[GRASSLIST:4414] v.to.sites error

Maciek_Sieczka · September 24, 2004, 6:16pm

Hi!

It seems that v.to.sites fails to translate some points near the nodes.

Ater doing as follows:

GRASS:~ > v.to.sites -a -i input=rzezba output=site_rzezba dmax=1

v.to.sites:

Number of vector nodes = 138

Number of sites generated = 162789
[Raster MASK present]

I get a result a piece of which you can see on the vtosites.png, where:
-RED is the input vector file "rzezba" (height contour lines and height
sites)
-WHITE is the output site file site_rzezba
-GREEN arrows indicate sections of the input vector not translated to
sites - near the nodes

Unclosed line's endings are not translated either.

This error does not depend on dmax.

DEVELOPERS! Should I give a bug report?

Maciek

P.S.
I wonder if v.surf.rst uses a smililar way to translate the input vector
lines to points thus commit the same error?

Maciek

(attachments)

Radim_Blazek · September 27, 2004, 8:14am

On Friday 24 September 2004 20:16, Maciek Sieczka wrote:

Hi!

It seems that v.to.sites fails to translate some points near the nodes.

Ater doing as follows:

GRASS:~ > v.to.sites -a -i input=rzezba output=site_rzezba dmax=1

v.to.sites:

Number of vector nodes = 138

Number of sites generated = 162789
[Raster MASK present]

I get a result a piece of which you can see on the vtosites.png, where:
-RED is the input vector file "rzezba" (height contour lines and height
sites)
-WHITE is the output site file site_rzezba
-GREEN arrows indicate sections of the input vector not translated to
sites - near the nodes

Unclosed line's endings are not translated either.

This error does not depend on dmax.

DEVELOPERS! Should I give a bug report?

Maciek

P.S.
I wonder if v.surf.rst uses a smililar way to translate the input vector
lines to points thus commit the same error?

Similar maybe, but not the same. Are you sure, that those lines have a category?

Radim

Maciek_Sieczka · September 27, 2004, 10:08pm

> It seems that v.to.sites fails to translate some points near the nodes.
> (...)
> Unclosed line's endings are not translated either.
>
> This error does not depend on dmax.
>
> DEVELOPERS! Should I give a bug report?
>
> Maciek
>
> P.S.
> I wonder if v.surf.rst uses a smililar way to translate the input vector
> lines to points thus commit the same error?

Radim wrote:

Similar maybe, but not the same. Are you sure, that those lines have a
category?

yes I'm sure
digitized and labelled in v.digit

each closed line has one node
the closed lines fail to convert to sites near that node

open lines have two nodes - one on each ending
open lines fail to convert to sites near the nodes on the line's endings

I worked the error around doing v.to.rast and then r.to.sites

Maciek

Maciek_Sieczka · September 28, 2004, 8:16pm

Radim,

one new thing about this error

v.to.sites fails near the nodes as described before if the vector lines are
prepared in v.digit but it doesn't when the vector lines are produced using
the r.contour on a dem raster

please see the attached vtosites_2.png for comaprison, where
VIOLET - vector lines
WHITE - sites produced with v.to.sites

the inner circle was digitized and labelled in v.digit, then v.to.sites
the outer circle was derived from a DEM with r.contour, then v.to.sites

could it be that v.digit is guilty?

(all refers to Grass5.3)

Maciek

----- Original Message -----
From: "Maciek Sieczka" <werchowyna@pf.pl>
To: <grasslist@baylor.edu>
Sent: Tuesday, September 28, 2004 12:08 AM
Subject: [GRASSLIST:4419] Re: v.to.sites error

> > It seems that v.to.sites fails to translate some points near the

nodes.

> > (...)
> > Unclosed line's endings are not translated either.
> >
> > This error does not depend on dmax.
> >
> > DEVELOPERS! Should I give a bug report?
> >
> > Maciek
> >
> > P.S.
> > I wonder if v.surf.rst uses a smililar way to translate the input

vector

> > lines to points thus commit the same error?

