[R-sig-Geo] Is there a function to correct arc-node topology of network

[Barry Rowlingson]

On Thu, Sep 9, 2010 at 5:33 PM, Roger Bivand <Roger.Bivand at nhh.no> wrote:

> Barry,
>
> Are you thinking of the excellent:
>
> http://www2.research.att.com/areas/stat/doc/93.2.ps
>
> http://www2.research.att.com/areas/stat/doc/95.2.ps
>
> mostly in awk and shellscripts in the second report? Then the building
> compiled programs that ship with the source maps package? They are elegant,
> but do involve manual editing of text files to remove dangles, etc. Not a
> menu in sight!
>

 Yes, thats the stuff. I can't imagine they'll be too quick, since I
don't think they use spatial indexing or any craft trick to improve on
O(N^2) performance when looking at intersections of segments - but
maybe there's some sorting going on...

Barry