[R-sig-Geo] Problem in converting SpatialPolygonsDataFrame to owin object

Roger Bivand Roger.Bivand at nhh.no
Thu Sep 14 12:19:18 CEST 2006


On Thu, 14 Sep 2006, Edzer J. Pebesma wrote:

> Debarchana Ghosh wrote:
> >
> >   
> >  
> 
> >     works fine; I modified the code such that
> >     xx1<-as(xx, "owin")
> >
> >     will work fine as well in the future.
> >     --
> >
> >
> >
> > So I just download and install the updated spspatstat package?
> I said "future", not "near future" or "now"; spspatstat isn't on CRAN. 
> If you wish you could download the source from anonymous cvs on the 
> r-spatial sourceforge project. It is not there where your problem is, 
> however, as you were able to transform to SpatialPolygons anyway.
> > Also, I know attachment is not the right way to converse on this 
> > list-serve but I couldn't think of any other way to show my study 
> > area. So I've attached the image of my study area (polygon). As i've 
> > mentioned in my earlier email, I've read a shapefile in R and created 
> > a object (xx)  of class SpatialPolygonDataFrame. Howver when I try to 
> > convert xx to owin I get the following error. As shown in the 
> > attachment, the lines do not intersect but has 7 contigous US counties.
> >
> > Any solution to this problem will be very helpful.
> The problem may be that polygons are only slightly intersecting (which 
> you cannot find out from a bitmap image), or that some code is making 
> wrong assumptions. The error message comes from the owin function in the 
> spatstat package. If the author of the package doesn't listen to this 
> list, you could contact him directly (I Cc:d this message). To me, at 
> first glance it looked like the error generating code is a fairly recent 
> add to spatstat, but I may be wrong on that.

Is it possible to make the offending shapefiles available to me/others? 
Without them it is very difficult to debug - SpatialPolygons do not check 
for intersections. I would suggest using spgpc (also on sourceforge) to 
clean the polygons, but first we need to find out if they do intersect.

Roger

> --
> Edzer
> 
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
> 

-- 
Roger Bivand
Economic Geography Section, Department of Economics, Norwegian School of
Economics and Business Administration, Helleveien 30, N-5045 Bergen,
Norway. voice: +47 55 95 93 55; fax +47 55 95 95 43
e-mail: Roger.Bivand at nhh.no




More information about the R-sig-Geo mailing list