[GRASS-user] SamplingStrata

[Markus Neteler]

On Thu, Apr 3, 2014 at 5:15 PM, Matteo Mura <mur.teo at gmail.com> wrote:
> Dear all,
>
> I am going to quickly explain you what I am trying to achieve.
>
> I have a grid of 4693 hexagons (see
> https://www.dropbox.com/s/rkim7v21i1p0bet/grid.jpeg). Each hexagon has
> its univocal id (ID_ES).
>
> We are planning a 2-phase sampling procedure. In the first phase we
> identified some feature inside each hexagon (namely trees out of
> forest features). Now we should carry out the second phase with ground
> survey. Obviously we cannot visit each hexagon but just a sample of
> them.
>
> In doing so we need to divide the hexagon grid in strata composed by
> compact (i.e. not too much elongated in vertical or horizontal
> direction) groups of adjacent hexagons, of course without replacement,
> inside of each strata an hexagon will be randomly choosen and visited.

... FWIW, as of today (kudos to Markus Metz) you can also contrain
v.random to single polygons:

http://grass.osgeo.org/grass70/manuals/v.random.html#examples

In your case you could generate a sufficiently low number of random
points over your hexagon area, then intersect the resulting points
with the hexagons to select them for your analysis. Something like
this :)

markusN