[GRASS-user] v.generalize issue

Markus Metz markus.metz.giswork at googlemail.com
Wed Oct 26 16:38:05 EDT 2011


Roger André wrote:
> Ah, ok.  The causes, please.
> It's a really powerful tool, and I would like to write some additional
> documentation on how to use it for generalizing and smoothing large polygon
> sets, but first I need to make sure that I can anticipate it's behavior a
> little better.  The area attribute swapping was something that I had noticed
> as well, but I figured that was just something to be expected.
> Also, would it be helpful to have additional testers for some of these?  I'm
> not using the raster tools at the moment, but am using the vector tools
> quite a bit (console only though).

Please use GRASS 6.4.2 for testing. A number of bugs present in 6.4.1
have been fixed in 6.4.2.

With regard to v.generalize, randomly swapping of area attributes was
a serious bug, this was by no means to be expected. This swapping
occurred because the original author used area ids instead of
categories to assign attributes to the smoothed vector.  This is
within the logic of the GRASS vector model nonsense because area ids
are randomly assigned, area categories are instead fixed because
user-defined. The disappearance of polygons happened because smoothed
boundaries that violated vector topology were deleted, after that all
boundaries that were topologically incorrect. Boundaries are now only
smoothed/simplified if the modification does not violate vector
topology.

HTH,

Markus M


> Roger
> --
> On Wed, Oct 26, 2011 at 11:51 AM, Markus Metz
> <markus.metz.giswork at googlemail.com> wrote:
>>
>> Roger André wrote:
>> > Hi Markus,
>> > That's good to know.  Can you tell me the bug number?  I'd like to
>> > review
>> > the details of it.
>>
>> There is no bug number, at least I could not find a corresponding bug
>> number. There were a few complaints about v.generalize in the mailing
>> lists. In short, you are not the first one to discover this bug.
>> Additionally, v.generalize was randomly swapping area attributes. This
>> is also fixed.
>>
>> Since you are asking about a bug number, are you interested in the
>> effects or the causes?
>>
>> Markus M
>>
>>
>> > Thanks,
>> > Roger
>> > --
>> >
>> > On Wed, Oct 26, 2011 at 12:23 AM, Markus Metz
>> > <markus.metz.giswork at googlemail.com> wrote:
>> >>
>> >> On Tue, Oct 25, 2011 at 11:07 PM, Roger André <randre at gmail.com> wrote:
>> >> > Hello GRASS Community,
>> >> > I'm having a problem that I'm hoping you can help me find the source
>> >> > of.
>> >> > I think that either A) I am trying to use the v.generalize tool
>> >> > incorrectly,
>> >> > B) that my data has some sort of topological problem which I'm
>> >> > unaware
>> >> > of,
>> >> > or C) that there is a bug in v.generalize.  The simplest way to
>> >> > describe
>> >> > what's happening is that I am consistently unable to generalize a
>> >> > large
>> >> > group of contiguous polygons (African states most recently) without
>> >> > losing a
>> >> > large number of polygons from my data.  I have tried a variety of
>> >> > different
>> >> > things, all without success, and now I'd like to ask if someone would
>> >> > be
>> >> > willing to verify that what I'm seeing isn't a problem of my own
>> >> > creation.
>> >> >  I've posted my data files at www.macrogis.net/africa, along with a
>> >> > before-and-after image of the specific problem I've encountered.  I'm
>> >> > using
>> >> > GRASS v 6.4.1 which I downloaded from the main website 2 months ago.
>> >>
>> >> This bug has been fixed in GRASS 6.4.2
>> >>
>> >> Markus M
>> >
>> >
>
>


More information about the grass-user mailing list