Re: postgis-users Digest, Vol 55, Issue 3

[Martin Hoegh]

Jeg er på forældreorlov indtil d. 14. maj. Mine kollegaer kan traeffes paa tlf. 98129300.

I am on parental leave until May 14. If you have any matter of urgency please do not hesitate to call my collegues on [phone]

_______________________________________________
postgis-users mailing list
postgis-users@...
http://postgis.refractions.net/mailman/listinfo/postgis-users

[pcreso]

Hi,

I'm not sure what the problem is here, hopefully someone can help.

I have some polygons cropped to cell boundaries. (select intersection...)
I want to generate the geomunion of these for each cell, which seems
straightforward, but GEOS is throwing an error...

   select astext(geomunion(track_cell_geom)) from effort_cells3 where cell_id =
73721;
NOTICE:  TopologyException: no outgoing dirEdge found
(168.077,-40.1961,-40.1961)
ERROR:  GEOS union() threw an error!



Can anyone say what this means & what I need to do to work around it?

Thanks,

  Brent Wood
_______________________________________________
postgis-users mailing list
postgis-users@...
http://postgis.refractions.net/mailman/listinfo/postgis-users

[Martin Davis-2]

Let me guess - the two polygons causing this problem had a collinear
edge which crosses the cell boundary?

If so, this is the *toughest* situation for the union code to handle
robustly, since it is trying to union along two edges which are almost
collinear, but not quite.

Tough to suggest a fix... One possible way is to do the union of the
orginal polygons first, and then clip - but this may or may not be
possible in your situation.

If all you want is an area result, you could buffer the polys by a very
small amount, union, and then negative-buffer the result.

If you can post examples of the failure case, more ideas might come to
mind...

Brent Wood wrote:

> Hi,
>
> I'm not sure what the problem is here, hopefully someone can help.
>
> I have some polygons cropped to cell boundaries. (select intersection...)
> I want to generate the geomunion of these for each cell, which seems
> straightforward, but GEOS is throwing an error...
>
>    select astext(geomunion(track_cell_geom)) from effort_cells3 where cell_id =
> 73721;
> NOTICE:  TopologyException: no outgoing dirEdge found
> (168.077,-40.1961,-40.1961)
> ERROR:  GEOS union() threw an error!
>
>
>
> Can anyone say what this means & what I need to do to work around it?
>
> Thanks,
>
>   Brent Wood
> _______________________________________________
> postgis-users mailing list
> postgis-users@...
> http://postgis.refractions.net/mailman/listinfo/postgis-users
>
>  

--
Martin Davis
Senior Technical Architect
Refractions Research, Inc.
[phone]

_______________________________________________
postgis-users mailing list
postgis-users@...
http://postgis.refractions.net/mailman/listinfo/postgis-users

[Martin Davis-2]

Another idea - are you using the latest version of PostGIS/GEOS?  The
recent release added some GEOS enhancements which were designed to
address this issue.

Brent Wood wrote:

> Hi,
>
> I'm not sure what the problem is here, hopefully someone can help.
>
> I have some polygons cropped to cell boundaries. (select intersection...)
> I want to generate the geomunion of these for each cell, which seems
> straightforward, but GEOS is throwing an error...
>
>    select astext(geomunion(track_cell_geom)) from effort_cells3 where cell_id =
> 73721;
> NOTICE:  TopologyException: no outgoing dirEdge found
> (168.077,-40.1961,-40.1961)
> ERROR:  GEOS union() threw an error!
>
>
>
> Can anyone say what this means & what I need to do to work around it?
>
> Thanks,
>
>   Brent Wood
> _______________________________________________
> postgis-users mailing list
> postgis-users@...
> http://postgis.refractions.net/mailman/listinfo/postgis-users
>
>  

--
Martin Davis
Senior Technical Architect
Refractions Research, Inc.
[phone]

_______________________________________________
postgis-users mailing list
postgis-users@...
http://postgis.refractions.net/mailman/listinfo/postgis-users

[pcreso]

--- Martin Davis <mbdavis@...> wrote:

> Another idea - are you using the latest version of PostGIS/GEOS?  The
> recent release added some GEOS enhancements which were designed to
> address this issue.

