[GRASS-user] Spatial Dissolve

Wed May 30 09:07:01 EDT 2007

OK, I'll try again.  Comments below:

> Well, the example is a bit different:
>
> http://grass.itc.it/gdp/html_grass63/v.dissolve.html
> # remove duplicated tile boundaries:
> v.clean clc2000_patched out=clc2000_clean tool=snap,break,rmdupl thresh=.01
>
> Please note that you need tool=snap,break,rmdupl, otherwise it won't
> work. Attributes should survive (they do for me in various maps).

I did try the full v.clean call the first time, and it hung part way
through processing (i.e. stopped responding) so I tried again with
only the rmdupl (incorrectly making the assumption that it was the
most important parameter for this task).  I'll give it a try again now
with the full v.dissolve call, and wait it out.

I ran (v.clean imported out=cleaned tool=snap,break,rmdupl
thresh=0.01).  The longest wait is while processing the "Break Lines
at Intersections" portion of the clean, which runs for ~20 mins on a
very small vector file, although this time it did complete
successfully.

> >  - at this point I noticed that the attribute values are not carried
> >  through and I'm left with only "id" and "cat" - why is this, and will
> >  it cause problems later on?
>
> The problem originates from the incomplete v.clean call, I guess.

This is still an issue, even with the full v.clean call.  When adding
the new "cleaned" vector to my project, I get the error "Warning:
Cannot select attributes from table 'cleaned'"  Is this a GRASS/QGIS
integration problem?  Is there another way to check if the attributes
are there, but not coming through?

> # dissolve based on column attributes:
> v.dissolve input=clc2000_clean output=clc2000_final col=CODE_00

Here's where I run into problems.  I do not have a "col" that I wish
to dissolve on, rather I wish to do a "spatial" dissolve i.e. not
based on attributes.  However, even if I did want to dissolve on
attributes, I would be unable to, as they do not come through the
clean process.  Also, the help for v.dissolve in the version of GRASS
I'm using does not even list "col", as a parameter, rather it has only
'input', 'output' and 'layer.'  Please advise as to the next steps.

> Please try again and let us know where you found the misleading
> documentation (so that we can fix it).

As I mentioned above, there is no misleading documentation, just my
incorrect assumption based on the CORINE example.

Thanks again or your help!  It will be nice to have this figured out!

Cheers,

Darren

> >  On 5/28/07, Markus Neteler <neteler at itc.it> wrote:
> > > Darren,
> > >
> > > could you be more specific about what you did (commands + flags/params),
> > > which GRASS version etc? I just used this method recently again with
> > > success.
> > > The v.patch step in the CORINE example is necessary for *CORINE* since
> > > you get vector tiles. If you have already a patched map, don't use it.
> > >
> > > Markus
> > >
> > > Darren Cope wrote on 05/28/2007 02:01 PM:
> > > > Hi Markus,
> > > >
> > > > I must be missing something, because this still does not seem to work.
> > > > I found another (non-GRASS) way to accomplish this, so the priority
> > > > is not very high, but I'd still like to figure out the GRASS method.
> > > > Is the v.patch step in the CORINE example necessary, or mostly just
> > > > the v.clean step with rmdupl?
> > > >
> > > > Cheers,
> > > >
> > > > Darren
> > > >
> > > > On 5/16/07, Markus Neteler <neteler at itc.it> wrote:
> > > >>
> > > >> Hi Darren,
> > > >>
> > > >> I also stumbled over this problem, but we solved it. The trick is that
> > > >> you first have to remove the duplicated tile boundaries, then dissolve
> > > >> works.
> > > >> See example for CORINE Landcover SHAPE files:
> > > >>  http://grass.itc.it/grass63/manuals/html63_user/v.dissolve.html
> > > >>
> > > >> Cheers
> > > >> Markus
> > > >>
> > > >>
> > > >> Darren Cope wrote:
> > > >> >
> > > >> > Hi all,
> > > >> >
> > > >> > I'm new to GRASS, and using it in Windows via QGIS.
> > > >> >
> > > >> > I am attempting to do a "spatial dissolve" (i.e. not dissolve on an
> > > >> > attribute, but rather just remove unnecessary tile boundaries) on a
> > > >> > file that I've imported to GRASS from a shapefile, and am so far
> > > >> > unsuccessful.  The file contains waterbodies, and some of them are cut
> > > >> > in pieces by tile boundaries.  I would like to remove these
> > > >> > boundaries.  Can someone please help me out?  I've tried combinations
> > > >> > of v.extract -d with no success, and am not sure how the "category"
> > > >> > value apply when I'm not dissolving on an attribute.
> > > >> >
> > > >> > So far my search of the archives has turned up something close, but
> > > >> > nothing that seems to fit my situation exactly.
> > > >> >
> > > >> > Cheers,
> > > >> >
> > > >> > Darren Cope
> > > >> > darrencope at gmail.com
> > > >> > http://dmcope.freeshell.org
> > >
> > >
> > > ------------------
> > > ITC -> dall'1 marzo 2007 Fondazione Bruno Kessler
> > > ITC -> since 1 March 2007 Fondazione Bruno Kessler
> > > ------------------
> > >
> > > _______________________________________________
> > > grassuser mailing list
> > > grassuser at grass.itc.it
> > > http://grass.itc.it/mailman/listinfo/grassuser
> > >
> >
> >
> >  --
> >  Darren Cope
> >  darrencope at gmail.com
> >  http://dmcope.freeshell.org
> >
> >  _______________________________________________
> >  grassuser mailing list
> >  grassuser at grass.itc.it
> >  http://grass.itc.it/mailman/listinfo/grassuser
>
> _______________________________________________
> grassuser mailing list
> grassuser at grass.itc.it
> http://grass.itc.it/mailman/listinfo/grassuser
>

-- 
Darren Cope
darrencope at gmail.com
http://dmcope.freeshell.org