Comments on: FWTools & MySQL leaves ArcGIS geoprocessor in the dust http://priour.wordpress.com/2008/01/22/fwtools-mysql-leaves-arcgis-geoprocessor-in-the-dust/ The (mis)Adventures of a self-taught developer Fri, 04 Jul 2008 06:33:24 +0000 http://wordpress.org/?v=MU By: priour http://priour.wordpress.com/2008/01/22/fwtools-mysql-leaves-arcgis-geoprocessor-in-the-dust/#comment-3623 priour Thu, 13 Mar 2008 19:03:03 +0000 http://priour.wordpress.com/2008/01/22/fwtools-mysql-leaves-arcgis-geoprocessor-in-the-dust/#comment-3623 @Tripp I honestly don't know how you would go about forcing ArcGIS to use BBOX intersection and not spatial intersection. I'm sure it is possible but probably not through any pre-installed tools or scripts. @Tripp
I honestly don’t know how you would go about forcing ArcGIS to use BBOX intersection and not spatial intersection. I’m sure it is possible but probably not through any pre-installed tools or scripts.

]]> By: Tripp http://priour.wordpress.com/2008/01/22/fwtools-mysql-leaves-arcgis-geoprocessor-in-the-dust/#comment-3622 Tripp Thu, 13 Mar 2008 18:34:05 +0000 http://priour.wordpress.com/2008/01/22/fwtools-mysql-leaves-arcgis-geoprocessor-in-the-dust/#comment-3622 Instead of trying to mimic the ArcGIS intersection in MySQL, can you mimic the MySQL intersection in ArcGIS (do the intersection in ArcGIS using the bounding box instead of the actual geometry)? Do you think that would make a difference? Instead of trying to mimic the ArcGIS intersection in MySQL, can you mimic the MySQL intersection in ArcGIS (do the intersection in ArcGIS using the bounding box instead of the actual geometry)? Do you think that would make a difference?

]]> By: Petz http://priour.wordpress.com/2008/01/22/fwtools-mysql-leaves-arcgis-geoprocessor-in-the-dust/#comment-3599 Petz Wed, 23 Jan 2008 10:05:37 +0000 http://priour.wordpress.com/2008/01/22/fwtools-mysql-leaves-arcgis-geoprocessor-in-the-dust/#comment-3599 Interesting benchmark, altough it's broken because MySQL uses bounding boxes instead of the actual geometry. Would be interesting to compare processing times on this with more spatial databases and GIS such as Oracle Spatial, IBM DB2, MS SQL Server 2008 and even maybe Manifold? Does anyone know if someone has done benchmark tests for basic GIS operations like this somewhere? Interesting benchmark, altough it’s broken because MySQL uses bounding boxes instead of the actual geometry. Would be interesting to compare processing times on this with more spatial databases and GIS such as Oracle Spatial, IBM DB2, MS SQL Server 2008 and even maybe Manifold?

Does anyone know if someone has done benchmark tests for basic GIS operations like this somewhere?

]]> By: dylan http://priour.wordpress.com/2008/01/22/fwtools-mysql-leaves-arcgis-geoprocessor-in-the-dust/#comment-3598 dylan Tue, 22 Jan 2008 20:09:40 +0000 http://priour.wordpress.com/2008/01/22/fwtools-mysql-leaves-arcgis-geoprocessor-in-the-dust/#comment-3598 Ok- it would be interesting to see how MySQL compares to the other spatial databases once it has proper geometry operators. Good luck! Ok- it would be interesting to see how MySQL compares to the other spatial databases once it has proper geometry operators.

Good luck!

]]> By: priour http://priour.wordpress.com/2008/01/22/fwtools-mysql-leaves-arcgis-geoprocessor-in-the-dust/#comment-3597 priour Tue, 22 Jan 2008 17:41:49 +0000 http://priour.wordpress.com/2008/01/22/fwtools-mysql-leaves-arcgis-geoprocessor-in-the-dust/#comment-3597 You are right that this operation is actually on MBR and not the specific geometeries so it is faster and less acurate than using PostGIS or ArcGIS. I didn't try it on PostGIS, just because I didn't already have the data in a table like I did for MySQL. However, I am certain than PostGIS, Oracle Spatial, or SQL Server 2008 would all perform this query in less than 1/4 the time it took ArcGIS. There is a beta or RC revision of the MySQL database which actually test geometries and not just MBR. However, I don't have it installed yet. You are right that this operation is actually on MBR and not the specific geometeries so it is faster and less acurate than using PostGIS or ArcGIS. I didn’t try it on PostGIS, just because I didn’t already have the data in a table like I did for MySQL. However, I am certain than PostGIS, Oracle Spatial, or SQL Server 2008 would all perform this query in less than 1/4 the time it took ArcGIS.

There is a beta or RC revision of the MySQL database which actually test geometries and not just MBR. However, I don’t have it installed yet.

]]> By: dylan http://priour.wordpress.com/2008/01/22/fwtools-mysql-leaves-arcgis-geoprocessor-in-the-dust/#comment-3596 dylan Tue, 22 Jan 2008 16:27:07 +0000 http://priour.wordpress.com/2008/01/22/fwtools-mysql-leaves-arcgis-geoprocessor-in-the-dust/#comment-3596 Nice post Matt. This closely mirrors my findings using GDAL, PostGIS, etc. for automated tasks on large datasets. It would be a good idea to verify that the intersects() operator is operating on geometries and not bounding boxes. Up until recently MySQL geometry operations were only performed on so-called MBR, or bounding boxes [1]. I think that they have updated this functionality, but haven't tried it- I have since moved on to PostGIS. 1. http://dev.mysql.com/doc/refman/5.0/en/functions-that-test-spatial-relationships-between-geometries.html Nice post Matt. This closely mirrors my findings using GDAL, PostGIS, etc. for automated tasks on large datasets. It would be a good idea to verify that the intersects() operator is operating on geometries and not bounding boxes. Up until recently MySQL geometry operations were only performed on so-called MBR, or bounding boxes [1]. I think that they have updated this functionality, but haven’t tried it- I have since moved on to PostGIS.

1. http://dev.mysql.com/doc/refman/5.0/en/functions-that-test-spatial-relationships-between-geometries.html

]]>