Thanks Martin,

The version I'm running is:
ben200601=# select postgis_geos_version();
 postgis_geos_version
----------------------
 2.2.3-CAPI-1.1.1

I believe this is pretty much the current release, apart from the RC's of V3.

Sigh, I need this to work within a week or so somehow, I was hoping to use this
project as a showcase of FOSS GIS, but murphy strikes :-)

Is GEOS v3RC going to be reliable enough for a mission critical production job
with several million polygons to be manipulated?

What is entailed in upgrading GEOS?
Do I need to rebuild & reinstall PostGIS?
Will the current DB/data still be accessible?

Or is there some way of tweaking the underlying polygons to be combined to
address the problem?


Appreciated,

   Brent

>
> Brent Wood wrote:
> > Hi,
> >
> > I'm not sure what the problem is here, hopefully someone can help.
> >
> > I have some polygons cropped to cell boundaries. (select intersection...)
> > I want to generate the geomunion of these for each cell, which seems
> > straightforward, but GEOS is throwing an error...
> >
> >    select astext(geomunion(track_cell_geom)) from effort_cells3 where
> cell_id =
> > 73721;
> > NOTICE:  TopologyException: no outgoing dirEdge found
> > (168.077,-40.1961,-40.1961)
> > ERROR:  GEOS union() threw an error!
> >
> >
> >
> > Can anyone say what this means & what I need to do to work around it?
> >
> > Thanks,
> >
> >   Brent Wood
> > _______________________________________________
> > postgis-users mailing list
> > postgis-users@...
> > http://postgis.refractions.net/mailman/listinfo/postgis-users
> >
>

_______________________________________________
postgis-users mailing list
postgis-users@...
http://postgis.refractions.net/mailman/listinfo/postgis-users

[Paul Ramsey]

On 3-May-07, at 6:36 PM, Brent Wood wrote:

> Is GEOS v3RC going to be reliable enough for a mission critical  
> production job
> with several million polygons to be manipulated?

Probably.

> What is entailed in upgrading GEOS?

./configure ; make ; make install

> Do I need to rebuild & reinstall PostGIS?

No, you can drop the new GEOS in behind.  Re-connect to the database  
and the new library will be loaded.

> Will the current DB/data still be accessible?

Yes.

> Or is there some way of tweaking the underlying polygons to be  
> combined to
> address the problem?

Martin gave that recipe:  buffer(geomunion(buffer(a,0.1),buffer(b,
0.1)),-0.1)

P

>
>
> Appreciated,
>
>    Brent
>
>
>>
>> Brent Wood wrote:
>>> Hi,
>>>
>>> I'm not sure what the problem is here, hopefully someone can help.
>>>
>>> I have some polygons cropped to cell boundaries. (select  
>>> intersection...)
>>> I want to generate the geomunion of these for each cell, which seems
>>> straightforward, but GEOS is throwing an error...
>>>
>>>    select astext(geomunion(track_cell_geom)) from effort_cells3  
>>> where
>> cell_id =
>>> 73721;
>>> NOTICE:  TopologyException: no outgoing dirEdge found
>>> (168.077,-40.1961,-40.1961)
>>> ERROR:  GEOS union() threw an error!
>>>
>>>
>>>
>>> Can anyone say what this means & what I need to do to work around  
>>> it?
>>>
>>> Thanks,
>>>
>>>   Brent Wood
>>> _______________________________________________
>>> postgis-users mailing list
>>> postgis-users@...
>>> http://postgis.refractions.net/mailman/listinfo/postgis-users
>>>
>>
> _______________________________________________
> postgis-users mailing list
> postgis-users@...
> http://postgis.refractions.net/mailman/listinfo/postgis-users

_______________________________________________
postgis-users mailing list
postgis-users@...
http://postgis.refractions.net/mailman/listinfo/postgis-users

[pcreso]

--- Paul Ramsey <pramsey@...> wrote:

Hi Paul,

Much appreciated.