Radim wrote:
> Similar maybe, but not the same. Are you sure, that those lines have a
> category?

yes I'm sure
digitized and labelled in v.digit

each closed line has one node
the closed lines fail to convert to sites near that node

open lines have two nodes - one on each ending
open lines fail to convert to sites near the nodes on the line's endings

I worked the error around doing v.to.rast and then r.to.sites

Maciek

(attachments)

Radim_Blazek · September 29, 2004, 8:35am

Categories in 5.0/5.3 can be easily lost and labeling
algorithm can differ in modules.
Important is output from v.what.vect on that part of line.

Radim

On Tuesday 28 September 2004 22:16, Maciek Sieczka wrote:

Radim,

one new thing about this error

v.to.sites fails near the nodes as described before if the vector lines are
prepared in v.digit but it doesn't when the vector lines are produced using
the r.contour on a dem raster

please see the attached vtosites_2.png for comaprison, where
VIOLET - vector lines
WHITE - sites produced with v.to.sites

the inner circle was digitized and labelled in v.digit, then v.to.sites
the outer circle was derived from a DEM with r.contour, then v.to.sites

could it be that v.digit is guilty?

(all refers to Grass5.3)

Maciek

----- Original Message -----
From: "Maciek Sieczka" <werchowyna@pf.pl>
To: <grasslist@baylor.edu>
Sent: Tuesday, September 28, 2004 12:08 AM
Subject: [GRASSLIST:4419] Re: v.to.sites error

> > > It seems that v.to.sites fails to translate some points near the

nodes.

> > > (...)
> > > Unclosed line's endings are not translated either.
> > >
> > > This error does not depend on dmax.
> > >
> > > DEVELOPERS! Should I give a bug report?
> > >
> > > Maciek
> > >
> > > P.S.
> > > I wonder if v.surf.rst uses a smililar way to translate the input

vector

> > > lines to points thus commit the same error?
>
> Radim wrote:
> > Similar maybe, but not the same. Are you sure, that those lines have a
> > category?
>
> yes I'm sure
> digitized and labelled in v.digit
>
> each closed line has one node
> the closed lines fail to convert to sites near that node
>
> open lines have two nodes - one on each ending
> open lines fail to convert to sites near the nodes on the line's endings
>
> I worked the error around doing v.to.rast and then r.to.sites
>
> Maciek

Venkatesh_Raghavan · September 29, 2004, 9:56am

Dear All,

It was wonderful meeting some of you at the
FOSS/GRASS 2004 Conference held in Bangkok between
12-14 Sept, 2004.

FOSS/GRASS 2004 Conference was a great success. Many
thanks to our numerous supporters and
Local Organizing Committee at Chulalongkorn University,
Thailand for the excellent arrangements.

The FOSS/GRASS attracted more that 100 participants working
in organizations in more than 16 countries (Thailand, Japan,
Italy, Indonesia, Malaysia, Canada, Spain, Philipines
US, India, Iran, Australia, Switzerland, Germany, Hong Kong, France).

The opening ceremony was graced by the Dean Faculty of
Engineering, Chulalongkorn University and the Director
of the GeoInformatics and Space Technology Development
Agency (GISTDA), Thailand.

Apart from the Keynote session with four invited
speakers (David Hastings, Dave McIlhagga, Hideo Nakano
and Jim Westervelt). A total of 25 Oral presentation and
16 poster presentations were made. A CD-ROM proceedings
was also brought out.

Full papers presented at the conference are available online
at
http://gisws.media.osaka-cu.ac.jp/grass04/papers.php?first_letter=all

Special Issues of selected papers presented at the FOSS/GRASS
2004 conference is planned to be published in in a scientific
journal in early 2003. We are looking for advertisers for the special
issue. Anyone on the list who maybe intrested to advertise his/her
company or product in the special issue can contact me off the list.

FOSS/GRASS 2006 is most likey to be held in Europe as suggested by
participants of FOSS/GRASS 2004. The exact venue is likely to be announced in the next few weeks.

