[postgis-users] Bug dropping the Z coordinate?

Paragon Corporation lr at pcorp.us
Mon Jan 11 08:54:23 PST 2010


Abe,

Good question?  Ask another question? So replace your first question with a
different one.  

What are you trying to achieve and why do you need it as a geometry?

Trying to stuff a volumetric object into a 2.5D polygon I fear is not
possible.  The polygon is a 2d object in 3D/4D space and well the box bounds
3D space.

In the case of 2D box that is realistic, but in 3D its not.  Perhaps if
PostGIS supported volumetric things that would be a possibility but it does
not.

Paul would be a more authoritative person on the subject.

Thanks,
Regina
 

-----Original Message-----
From: postgis-users-bounces at postgis.refractions.net
[mailto:postgis-users-bounces at postgis.refractions.net] On Behalf Of Abram
Gillespie
Sent: Monday, January 11, 2010 11:38 AM
To: PostGIS Users Discussion
Subject: Re: [postgis-users] Bug dropping the Z coordinate?

OK, I'll ask this then.  What's the best way to get this:

select ST_Extent3d(geometry_column) from my_table;

Into EWKB w/o losing the Z (or M) coordinate?

Doing:

select ST_AsEWKB(ST_Extent3D(geometry_column)) from my_table;

unfortunately falls prey to the issue we're discussing.

Thanks.
-Abe

On Sun, Jan 10, 2010 at 10:53 PM, Paragon Corporation <lr at pcorp.us> wrote:
> As a side note.  ST_Zmax doesn't exist for geometries.  It works 
> because geometries can be autocast to boxes.
>
> See matrix
> http://www.postgis.org/documentation/manual-svn/ch08.html#PostGIS_Type
> Functi
> onMatrix
>
> -----Original Message-----
> From: postgis-users-bounces at postgis.refractions.net
> [mailto:postgis-users-bounces at postgis.refractions.net] On Behalf Of 
> Paragon Corporation
> Sent: Sunday, January 10, 2010 10:50 PM
> To: 'PostGIS Users Discussion'
> Subject: Re: [postgis-users] Bug dropping the Z coordinate?
>
> Abe,
>
> It seems we are flattening the box3d geometry in to a 2D polygon 
> always instead of as a 3D/2.5D polygon.
> I'm not sure it's a bug though, because I think a box3d is an invalid 
> 2.5/3D polygon if you were to construct it.
>
> Can you envision it?
>
> Paul, Mark or Kevin might have a better clue, but I think it's at best 
> an ambiguous 3D polygon and that is the problem. A box3d isn't really 
> a 3D box, its more just a coordinate bounding thing.
>
> Hope that helps,
> Regina
>
> -----Original Message-----
> From: postgis-users-bounces at postgis.refractions.net
> [mailto:postgis-users-bounces at postgis.refractions.net] On Behalf Of 
> Abram Gillespie
> Sent: Sunday, January 10, 2010 4:31 PM
> To: PostGIS Users Discussion
> Subject: [postgis-users] Bug dropping the Z coordinate?
>
> Is this a bug?
>
> select ST_ZMax('BOX3D(0 0 0,0 0 1)'::box3d);
>
> versus
>
> select ST_ZMax(ST_GeomFromEWKB(ST_AsEWKB('BOX3D(0 0 0,0 0 
> 1)'::box3d)));
>
> The first query returns "1" as expected but the second returns "0" as 
> unexpected.
>
> I'm trying to add Z/M support to zigGIS and this is show-stopping me.
> Any work-arounds if this is a bug?  Also, I'll be happy to report the 
> bug if so.
>
> Thanks.
> -Abe
> _______________________________________________
> postgis-users mailing list
> postgis-users at postgis.refractions.net
> http://postgis.refractions.net/mailman/listinfo/postgis-users
>
>
>
> _______________________________________________
> postgis-users mailing list
> postgis-users at postgis.refractions.net
> http://postgis.refractions.net/mailman/listinfo/postgis-users
>
>
> _______________________________________________
> postgis-users mailing list
> postgis-users at postgis.refractions.net
> http://postgis.refractions.net/mailman/listinfo/postgis-users
>
_______________________________________________
postgis-users mailing list
postgis-users at postgis.refractions.net
http://postgis.refractions.net/mailman/listinfo/postgis-users






More information about the postgis-users mailing list