I never received Martin's first reply describing the buffer approach. I've
checked my trash folder & I did not delete it by mistake, I did recieve what
now appears to be his second message mentioning the version as a possible
issue, which was the one I replied to.

Thanks for all the info. I'll have a go with the buffer approach & see if that
works, if not I'll try v3.0.

I appreciate that v3 should work seamlessly, but I'm well down the track with a
very effective system so far & if I can make v2 do the job, then it ain't broke
enough yet :-)

The output is for reports to two Govt Depts, with about 2500 colour publication
quality maps. All work done in FOSS (postgis/GMT/R)

Many thanks for the help! As always, it is so much better than that for any
proprietary software I've ever seen!

 
Thanks again,

   Brent Wood

> On 3-May-07, at 6:36 PM, Brent Wood wrote:
>
> > Is GEOS v3RC going to be reliable enough for a mission critical  
> > production job
> > with several million polygons to be manipulated?
>
> Probably.
>
> > What is entailed in upgrading GEOS?
>
> ./configure ; make ; make install
>
> > Do I need to rebuild & reinstall PostGIS?
>
> No, you can drop the new GEOS in behind.  Re-connect to the database  
> and the new library will be loaded.
>
> > Will the current DB/data still be accessible?
>
> Yes.
>
> > Or is there some way of tweaking the underlying polygons to be  
> > combined to
> > address the problem?
>
> Martin gave that recipe:  buffer(geomunion(buffer(a,0.1),buffer(b,
> 0.1)),-0.1)
>
> P
>
> >
> >
> > Appreciated,
> >
> >    Brent
> >
> >
> >>
> >> Brent Wood wrote:
> >>> Hi,
> >>>
> >>> I'm not sure what the problem is here, hopefully someone can help.
> >>>
> >>> I have some polygons cropped to cell boundaries. (select  
> >>> intersection...)
> >>> I want to generate the geomunion of these for each cell, which seems
> >>> straightforward, but GEOS is throwing an error...
> >>>
> >>>    select astext(geomunion(track_cell_geom)) from effort_cells3  
> >>> where
> >> cell_id =
> >>> 73721;
> >>> NOTICE:  TopologyException: no outgoing dirEdge found
> >>> (168.077,-40.1961,-40.1961)
> >>> ERROR:  GEOS union() threw an error!
> >>>
> >>>
> >>>
> >>> Can anyone say what this means & what I need to do to work around  
> >>> it?
> >>>
> >>> Thanks,
> >>>
> >>>   Brent Wood
> >>> _______________________________________________
> >>> postgis-users mailing list
> >>> postgis-users@...
> >>> http://postgis.refractions.net/mailman/listinfo/postgis-users
> >>>
> >>
> > _______________________________________________
> > postgis-users mailing list
> > postgis-users@...
> > http://postgis.refractions.net/mailman/listinfo/postgis-users
>
> _______________________________________________
> postgis-users mailing list
> postgis-users@...
> http://postgis.refractions.net/mailman/listinfo/postgis-users
>

_______________________________________________
postgis-users mailing list
postgis-users@...
http://postgis.refractions.net/mailman/listinfo/postgis-users

[pcreso]

--- Paul Ramsey <pramsey@...> wrote:

Hi Paul,

Much appreciated.

I never received Martin's first reply describing the buffer approach. I've
checked my trash folder & I did not delete it by mistake, I did recieve what
now appears to be his second message mentioning the version as a possible
issue, which was the one I replied to.

Thanks for all the info. I'll have a go with the buffer approach & see if that
works, if not I'll try v3.0.

I appreciate that v3 should work seamlessly, but I'm well down the track with a
very effective system so far & if I can make v2 do the job, then it ain't broke
enough yet :-)

The output is for reports to two Govt Depts, with about 2500 colour publication
quality maps. All work done in FOSS (postgis/GMT/R)

Many thanks for the help! As always, it is so much better than that for any
proprietary software I've ever seen!

