[postgis-devel] Postgis topology creation - O(n-squared)? - creates problems with large datasets.

[Graeme B. Bell]

>> What about producing a syntetic dataset to serve as base for analisys
>> and see how it performs on differently configured systems ?
> 
> I had the same thought, and I'm making one right now.

http://graemebell.net/topologish/million_poly_topo1.zip
http://graemebell.net/topologish/million_poly_topo2.zip

If anyone has time to help, I would be super-interested to see other people's timings of building topologies with these datasets of topologish* data. > 1 million polys each.

e.g. pick polys, order by gid limit 5000, 10000, 20000, 40000, 80000, 160000, 320000, 640000.   SRID is 25833.

It will be next week before I can do it myself unfortunately. I'll check with my boss and get a CC-license for the files next week. 

Have a great weekend everyone, 

Graeme.


* not a mathematical term but ought to be