[postgis-users] Method to remove overlaps in a layer

[Sandro Santilli]

On Thu, Mar 15, 2012 at 11:26:22AM -0400, Pierre Racine wrote:
> Hi all,
> 
> Anybody know a well documented and efficient method to remove overlaps in a polygon table using PostGIS? 
> 
> I can think of two cases:
> 
> a) I may want to do it  on a single table containing overlaps.
> 
> b) I may want to do it because I want to update a clean (non-overlapping) table with a new one of smaller extent.
> 
> In both cases I would like to be able to define/use a best polygon selection strategy (based on a year attribute for example).
> 
> A small detail: I want to do this on a table containing 26 000 000 polygons...

I guess you'd better do this in steps:

 1. find out intersecting pairs into cache (id1,id2,geom1,geom2)
 2. update cache set geom1 = st_difference(geom1, geom2)
 3. update main set geom = c.geom1 from cache where id1 = id

Hey, step3 is probably correct SQL too ! :)

--strk; 

  ,------o-. 
  |   __/  |    Delivering high quality PostGIS 2.0 !
  |  / 2.0 |    http://strk.keybit.net - http://vizzuality.com
  `-o------'