[postgis-users] "summing" (polygon, float) pairs

[Martin Davis]

I read the original request as asking how to create a non-overlapping 
coverage from a set of overlapping polygons, and attribute each face of 
the coverage with the sum of its parent polygons' scores.  I think this 
is different to your solution.

AFAIK the best way to create a non-overlapping coverage from a set of 
overlapping polygons is to extract the linework (boundaries) from the 
polygons, node it together (by unioning all the lines) and then 
re-polygonize.  A point-in-polygon for each new face against the 
original set can be used to determine the parents, and then the summing 
is easy.

Obe, Regina wrote:
> If the below conditions are true, then I would say you can do this in
> PostGIS.
>
> A solution that comes to mind is this
>
> SELECT Sum(score) as totscore, ST_Union(the_geom) as newgeom
> FROM poly
> GROUP BY CAST(random()*10000 As int)
> HAVING All_Disjoint(the_geom)
>
>
> Set the 10000 so some function of the size of your set.  You always want
> it to be lower than the number of polygons you have. The larger the
> number the fewer polygons you would expect per match.
>
> All_Disjoint is a yet to be defined aggregate function you will need to
> create yourself that  checks each
> incoming geometry against the previous set and returns true if it is
> disjoint (not intersecting) any of the previous.
>
> Unfortunately the above will not use indexes.
>
> Aggregates are pretty easy to create in PostgreSQL so I'll leave that as
> an exercise for you to figure out.  The answers to the above will
> probably give different answers each time since it does a random walk.
>
> Hope that helps,
> Regina
>
>
>
>
>
> -----Original Message-----
> From: postgis-users-bounces at postgis.refractions.net
> [mailto:postgis-users-bounces at postgis.refractions.net] On Behalf Of Obe,
> Regina
> Sent: Thursday, October 02, 2008 10:28 AM
> To: PostGIS Users Discussion
> Subject: RE: [postgis-users] "summing" (polygon, float) pairs
>
> Reid,
>
> I assume
> 1) You don't care if polygons are reused to form your new sets 
> 2) You don't care how many polygons are in each set.  
> 3) You also are not looking for all possible permutations - just a
> random set of answers that solve your equation?
>
>
>
> -----Original Message-----
> From: postgis-users-bounces at postgis.refractions.net
> [mailto:postgis-users-bounces at postgis.refractions.net] On Behalf Of Reid
> Priedhorsky
> Sent: Wednesday, October 01, 2008 6:35 PM
> To: PostGIS Users Discussion
> Subject: [postgis-users] "summing" (polygon, float) pairs
>
> Dear all,
>
> I have a set of polygons. Associated with each polygon is a number; call
>
> it score. Each polygon can intersect with zero or more other polygons.
>
> What I would like to compute is a set of non-intersecting polygons with 
> summed scores -- each of these polygons would be a set of points where 
> the sum of all the original polygons' scores covered by the result 
> polygon are equal.
>
> Here's an example in one dimension:
>
> Input:
>
> ----3------
>       ----1----
>          ----4----
>                      ----2----
>
> Output:
>
> --3--
>       -4-
>          -8-
>             -5-
>                -4-
>                      ----2-----
>
> It is OK if two adjacent result polygons have equal sums.
>
> Can this be done in PostGIS?
>
> Reid
>
> p.s. thanks for the help on my question a few days ago! I felt a tad 
> foolish after learning it was just a missing isvalid() constraint. I 
> thought I had one but didn't look carefully enough.
> _______________________________________________
> postgis-users mailing list
> postgis-users at postgis.refractions.net
> http://postgis.refractions.net/mailman/listinfo/postgis-users
> -----------------------------------------
> The substance of this message, including any attachments, may be
> confidential, legally privileged and/or exempt from disclosure
> pursuant to Massachusetts law. It is intended
> solely for the addressee. If you received this in error, please
> contact the sender and delete the material from any computer.
> _______________________________________________
> postgis-users mailing list
> postgis-users at postgis.refractions.net
> http://postgis.refractions.net/mailman/listinfo/postgis-users
> _______________________________________________
> postgis-users mailing list
> postgis-users at postgis.refractions.net
> http://postgis.refractions.net/mailman/listinfo/postgis-users
>
>   

-- 
Martin Davis
Senior Technical Architect
Refractions Research, Inc.
(250) 383-3022