[GRASS-user] Buffering large grid maps

[Milton Cezar Ribeiro]

Hi Glynn Clements,

Thanks for your reply. In fact I need to generate a (uncontroled) distance
map from all my values=1.
I confess that I don´t understood how r.grow will help me on this task.
I need to build a for looping and grow and grow many times the new generated
map?
Or is there a a way of simple generate a distance map on one command?

Remembering that I am new in GRASS and that I run under windows.

Thanks in advance,

miltinho astronauta
brazil



2008/7/21, Glynn Clements <glynn at gclements.plus.com>:
>
>
> Milton Cezar Ribeiro wrote:
>
> > I just started to use GRASS, and I am working with a (almost to me) very
> > large dataset into GRASS 6.3.0 NATIVE WINDOWS
> > I imported an Erdas image file using Gdal. Apparently I looks fine,
> because
> > GRASS understood and create a Mapset with all informations contained on
> my
> > IMG file (like projection, number of cells, bounding box, etc.).
> >
> > The raster map is 60,000x48,000 pixels.
> >
> > I tryed display my raster map (it is a binary map with values 0 and 1),
> but
> > I can see the binary map.
> > How can I display my raster map on Map Display 1? I notice that "d.raster
> > teste" send the output to .tmp directory (or something like).
> >
> > But my biggest problem is that I need to generate a Distance Map from
> those
> > pixels which values is 1.
> >
> > I tryed "r.buffer input=test output=teste_buffer distances=50,100" (it is
> in
> > meters)
> > but GRASS return a erros message telling that can´t allocate memory
>
> r.buffer tries to read the entire map into memory, which isn't going
> to work with a map that large (it would need 2.8GB; even if you have
> that amount of RAM, a 32-bit OS probably isn't going to let you have
> that much in a single block).
>
> r.grow doesn't suffer from the memory issues of r.buffer. It only
> keeps in memory as many rows as are necessary given the size of the
> buffer (i.e. 2*radius+1 rows).
>
> It doesn't support multiple distance zones, so you would need to use
> multiple passes and patch the results together with r.patch.
>
> --
> Glynn Clements <glynn at gclements.plus.com>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.osgeo.org/pipermail/grass-user/attachments/20080721/00b4f9c6/attachment.html