Will be posting some video footage and pictures
taken at FOSS/GRASS 2004 on the Conference web-site soon.

Looking forward to your continued cooperation and support for
the FOSS/GRASS 2006 and future meetings too.

Raghavan

--------------
Dr. Venkatesh Raghavan
Co-Coordinator FOSS/GRASS 2004

Maciek_Sieczka · September 29, 2004, 9:36pm

Categories in 5.0/5.3 can be easily lost and labeling
algorithm can differ in modules.
Important is output from v.what.vect on that part of line.

Radim

v.what.vect?

if you mean an output from d.what.vect then it's ok within the whole line
lenght including the part where v.to.sites fails

also when I display labelled lines and their lables in v.digit all is fine

but after the v.to.sites such line is not converted near the node

here are two more pictures to see how is the node located (YELLOW on
vdigit.png) in respect to the v.to.sites result (WHITE sites on the
dsitesdvect.png)

On Tuesday 28 September 2004 22:16, Maciek Sieczka wrote:
> Radim,
>
> one new thing about this error
>
> v.to.sites fails near the nodes as described before if the vector lines

are

> prepared in v.digit but it doesn't when the vector lines are produced

using

> the r.contour on a dem raster
>
> please see the attached vtosites_2.png for comaprison, where
> VIOLET - vector lines
> WHITE - sites produced with v.to.sites
>
> the inner circle was digitized and labelled in v.digit, then v.to.sites
> the outer circle was derived from a DEM with r.contour, then v.to.sites
>
> could it be that v.digit is guilty?
>
> (all refers to Grass5.3)
>
> Maciek
>
> ----- Original Message -----
> From: "Maciek Sieczka" <werchowyna@pf.pl>
> To: <grasslist@baylor.edu>
> Sent: Tuesday, September 28, 2004 12:08 AM
> Subject: [GRASSLIST:4419] Re: v.to.sites error
>
> > > > It seems that v.to.sites fails to translate some points near the
>
> nodes.
>
> > > > (...)
> > > > Unclosed line's endings are not translated either.
> > > >
> > > > This error does not depend on dmax.
> > > >
> > > > DEVELOPERS! Should I give a bug report?
> > > >
> > > > Maciek
> > > >
> > > > P.S.
> > > > I wonder if v.surf.rst uses a smililar way to translate the input
>
> vector
>
> > > > lines to points thus commit the same error?
> >
> > Radim wrote:
> > > Similar maybe, but not the same. Are you sure, that those lines have

a

> > > category?
> >
> > yes I'm sure
> > digitized and labelled in v.digit
> >
> > each closed line has one node
> > the closed lines fail to convert to sites near that node
> >
> > open lines have two nodes - one on each ending
> > open lines fail to convert to sites near the nodes on the line's

endings

> >
> > I worked the error around doing v.to.rast and then r.to.sites
> >
> > Maciek

(attachments)

Camilo_Alcantara · September 30, 2004, 3:06am

Hello
I exported a vector coverage with v.out.ascii and it seems to export
very well, but when I try to import the files with v.in.ascii I got only
the vectors without centroids, even when I placed properly both files
(the ascii polygon file and the auxiliar one with the centroids)
in the dig_ascii and dig_att with the same name (just like the manual page
said).

How can I attach my centroids file to the vector coverage?

I'm using grass57_exp_2004_09_25 on one AMD 1.6 GZ with debian.

--
Pedro Camilo Alcántara Concepción
-----------------
camilo@pcbiol.posgrado.unam.mx
camiloalcantara@hotmail.com
-----------------
pcbiol.posgrado.unam.mx Wednesday Sep 29 2004 20:05:00 CST
-----------------
Algún día la ciencia puede llegar a tener la vida del hombre en sus manos y, haciendo estallar el mundo la especie humana puede incurrir en un suicidio colectivo.
-- Henry Adams. (1838-1918) Escritor e historiador estadounidense.