 
Thanks again,

   Brent Wood

> On 3-May-07, at 6:36 PM, Brent Wood wrote:
>
> > Is GEOS v3RC going to be reliable enough for a mission critical  
> > production job
> > with several million polygons to be manipulated?
>
> Probably.
>
> > What is entailed in upgrading GEOS?
>
> ./configure ; make ; make install
>
> > Do I need to rebuild & reinstall PostGIS?
>
> No, you can drop the new GEOS in behind.  Re-connect to the database  
> and the new library will be loaded.
>
> > Will the current DB/data still be accessible?
>
> Yes.
>
> > Or is there some way of tweaking the underlying polygons to be  
> > combined to
> > address the problem?
>
> Martin gave that recipe:  buffer(geomunion(buffer(a,0.1),buffer(b,
> 0.1)),-0.1)
>
> P
>
> >
> >
> > Appreciated,
> >
> >    Brent
> >
> >
> >>
> >> Brent Wood wrote:
> >>> Hi,
> >>>
> >>> I'm not sure what the problem is here, hopefully someone can help.
> >>>
> >>> I have some polygons cropped to cell boundaries. (select  
> >>> intersection...)
> >>> I want to generate the geomunion of these for each cell, which seems
> >>> straightforward, but GEOS is throwing an error...
> >>>
> >>>    select astext(geomunion(track_cell_geom)) from effort_cells3  
> >>> where
> >> cell_id =
> >>> 73721;
> >>> NOTICE:  TopologyException: no outgoing dirEdge found
> >>> (168.077,-40.1961,-40.1961)
> >>> ERROR:  GEOS union() threw an error!
> >>>
> >>>
> >>>
> >>> Can anyone say what this means & what I need to do to work around  
> >>> it?
> >>>
> >>> Thanks,
> >>>
> >>>   Brent Wood
> >>> _______________________________________________
> >>> postgis-users mailing list
> >>> postgis-users@...
> >>> http://postgis.refractions.net/mailman/listinfo/postgis-users
> >>>
> >>
> > _______________________________________________
> > postgis-users mailing list
> > postgis-users@...
> > http://postgis.refractions.net/mailman/listinfo/postgis-users
>
> _______________________________________________
> postgis-users mailing list
> postgis-users@...
> http://postgis.refractions.net/mailman/listinfo/postgis-users
>

_______________________________________________
postgis-users mailing list
postgis-users@...
http://postgis.refractions.net/mailman/listinfo/postgis-users

[pcreso]

--- Paul Ramsey <pramsey@...> wrote:

<snip>

> > Or is there some way of tweaking the underlying polygons to be  
> > combined to address the problem?
>
> Martin gave that recipe:  buffer(geomunion(buffer(a,0.1),buffer(b,
> 0.1)),-0.1)



Hi guys,

It looks like I'll need to upgrade GEOS to v3, as the buffer command has failed
to fix the problem with geomunion:

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
select buffer(geomunion(buffer(track_cell_geom,0.1)),-0.1)
from effort_cells3
where cell_id = 73722
  and isvalid(track_cell_geom);

NOTICE:  TopologyException: no outgoing dirEdge found
(168.129,-40.0659,-40.0659)
ERROR:  GEOS union() threw an error!

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Cheers,

  Brent
_______________________________________________
postgis-users mailing list
postgis-users@...
http://postgis.refractions.net/mailman/listinfo/postgis-users

[pcreso]

Hi all,

GEOS v3rc3 seems to have fixed the previous problem, but now I have another
one!
The usual Hydra like situation :-)

geomunion() sometimes returns a multipolygon, & at other times it returns a
polygon, depending I guess, on what it finds when it does it's thing.

Is it possible to force all the output to multipolygon, so both single polygons
& multi's can go in the same column?

Or can I populate two columns, then somehow insert the polygons as multi's into
the multi column?

I could do this by exporting as text, pass the file through awk/sed to turn
polygons into multi's, then insert them into the multi table.


I guess another approach might be a view unioning the two columns into one, as
the are both geometry datatypes as far as postgres is concerned, but this has
other impacts which are not very nice.

Any suggestions/advice appreciated.


Thanks (again!)


