[GRASS5] Voxels/3D rasters in GRASS 6.1

This is another mail in my desperate quest for handling 3D data
in GRASS. I am trying voxels this time and got stuck just as
quickly as with 3D vectors.
Maybe someone could answer some of the questions I have.
I am using GRASS 6.1 CVS (about 2 weeks old).

1. Do all commands that relate to voxels have r3. prefix now? Which
other commands relate to voxels?

2. r3.in.ascii mentions s.to.rast3 is there an equivalent command for
   this in 6.1?

3. When trying to visualise a voxel in nviz, I get:

  Illegal filename. character <> not allowed.G__open(r): mapset (benni)
doesn't match xmapset (ô¯@DHäÿ¿|v@5&òr@ÿÿÿÿ@äÿ)

  on the console. Am I using a broken version of nviz?

4. Why does r3.info not give information about the data range?

5. How do I get actual volume information, do I have to count the
   non NULL cells in the voxel map, then add them up? Is there
   any command that can give me 3D raster voxel statistics?

6. Is there any CURRENT information about voxels and GRASS on the
   web?

I would be very happy indeed about any information on this subject!

Cheers,

Benjamin

On Fri, Jul 01, 2005 at 11:50:57AM +0200, benducke@compuserve.de wrote:

This is another mail in my desperate quest for handling 3D data
in GRASS. I am trying voxels this time and got stuck just as
quickly as with 3D vectors.
Maybe someone could answer some of the questions I have.
I am using GRASS 6.1 CVS (about 2 weeks old).

1. Do all commands that relate to voxels have r3. prefix now? Which
other commands relate to voxels?

Yes: r3.*
- plus v.vol.*
- plus g.* for file/region management (was g3.* in earlier days)

2. r3.in.ascii mentions s.to.rast3 is there an equivalent command for
   this in 6.1?

No. I think that s.to.rast3 still needs to be ported as v.to.rast3.

3. When trying to visualise a voxel in nviz, I get:

  Illegal filename. character <> not allowed.G__open(r): mapset (benni)
doesn't match xmapset (ô¯@DHäÿ¿|v@5&òr@ÿÿÿÿ@äÿ)

  on the console. Am I using a broken version of nviz?

Doesn't look good. I used the Slovakia data set a few days
ago without problems.

4. Why does r3.info not give information about the data range?

AFAIK: No, that's a TODO.

5. How do I get actual volume information, do I have to count the
   non NULL cells in the voxel map, then add them up? Is there
   any command that can give me 3D raster voxel statistics?

There is not too much yet, I believe.

6. Is there any CURRENT information about voxels and GRASS on the
   web?

Should be. Please help to improve the documentation...

I would be very happy indeed about any information on this subject!

Cheers,

Benjamin

Markus

Benjamin,

I got 3-D points to display in NVIZ pretty readily. I was testing this out
several months ago. I suppose that it still works OK, but haven't tested it
lately. The thing that tripped me up for awhile was the fact that my z-scale
(total range of a couple meters) was much less than my xy scale (range of a
couple hundred meters). I had to greatly increase the z exaggeration to see
the 3D points.

For voxels, I had more trouble. I got the Slovakia dataset to work well
once, but not a second time. However, this was several months ago and it may
work better now. Again, there are some tricks. In the case of voxels, you
need to set the isosurfaces to values that actually show something. I did
this by trial and error, but you should be able to do this more
intelligently with some practice.

(Without checking) I think that all voxel commands are now under the r3.*
designation. I don't know why the information you are seeking is missing,
except to say that the true 3D functions are still relatively rough and
experimental compared with the 2D/2.5D in GRASS (as they are in other GIS's
if they are even present at all).

Hope this is helpful.

Michael

On 7/1/05 2:50 AM, "benducke@compuserve.de" <benducke@compuserve.de> wrote:

This is another mail in my desperate quest for handling 3D data
in GRASS. I am trying voxels this time and got stuck just as
quickly as with 3D vectors.
Maybe someone could answer some of the questions I have.
I am using GRASS 6.1 CVS (about 2 weeks old).

1. Do all commands that relate to voxels have r3. prefix now? Which
other commands relate to voxels?

2. r3.in.ascii mentions s.to.rast3 is there an equivalent command for
   this in 6.1?

3. When trying to visualise a voxel in nviz, I get:

  Illegal filename. character <> not allowed.G__open(r): mapset (benni)
doesn't match xmapset (ô¯@DHäÿ¿|v@5&òr@ÿÿÿÿ@äÿ)

  on the console. Am I using a broken version of nviz?

4. Why does r3.info not give information about the data range?

5. How do I get actual volume information, do I have to count the
   non NULL cells in the voxel map, then add them up? Is there
   any command that can give me 3D raster voxel statistics?

6. Is there any CURRENT information about voxels and GRASS on the
   web?

I would be very happy indeed about any information on this subject!

Cheers,

Benjamin

____________________
C. Michael Barton, Professor of Anthropology
School of Human Evolution and Social Change
PO Box 872402
Arizona State University
Tempe, AZ 85287-2402
USA

Phone: 480-965-6262
Fax: 480-965-7671
www: <www.public.asu.edu/~cmbarton>

On Sat, Jul 02, 2005 at 09:16:47PM -0700, Michael Barton wrote:

Benjamin,

I got 3-D points to display in NVIZ pretty readily. I was testing this out
several months ago. I suppose that it still works OK, but haven't tested it
lately. The thing that tripped me up for awhile was the fact that my z-scale
(total range of a couple meters) was much less than my xy scale (range of a
couple hundred meters). I had to greatly increase the z exaggeration to see
the 3D points.

For voxels, I had more trouble. I got the Slovakia dataset to work well
once, but not a second time.

Be sure to get the latest version (27-Apr-2005):

http://mpa.itc.it/grasstutor/data_menu2nd.phtml
-> slovakia3d.tar.gz, Slovakia 3D precipitation data set

However, this was several months ago and it may
work better now. Again, there are some tricks. In the case of voxels, you
need to set the isosurfaces to values that actually show something. I did
this by trial and error, but you should be able to do this more
intelligently with some practice.

Just read the README :slight_smile:
http://mpa.itc.it/grasstutor/data7/slovakia3d_readme.txt

(Without checking) I think that all voxel commands are now under the r3.*
designation. I don't know why the information you are seeking is missing,
except to say that the true 3D functions are still relatively rough and
experimental compared with the 2D/2.5D in GRASS (as they are in other GIS's
if they are even present at all).

I have updated the " GRASS 5.3 modules porting status to 6.0" WIKI page:
http://grass.gdf-hannover.de/twiki/bin/view/GRASS/ModulePortingList#Volume_modules

Markus