I think your approach with two runs is adequate, although I am not
sure which tool you would use to extend your polygon as I have not
looked into that issue recently. I have similar questions that I
address. In order to get names of the polygons (conservation areas for
you), I go beyond v.select to populate a field with the names of the
areas being queried. I then dump to a text file that I can open in a
spreadsheet application for summarizing. My example assumes you have a
field called 'name' in your input polygon file:
v.select ainput=in_points binput=conservation_polygons
out=points_in_polygons
v.db.addcol points_in_polygons col="polygon_name VARCHAR(40)"
v.distance from= points_in_polygons to=conservation_polygons dmax=0
upload=to_attr to_column=name col=polygon_name
v.db.select points_in_polygons > points_in_polygons.txt
Hope that helps. Maybe others will have similar or improved solutions,
but this has worked for me.
John
On Mar 4, 2008, at 3:00 AM, Corrado wrote:
Dear John, Rick, Hamish,
thanks for your kind answers.
I explain the problem:
I have a file with some hundreds of species and 58,000 observations.
For each
species, I have the coordinates (X,Y) of the sites where they were
observed.
I also have a shape file, with polygons representing conservation
areas, like
natural reserves.
I need to know for each species, how many point fall inside
conservation areas
and how many points fall outside conservation areas. Also, I need to
run the
same test on the conservation areas with a buffer of 100 m and with
a buffer
of 500 m.
This is what I thought: I test if each point (that is site) is
inside the
polygon (that is conservation area), and I have an answer like yes
or no (or
0 or 1). I write a routine that does test all the points for each
species.
I add the buffer (I do not know how to do it, but I think it is
possible). I
run the test again.
What do you think of this approach?
If I use v.select, would that return me 0 or 1 (or yes or no), when
I try to
overlap one point with the vector map?
Or do you think it would be better to use R?
Or would it be better to prepare a vector map for each species with
all the
sites, and overlap species by species instead of point by point?
I apologise for my questions, but I am trying to use GRASS for the
first time.
I would like to switch from ArcGIS and convince also the other
people in the
research group (there is at least two of us campaigning now).
Thanks!
_______________________________________________
grass-user mailing list
[email protected]
http://lists.osgeo.org/mailman/listinfo/grass-user
