_geom)<BR>
> from usstatebounds;<BR>
><BR>
> --time 121719 ms = SELECT 121719/1000.0/60 = 2.02 minutes<BR>
> SELECT st_unitecascade_garray_sort(ARRAY(SELECT the_geom FROM<BR>
> usstatebounds));<BR>
><BR>
> In OJ - it gets to the last round and then fails with a Java out of<BR>
> Heap space error.<BR>
><BR>
><BR>
> ------------------------------------------------------------------------<BR>
><BR>
> *The substance of this message, including any attachments, may be<BR>
> confidential, legally privileged and/or exempt from disclosure<BR>
> pursuant to Massachusetts law. It is intended solely for the<BR>
> addressee. If you received this in error, please contact the sender<BR>
> and delete the material from any computer. *<BR>
><BR>
> ------------------------------------------------------------------------<BR>
><BR>
> * Help make the earth a greener place. If at all possible resist<BR>
> printing this email and join us in saving paper. *<BR>
><BR>
> * *<BR>
><BR>
> * *<BR>
><BR>
> ------------------------------------------------------------------------<BR>
><BR>
> _______________________________________________<BR>
> postgis-devel mailing list<BR>
> postgis-devel@postgis.refractions.net<BR>
> <A HREF="http://postgis.refractions.net/mailman/listinfo/postgis-devel">http://postgis.refractions.net/mailman/listinfo/postgis-devel</A><BR>
> <BR>
<BR>
--<BR>
Martin Davis<BR>
Senior Technical Architect<BR>
Refractions Research, Inc.<BR>
(250) 383-3022<BR>
<BR>
_______________________________________________<BR>
postgis-devel mailing list<BR>
postgis-devel@postgis.refractions.net<BR>
<A HREF="http://postgis.refractions.net/mailman/listinfo/postgis-devel">http://postgis.refractions.net/mailman/listinfo/postgis-devel</A><BR>
<BR>
<BR>
</FONT>
</P>
</BODY>
</HTML>