[postgis-users] ST_Transform Incorrect Results?

[Josh Abts]

Hello,

I have been having a problem where I am getting mixed results from two
servers using the same dataset.  The command I am running is as follows:

select astext(st_transform(start_point,
4326)),astext(st_transform(end_point, 4326)) from mytable where gid=123

The two columns are originally using SRID 2163 so I am transforming them
into 4326.  But the results on one server (correct) are different than the
results on a second server (incorrect).  They are off by .2 latitude
roughly.

For example one result set is as follows:

Server 1 (correct): "POINT(-88.11068 42.2810949683921)";"POINT(-88.10731
42.2844149305216)"
Server 2 (incorrect): "POINT(-88.11068 42.0896736466198)";"POINT(-88.10731
42.0929914612221)"

As you can see the longitudes are correctly transformed, but the latitude is
not.

I have also included the postgis_full_version output in case a version issue
might be the culprit.

Server 1: "POSTGIS="1.3.2" GEOS="3.0.0rc4-CAPI-1.3.3" PROJ="Rel. 4.5.0, 22
Oct 2006" USE_STATS"
Server 2: "POSTGIS="1.5.2" GEOS="3.2.2-CAPI-1.6.2" PROJ="Rel. 4.7.1, 23
September 2009" LIBXML="2.7.5" USE_STATS (procs from 1.5 r5385 need
upgrade)"

So you can see I am using a newer version of PostGIS as well as Proj4 but I
can't seem to figure out why this would result in an incorrect St_Transform
result.

I also checked further into the spatial_ref_sys table on both servers and
Server 1 has slightly more detailed srtext but they have the same proj4text
lines.  I tried copying the srtext from Server 1 to Server 2 and there was
no change.

The only other difference would be the system platforms themselves which are
below.

Server 1: "PostgreSQL 8.3.4, compiled by Visual C++ build 1400" (this is a
windows 2003 server)
Server 2: "PostgreSQL 8.4.5 on x86_64-pc-linux-gnu, compiled by GCC
gcc-4.4.real (Ubuntu 4.4.1-4ubuntu9) 4.4.1, 64-bit"

Anyone have any ideas of what I could look into to see why I might be
getting this slight variation in results?  I can't find anything that would
seem to be the problem or cause of a varying transformation...

Thanks!
Josh
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/postgis-users/attachments/20101020/3f27f746/attachment.html>