[postgis-users] SRID not matching error

[Åsmund Tokheim]

Hi

Am I correct in assuming that you previously ran PostGIS 2.0 and now use
2.1? I'm only speculating as I'm not that familiar with the raster
features, but according to the docs (http://postgis.net/docs/RT_ST_Clip.html),
st_clip was rewritten in C for the 2.1 version, and the change in behavior
probably stems from that. While such undocumented changes between versions
are unfortunate, I personally think that explicitly calling the transform
function is more in the spirit with how the rest of the PostGIS functions
handle srid mismatches.

Åsmund


On Thu, May 22, 2014 at 5:14 PM, Jayson Gallardo <jaysontrades at gmail.com>wrote:

> Well, the thing is, that's how it's been for the last year and has worked
> fine. The difference now is that our server was rebuilt from scratch and I
> had to reload a dump of our database. The other difference is that the
> previous database was on PostgreSQL 9.1, and now we're on 9.3.
>
>
> On Wed, May 21, 2014 at 6:45 PM, Åsmund Tokheim <asmundto at gmail.com>wrote:
>
>> Hi
>>
>> It seems like you transformed the polygon used in the st_intersects
>> function, but the polygon used in the st_clip function still has the 3857
>> srid.
>>
>> Åsmund
>>
>>
>> On Thu, May 22, 2014 at 1:04 AM, Jayson Gallardo <jaysontrades at gmail.com>wrote:
>>
>>> I was finally able to restore my database dump today. However, upon
>>> testing, I get this error:
>>>
>>>
>>> production=# SELECT ST_AsGDALRaster(ST_CLIP(ST_Union(rast),
>>> ST_GeomFromText('POLYGON((-10366992.073461 5117898.5805526,-10366635.506271
>>> 5117898.5805526,-10366635.506271 5118206.0551587,-10366992.073461
>>> 5118206.0551587,-10366992.073461 5117898.5805526))',3857)),'GTiff') FROM
>>> "dem_elevation" WHERE ST_Intersects(rast,
>>> ST_Transform(ST_GeomFromText('POLYGON((-10366992.073461
>>> 5117898.5805526,-10366635.506271 5117898.5805526,-10366635.506271
>>> 5118206.0551587,-10366992.073461 5118206.0551587,-10366992.073461
>>> 5117898.5805526))',3857),4269));
>>> NOTICE:  Geometry provided does not have the same SRID as the raster.
>>> Returning NULL
>>>
>>>
>>> However, all my DEMs are in the raster with SRID 4269. My DEMs have been
>>> partitioned however, but the restore should have restored all that back
>>> into place, right?
>>>
>>> This is what I have for dem_elevation in pgAdmin:
>>>
>>> -- Table: dem_elevation
>>>
>>> -- DROP TABLE dem_elevation;
>>>
>>> CREATE TABLE dem_elevation
>>> (
>>>   rid integer NOT NULL,
>>>   rast raster,
>>>   CONSTRAINT dem_elevation_pkey PRIMARY KEY (rid)
>>> )
>>> WITH (
>>>   OIDS=FALSE
>>> );
>>> ALTER TABLE dem_elevation
>>>   OWNER TO postgres;
>>> GRANT ALL ON TABLE dem_elevation TO postgres;
>>> GRANT SELECT, REFERENCES, TRIGGER ON TABLE dem_elevation TO
>>> postgis_reader;
>>>
>>> And one of the dependents:
>>>
>>> -- Table: dem_elevation_n33w090
>>>
>>> -- DROP TABLE dem_elevation_n33w090;
>>>
>>> CREATE TABLE dem_elevation_n33w090
>>> (
>>>   rid serial NOT NULL,
>>>   rast raster,
>>>   CONSTRAINT dem_elevation_n33w090_pkey PRIMARY KEY (rid),
>>>   CONSTRAINT enforce_height_rast CHECK (st_height(rast) = 50),
>>>   CONSTRAINT enforce_max_extent_rast CHECK
>>> (st_coveredby(st_convexhull(rast),
>>> '0103000020AD1000000100000005000000863D2B1A098056C03B126100F5FE3F40863D2B1A098056C09578563412804040CA451840BD3F56C09578563412804040CA451840BD3F56C03B126100F5FE3F40863D2B1A098056C03B126100F5FE3F40'::geometry)),
>>>   CONSTRAINT enforce_num_bands_rast CHECK (st_numbands(rast) = 1),
>>>   CONSTRAINT enforce_out_db_rast CHECK (_raster_constraint_out_db(rast)
>>> = '{f}'::boolean[]),
>>>   CONSTRAINT enforce_pixel_types_rast CHECK
>>> (_raster_constraint_pixel_types(rast) = '{32BF}'::text[]),
>>>   CONSTRAINT enforce_same_alignment_rast CHECK (st_samealignment(rast,
>>> '010000000097C6CEA0C845183F97C6CEA0C84518BF863D2B1A098056C0957856341280404000000000000000000000000000000000AD10000001000100'::raster)),
>>>   CONSTRAINT enforce_scalex_rast CHECK (st_scalex(rast)::numeric(16,10)
>>> = 0.000092592592593::numeric(16,10)),
>>>   CONSTRAINT enforce_scaley_rast CHECK (st_scaley(rast)::numeric(16,10)
>>> = (-0.000092592592593)::numeric(16,10)),
>>>   CONSTRAINT enforce_srid_rast CHECK (st_srid(rast) = 4269),
>>>   CONSTRAINT enforce_width_rast CHECK (st_width(rast) = 50)
>>> )
>>> INHERITS (dem_elevation)
>>> WITH (
>>>   OIDS=FALSE
>>> );
>>> ALTER TABLE dem_elevation_n33w090
>>>   OWNER TO postgres;
>>> GRANT ALL ON TABLE dem_elevation_n33w090 TO postgres;
>>> GRANT SELECT, REFERENCES, TRIGGER ON TABLE dem_elevation_n33w090 TO
>>> postgis_reader;
>>>
>>> -- Index: dem_elevation_n33w090_rast_gist
>>>
>>> -- DROP INDEX dem_elevation_n33w090_rast_gist;
>>>
>>> CREATE INDEX dem_elevation_n33w090_rast_gist
>>>   ON dem_elevation_n33w090
>>>   USING gist
>>>   (st_convexhull(rast));
>>>
>>>  Thanks for your help in advance,
>>> Jayson
>>>
>>>
>>>
>>> _______________________________________________
>>> postgis-users mailing list
>>> postgis-users at lists.osgeo.org
>>> http://lists.osgeo.org/cgi-bin/mailman/listinfo/postgis-users
>>>
>>
>>
>> _______________________________________________
>> postgis-users mailing list
>> postgis-users at lists.osgeo.org
>> http://lists.osgeo.org/cgi-bin/mailman/listinfo/postgis-users
>>
>
>
> _______________________________________________
> postgis-users mailing list
> postgis-users at lists.osgeo.org
> http://lists.osgeo.org/cgi-bin/mailman/listinfo/postgis-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/postgis-users/attachments/20140522/5836044e/attachment.html>