[R-sig-Geo] shortest path

[Barry Rowlingson]

On Mon, Sep 10, 2012 at 11:47 AM, Mathieu Rajerison
<mathieu.rajerison at gmail.com> wrote:
> Barry,
>
> I looked at your code and I wondered if it would not be more convenient to
> find the closest point on a an edge rather than the closest vertex?
>
> In this case, what technique would be the best? I thought of
> spatstat::project2Segment that I used in my script (but Edzer says it
> doesn't work - I have to check that)
>

 I wasn't really concerned about that since it was extraneous to the
core job of constructing an igraph from the lines, but should be
trivial.

 There's also possible work on adding tolerances to the nodes and
snapping line ends that don't quite make it to other lines, but that's
all topological cleanup stuff that could expand into cleaning dangles
and jumpbacks and other digitizing errors - I didn't want to get into
all that at the moment! I have other fish to fry...

Barry