[GRASS-user] how to measure distances in a constrained space

[Hamish]

Hamish escribió:
>> - See the v.surf.icw script in wiki addons. it is like IDW interpolation
>> but replaces euclidean distance with true (cost) distance and goes
>> around corners/coastlines.

Moritz:
> Interesting, but:
> "Input data points [...] but should be kept within a few dozen as the
> module becomes very computationally expensive"

Partial weights are constructed by running r.cost starting from each
input point. A much faster r.cost will mean it is practical to run it
with many more points, but disk space/IO may still be a problem.
(mapset on ramdisk solution)

I will have some new data to process with it in the next weeks, but
perhaps ~200 points might be ok now. Of course resolution matters as much
as number of starting points, I speak of 2000x2000 or so raster region
sizes.


Hamish