P Kishor wrote:
I actually wrote a pretty useful point-in-polygon routine using Perl
DBD::SQLite unwrapping Shapefiles into a SQLite db and then using
SQLite for boundary matching. It was for a very large p-in-p (7.5
million points against 250k polys) that ArcGIS was choking over. Works
very sweet in Perl/SQLite in about 20% to 30% of the time taken.
I'm curious about your SQLite spatial solution.
What approach did you use for storing geometry? How did you load the
shapefiles?
Since SQLite has no spatial index, did you just iterate over all
geometries for you PIP test? What was your PIP test written in?
What was the advantage of doing this over just using the raw shapefiles
- was it performance, or ease of data management?
The SQLite Spatial direction seems like an interesting one to pursue for
just this kind of bulk processing use case.
--
Martin Davis
Senior Technical Architect
Refractions Research, Inc.
(250) 383-3022
_______________________________________________
postgis-users mailing list
postgis-users@postgis.refractions.net
http://postgis.refractions.net/mailman/listinfo/postgis-users
