[pgpointcloud] Questions about indexing and querying

[Paul Ramsey]

> So I have endeavored to process and load the entire State of Iowa's 
> LiDAR data into pgpointcloud using PDAL. I currently have 1.24  
> million patches at ~120k points per patch. I have some questions  
> on how to use this holding more effectively.
>  
> 1) Is there an equivalent to this when the pointcloud_postgis  
> extension is enabled? Or did I miss something? The docs don't  
> say too much about indexing other than using the chipper to create  
> patches.

Arg, yeah, there’s a big hole in indexing currently. Basically you have to enable pointcloud_postgis and then

CREATE INDEX patches_idx on cloud using GIST (Geometry(patchcol));

And since it’s a functional index, you then need to make sure than one of your terms is “Geometry(patches)” when trying to invoke it.

> 2) Does a simple count, ala "select sum(PC_NumPoints(pa)) from  
> cloud", mean unpacking every patch all the way?

No, the npoints is stored in the header, so it should only mean a peek into the front of the page. 

P.

> Howard
> - signature.asc, 506 bytes _______________________________________________  
> pgpointcloud mailing list
> pgpointcloud at lists.osgeo.org
> http://lists.osgeo.org/cgi-bin/mailman/listinfo/pgpointcloud