R help - Feb 2012 - SPATIAL QUESTION: HOW TO MAKE POLYGONS AROUND CLUSTERS OF POINTS AND EXTRACT AREAS AND COORDINATES OF THESE POLYGONS?

Imagine that I have a large number of points (given by coordinates x and y) that
vary in density per space. For the purpose of demonstration it could be
generated like this: s <-
data.frame(x=runif(10000,0,900),y=runif(10000,0,900)); plot(s)

I want to create polygons around the points where point density is greater than
a selected threshold (for example, by using krieging or equivalent method). For
these polygons, I want to have the centre coordinates and the size of the area
for further use in analyses.

I would be very grateful if I could be shown the R packages and functions I
should use to accomplish this, and even some outline of the code. Is it
possible?

Best regards
Bjørn


	[[alternative HTML version deleted]]

I would suggest asking this question on r-sig-geo.
-Don

-- 
Don MacQueen

Lawrence Livermore National Laboratory
7000 East Ave., L-627
Livermore, CA 94550
925-423-1062





On 2/3/12 6:59 AM, "Bj?rn ?kland" <okb at skogoglandskap.no>
wrote:
>Imagine that I have a large number of points (given by coordinates x and
>y) that vary in density per space. For the purpose of demonstration it
>could be generated like this: s <-
>data.frame(x=runif(10000,0,900),y=runif(10000,0,900)); plot(s)
>
>I want to create polygons around the points where point density is
>greater than a selected threshold (for example, by using krieging or
>equivalent method). For these polygons, I want to have the centre
>coordinates and the size of the area for further use in analyses.
>
>I would be very grateful if I could be shown the R packages and functions
>I should use to accomplish this, and even some outline of the code. Is it
>possible?
>
>Best regards
>Bj?rn
>
>
>	[[alternative HTML version deleted]]
>

On 04/02/12 03:59, Bj?rn ?kland wrote:
> Imagine that I have a large number of points (given by coordinates x and y)
that vary in density per space. For the purpose of demonstration it could be
generated like this: s<-
data.frame(x=runif(10000,0,900),y=runif(10000,0,900)); plot(s)
>
> I want to create polygons around the points where point density is greater
than a selected threshold (for example, by using krieging or equivalent method).
For these polygons, I want to have the centre coordinates and the size of the
area for further use in analyses.
>
> I would be very grateful if I could be shown the R packages and functions I
should use to accomplish this, and even some outline of the code. Is it
possible?
You can do this using the spatstat package.  E.g.:

     require(spatstat)
     set.seed(42)
     X <- rpoisspp(100)
     Y <- density(X)
     Z <- levelset(Y,130,">")
     U <- connected(Z)
     V <- tiles(tess(image=U))
     lapply(V,centroid.owin)
     lapply(V,area.owin)

You should read the help on the functions invoked above.  It might be
illuminating to do some plotting along the way (of X, Y, Z, U, and V).

     cheers,

         Rolf Turner

It has been pointed out to me (in a private email) that there was
a typo in my previous post;

     X <- rpoisspp(100)

should read

     X <- rpoispp(100)

(i.e. just one "s").  Sorry 'bout that! :-)

     cheers,

         Rolf