   Brent Wood
_______________________________________________
postgis-users mailing list
postgis-users@...
http://postgis.refractions.net/mailman/listinfo/postgis-users

[Paul Ramsey]

Brent, wrap your output in the "multi()" function, which will fluff
everything up to aggregate geometries.

Brent Wood wrote:

>
> Hi all,
>
> GEOS v3rc3 seems to have fixed the previous problem, but now I have another
> one!
> The usual Hydra like situation :-)
>
> geomunion() sometimes returns a multipolygon, & at other times it returns a
> polygon, depending I guess, on what it finds when it does it's thing.
>
> Is it possible to force all the output to multipolygon, so both single polygons
> & multi's can go in the same column?
>
> Or can I populate two columns, then somehow insert the polygons as multi's into
> the multi column?
>
> I could do this by exporting as text, pass the file through awk/sed to turn
> polygons into multi's, then insert them into the multi table.
>
>
> I guess another approach might be a view unioning the two columns into one, as
> the are both geometry datatypes as far as postgres is concerned, but this has
> other impacts which are not very nice.
>
> Any suggestions/advice appreciated.
>
>
> Thanks (again!)
>
>
>    Brent Wood
> _______________________________________________
> postgis-users mailing list
> postgis-users@...
> http://postgis.refractions.net/mailman/listinfo/postgis-users

--

   Paul Ramsey
   Refractions Research
   http://www.refractions.net
   pramsey@...
   Phone: [phone]
   Cell: 250-885-0632
_______________________________________________
postgis-users mailing list
postgis-users@...
http://postgis.refractions.net/mailman/listinfo/postgis-users

[Rob Agar-2]

hi Brent

Would just wrapping the call to GeomUnion with a call to Multi do the trick?

 From the docs:
Multi(geometry) - Returns the geometry as a MULTI* geometry. If the
geometry is already a MULTI*, it is returned unchanged.

hth
Rob

Brent Wood wrote:

>
> Hi all,
>
> GEOS v3rc3 seems to have fixed the previous problem, but now I have another
> one!
> The usual Hydra like situation :-)
>
> geomunion() sometimes returns a multipolygon, & at other times it returns a
> polygon, depending I guess, on what it finds when it does it's thing.
>
> Is it possible to force all the output to multipolygon, so both single polygons
> & multi's can go in the same column?
>
> Or can I populate two columns, then somehow insert the polygons as multi's into
> the multi column?
>
> I could do this by exporting as text, pass the file through awk/sed to turn
> polygons into multi's, then insert them into the multi table.
>
>
> I guess another approach might be a view unioning the two columns into one, as
> the are both geometry datatypes as far as postgres is concerned, but this has
> other impacts which are not very nice.
>
> Any suggestions/advice appreciated.
>
>
> Thanks (again!)
>
>
>    Brent Wood
> _______________________________________________
> postgis-users mailing list
> postgis-users@...
> http://postgis.refractions.net/mailman/listinfo/postgis-users
>

_______________________________________________
postgis-users mailing list
postgis-users@...
http://postgis.refractions.net/mailman/listinfo/postgis-users

[pcreso]

--- Rob Agar <robagar@...> wrote:

> hi Brent
>
> Would just wrapping the call to GeomUnion with a call to Multi do the trick?
>
>  From the docs:
> Multi(geometry) - Returns the geometry as a MULTI* geometry. If the
> geometry is already a MULTI*, it is returned unchanged.

Thanks Paul & Rob,

That seems absolutely perfect!!!

(I'm running it right now with separate poly & multi columns, so I'll just
update the multi column with the multi(poly) when it finishes....


Much appreciated guys!!!


Cheers,

  Brent
_______________________________________________
postgis-users mailing list
postgis-users@...
http://postgis.refractions.net/mailman/listinfo/postgis-users

[SangJun]

Brent Wood-2 wrote:

--- Rob Agar <[e-mail]> wrote:

> hi Brent
>
> Would just wrapping the call to GeomUnion with a call to Multi do the trick?
>
>  From the docs:
> Multi(geometry) - Returns the geometry as a MULTI* geometry. If the
> geometry is already a MULTI*, it is returned unchanged.

