[GRASSLIST:575] Re:

David Finlayson david.p.finlayson at gmail.com
Wed Apr 5 01:58:21 EDT 2006


If white == 1 and black == 0 and you want all the "green" area
half-way between the white point and the black point, why don't you
try selecting all areas "green" >= 0.5?

Step 1: interpolate raster
Step 2: r.mapcalc "test = if(green >= 0.5, green, null())"

Alternatively, you could buffer the white points by some reasonable
distance and select for all cells within X meters of the white point.

David

On 4/4/06, lponti at infinito.it <lponti at infinito.it> wrote:
> Dear list,
>
> so far, I have been masking either via MASK or r.mapcalc.
> This way, only cells corresponding to MASK would display
> in the masked raster.
>
> In contrast, now I would like cells corresponding to a
> "mask" NOT to display* in the main raster layer. Is there
> a way to do it?
>
> This is part of a larger issue I am addressing and that I
> will briefly describe. In the attached sample image, you
> see points imported via v.in.ascii and used to get the
> green interpolated raster. Values at black points are
> greater than zero, whereas the value of the white point is
> zero (and it is not used for interpolation since it has no
> biological meaning).
>
> The end result would be to clip out a portion of the green
> raster around the white point. This portion would be
> delimited by midpoints between the white point and the
> closest neighboring black points.
>
> So far, I have been trying to get a vector buffer around
> the white point and then v.to.rast for some kind of
> subsequent clipping (described above*). However I have no
> cue on how to get a comlex shape from neighboring points
> as described. Is there a method already in place to do
> this?
>
> Many thanks in advance and any hints much appreciated,
>
> Luigi
>
> grass6/cygwin user
> INFINITO ADSLFLAT 4 MEGA: SOLO 27,90 EURO AL MESE IVA INCLUSA IP STATICO, BANDA GARANTITA 256Kbps ANTIVIRUS E FIREWALL INCLUSI NEL PREZZO
>
> http://adsl.infinito.it
>
>
>
>
>
>


--
David Finlayson




More information about the grass-user mailing list