[GRASS-user] obtaining pairwise least-cost distances in meters

[Facundo Muñoz]

  Hi Anders,
I think that this GRASS script (v.costdist.mat) that I did a couple of 
years ago may fit.
You can get it from 
http://www.geeitema.org/guenmap/index.jsp?opcion=resultados&idioma=en.

v.costdist.mat determines de cumulative cost of moving between each pair 
of points from two given maps over a (single) cost surface.
The results are stored in the attribute table of the second set of 
points (hence, you get a matrix of values).
You can only feed a single cost surface (friction map) at a time, so you 
should make a script that loops over your set of friction maps.

Hope it helps.
ƒacu.-



On 25/08/10 15:52, Anders Gonçalves da Silva wrote:
> Hello all.
>
> I have a set of samples in a landscape. I want to create different friction maps based on landscape features (i.e., presence/absence of roads, vegetation type, etc.) to understand animal movement between habitat patches. I have used v.distance to calculate the pairwise distance among the samples (which I assume is the Euclidean distance). But, I would like to incorporate other features.
>
> I have a dem, a vector map with points, and a vector map with vegetation types and other land use types all in one mapset.
>
> I tried to use r.walk to calculate the distance among the sample points based solely on the topography by creating a friction map with all cells =1:
>
> r.walk elevation=dem friction=area1 out=null.cost start_points=points stop_points=points #similarly to what v.distance does, ie, the start points are the same as the end points.
>
> Computations proceed to the end, but when I try to visualise the results, nothing appears on the d.mon.
>
> Ideally, I would end up with different matrices of pairwise distances for each different friction map, which represent different hypotheses about how the animals interact with the landscape.
>
> I am not sure what I am doing wrong, or even if this is the most efficient way of approaching this problem. Any assistance would be appreciated.
>
> Thank you.
>
> Anders._______________________________________________
> grass-user mailing list
> grass-user en lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/grass-user
>