Thanks Paul & Rob,

That seems absolutely perfect!!!

(I'm running it right now with separate poly & multi columns, so I'll just
update the multi column with the multi(poly) when it finishes....


Much appreciated guys!!!


Cheers,

  Brent
_______________________________________________
postgis-users mailing list
[e-mail]
http://postgis.refractions.net/mailman/listinfo/postgis-users

Hi,

I have the same error (NOTICE:  TopologyException: no outgoing dirEdge found (3.07069,50.2997,50.2997)
SELECT failed: ERROR:  GEOS union() threw an error!) with one of my requests (select GeomUnion(GeomFromText('POLYGON((3.064484 50.307842, 3.076605 50.291935, 3.076635 50.291958, 3.064514 50.307865, 3.064484 50.307842))')),('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')).
Each part is valid. I don't understand how I can solve this problem (and I'm French).

[Nicolas Ribot]

> Hi,
>
> I have the same error (NOTICE:  TopologyException: no outgoing dirEdge found
> (3.07069,50.2997,50.2997)
> SELECT failed: ERROR:  GEOS union() threw an error!) with one of my requests
> (select GeomUnion(GeomFromText('POLYGON((3.064484 50.307842, 3.076605
> 50.291935, 3.076635 50.291958, 3.064514 50.307865, 3.064484
> 50.307842))')),('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')).
> Each part is valid. I don't understand how I can solve this problem (and I'm
> French).
> :-|

Works for me in this env (Mac OS X 10.4):
 PostgreSQL 8.2.3 on i386-apple-darwin8.8.1, compiled by GCC
i686-apple-darwin8-gcc-4.0.1 (GCC) 4.0.1 (Apple Computer, Inc. build
5367)
POSTGIS="1.2.1" GEOS="3.0.0rc4-CAPI-1.3.3" PROJ="Rel. 4.5.0, 22 Oct
2006" USE_STATS

GEOS 3 is much more robust than older versions.

result is:
geomunion|?column?
0103000000010000000500000010ECF82F10840840E92ADD5D672749403EAE0D15E39C084065C746205E254940E9D495CFF29C08401DE736E15E254940BA1281EA1F840840A14ACD1E6827494010ECF82F10840840E92ADD5D67274940|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

Nicolas
_______________________________________________
postgis-users mailing list
postgis-users@...
http://postgis.refractions.net/mailman/listinfo/postgis-users

[SangJun]

Nicolas Ribot wrote:

> Hi,
>
> I have the same error (NOTICE:  TopologyException: no outgoing dirEdge found
> (3.07069,50.2997,50.2997)
> SELECT failed: ERROR:  GEOS union() threw an error!) with one of my requests
> (select GeomUnion(GeomFromText('POLYGON((3.064484 50.307842, 3.076605
> 50.291935, 3.076635 50.291958, 3.064514 50.307865, 3.064484
> 50.307842))')),('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')).
> Each part is valid. I don't understand how I can solve this problem (and I'm
> French).
> :-|

Works for me in this env (Mac OS X 10.4):
 PostgreSQL 8.2.3 on i386-apple-darwin8.8.1, compiled by GCC
i686-apple-darwin8-gcc-4.0.1 (GCC) 4.0.1 (Apple Computer, Inc. build
5367)
POSTGIS="1.2.1" GEOS="3.0.0rc4-CAPI-1.3.3" PROJ="Rel. 4.5.0, 22 Oct
2006" USE_STATS

GEOS 3 is much more robust than older versions.

result is:
geomunion|?column?
0103000000010000000500000010ECF82F10840840E92ADD5D672749403EAE0D15E39C084065C746205E254940E9D495CFF29C08401DE736E15E254940BA1281EA1F840840A14ACD1E6827494010ECF82F10840840E92ADD5D67274940|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

Nicolas
_______________________________________________
postgis-users mailing list
[e-mail]
http://postgis.refractions.net/mailman/listinfo/postgis-users

Thanks

I'm waiting a update from my root. He said me that GEOS'version isn't stable.

EDIT : It's ok now. Thanks you very much