[postgis-devel] [PostGIS] #1639: [raster] MapAlgebraExpr can't handle conditions
PostGIS
trac at osgeo.org
Fri Mar 2 02:40:36 PST 2012
#1639: [raster] MapAlgebraExpr can't handle conditions
---------------------+------------------------------------------------------
Reporter: nicklas | Owner: pracine
Type: defect | Status: new
Priority: high | Milestone: PostGIS 2.0.0
Component: raster | Version: trunk
Keywords: |
---------------------+------------------------------------------------------
Comment(by nicklas):
Great Pierre
almost there
First 1-raster version is still not working as I understand dustymugs
comments above.
Second:
You are right I misunderstood things, but still, the casts doesn't make
sense.
If I cast to Integer and put it in '8BSI' like this:
{{{
WITH foo AS (
SELECT
ST_MapAlgebraExpr(a.rast, b.rast, '(([rast1]-[rast2]) >
5)::int', '8BSI') AS rast
FROM test_mapalgebra a
CROSS JOIN test_mapalgebra b
WHERE a.rid = 1
AND b.rid = 2
)
SELECT
(ST_BandMetadata(rast)).*,
(ST_SummaryStats(rast)).*
FROM foo;
}}}
Then I would expect the false results to show as 0, right?
But I still get the same result (nodata as I have learned now :-) )
I also wonder if something is wrong because of performance. I am using
this expressions to find clear cuttings (I will give a PostGIS talk next
week and will show the functionality)
To do this calculation in PostGIS it takes about 45 seconds, but in QGIS
with the same images as geotiff it takes less than 2 seconds.
But it can be because PostGIS can't gain fully from the SATA3 ssd on this
box while maybe the file based gdal calculation through QGIS can?
Or is it some casts that is giving a false bottleneck?
--
Ticket URL: <http://trac.osgeo.org/postgis/ticket/1639#comment:14>
PostGIS <http://trac.osgeo.org/postgis/>
The PostGIS Trac is used for bug, enhancement & task tracking, a user and developer wiki, and a view into the subversion code repository of PostGIS project.
More information about the postgis-devel
mailing list