| Hi, I’m a GIS newbie that has TIGER_2010 data for U.S.-California loaded in PostGIS as a pilot. For California alone, the edges table stores the multilinestring data has 3 million rows. I need to load/use all 50 States when I am done. Attempts to run assign_vertex_id on California alone go on for a few days then just dies with nothing inserted. It seems to me that I should just add a column to the edges table (or a view or something related to edges) to permanently store the value returned by the point_to_id function instead of trying to generate a giant temporary table. Am I headed in the right direction? What are the consequences I have to work through? I am competent at editing the SQL functions. I just don’t know what the consequences of storing the data generated in point_to_id function somewhere else has on other pg_routing functions. Thanks in advance for your patience answering a newbie question. Michael. |
|---|
Hi Michael,
Not a direct answer, but I assume you are using the TIGER/Line all lines shapefiles to start with. One question is have you taken out all the non-street edges? If you haven’t you can greatly reduce the size of table you are working with, so you may be able to use a more direct approach. A second question is are you going to use this for local or for long distance routing as well? The TIGER road network is not designed for routing (it would have z values in addition to x and y values if it was). As a result, there are “intersections” in the TIGER data that are really over/under passes, to say nothing of one-way streets that aren’t identified. For routing over very small areas, this isn’t fatal, but things get funny as the size of the area increases.
Dan
Dan
— On Fri, 9/9/11, Michael Papet mpapet@yahoo.com wrote:
From: Michael Papet mpapet@yahoo.com
Subject: [pgrouting-users] Newbie: point_to_id and very large tiger_2010 tables
To: pgrouting-users@lists.osgeo.org
Received: Friday, September 9, 2011, 11:21 AMHi,
I’m a GIS newbie that has TIGER_2010 data for U.S.-California loaded in PostGIS as a pilot.
For California alone, the edges table stores the multilinestring data has 3 million rows. I need to load/use all 50 States when I am done. Attempts to run assign_vertex_id on California alone go on for a few days then just dies with nothing inserted.
It seems to me that I should just add a column to the edges table (or a view or something related to edges) to permanently store the value returned by the point_to_id function instead of trying to generate a giant temporary table.
Am I headed in the right direction? What are the consequences I have to work through? I am competent at editing the SQL functions. I just don’t know what the consequences of storing the data generated in point_to_id function somewhere else has on other pg_routing functions.
Thanks in advance for your patience answering a newbie question.
Michael.
-----Inline Attachment Follows-----
Pgrouting-users mailing list
Pgrouting-users@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/pgrouting-users