Winding Number

I have shapefiles for the state climatic divisions for the United States
and read.shape brings them in wonderfully.  Now I wish to run through a
list of several thousand observation sites to find out in which division
each is located.  I figure that I can compute the winding number for each
site and be done.  However a search doesn't find any references and I
can't find a winding number function among the map/tools/stats.  I have
the code for an efficient C++ but was expecting that it would already be
implemented as an R function

Since I haven't used the map/tools/stats collection before, I suspect I'm 
overlooking the function and would be thankful for a pointer.

This is work in progress - where all good ideas and contributions will be
welcome. If you look on http://sourceforge.net/projects/r-spatial/, you
will see an "alpha" package called "sp", which already has a
point-in-polygon facility, but which may not scale up to the kinds of data
volumes you have, but which invites a spatial query (match polygon?)
function between a SpatialDataFrame with point coordinates and a
SpatialDataPolygons object (sometime). This is only as source packages so 
far, so Windows binaries are not yet available.

I have used the splancs package inout() function before, trying the points
coordinate matrix on each polygon in turn; splancs is available as a
Windows binary. This ought to be less "rough at the edges", and in time
will be. The immediate solution is to use splancs, but this will not work
if the Shapes have multiple polygons. There is a more specialised list for
these kinds of questions in addition to r-help:

https://www.stat.math.ethz.ch/mailman/listinfo/r-sig-geo

and since yesterday (thanks to Jonathan Baron), its archives are also 
searchable from:

http://finzi.psych.upenn.edu/search.html

This question hasn't come up there, but maybe we could move further 
discussion there - posting only for subscribers?
TIA

Clint

Roger Bivand
Economic Geography Section, Department of Economics, Norwegian School of
Economics and Business Administration, Breiviksveien 40, N-5045 Bergen,
Norway. voice: +47 55 95 93 55; fax +47 55 95 93 93
e-mail: Roger.Bivand at nhh.no
Roger,

Thanks for your reference.  Since I can get the polygon coordinates (and 
have the coordinates of my sites, I can cobble together a function that 
will do the trick.

Again, thanks,

Clint

On Fri, 2 Apr 2004, Clint Bowman wrote:

I have shapefiles for the state climatic divisions for the United States
and read.shape brings them in wonderfully.  Now I wish to run through a
list of several thousand observation sites to find out in which division
each is located.  I figure that I can compute the winding number for each
site and be done.  However a search doesn't find any references and I
can't find a winding number function among the map/tools/stats.  I have
the code for an efficient C++ but was expecting that it would already be
implemented as an R function

Since I haven't used the map/tools/stats collection before, I suspect I'm 
overlooking the function and would be thankful for a pointer.

This is work in progress - where all good ideas and contributions will be
welcome. If you look on http://sourceforge.net/projects/r-spatial/, you
will see an "alpha" package called "sp", which already has a
point-in-polygon facility, but which may not scale up to the kinds of data
volumes you have, but which invites a spatial query (match polygon?)
function between a SpatialDataFrame with point coordinates and a
SpatialDataPolygons object (sometime). This is only as source packages so 
far, so Windows binaries are not yet available.

I have used the splancs package inout() function before, trying the points
coordinate matrix on each polygon in turn; splancs is available as a
Windows binary. This ought to be less "rough at the edges", and in time
will be. The immediate solution is to use splancs, but this will not work
if the Shapes have multiple polygons. There is a more specialised list for
these kinds of questions in addition to r-help:

https://www.stat.math.ethz.ch/mailman/listinfo/r-sig-geo

and since yesterday (thanks to Jonathan Baron), its archives are also 
searchable from:

http://finzi.psych.upenn.edu/search.html

This question hasn't come up there, but maybe we could move further 
discussion there - posting only for subscribers?

TIA

Clint

Clint Bowman			INTERNET:	clint at ecy.wa.gov
Air Quality Modeler		INTERNET:	clint at math.utah.edu
Department of Ecology		VOICE:		(360) 407-6815
PO Box 47600			FAX:		(360) 407-7534
Olympia, WA 98504-7600
Clint,

Could you please report back to the r-sig-geo list on how you got this 
done so that others can benefit?

Roger

Roger,

Thanks for your reference.  Since I can get the polygon coordinates (and 
have the coordinates of my sites, I can cobble together a function that 
will do the trick.

Again, thanks,

Clint

On Fri, 2 Apr 2004, Roger Bivand wrote:

On Fri, 2 Apr 2004, Clint Bowman wrote:

I have shapefiles for the state climatic divisions for the United States
and read.shape brings them in wonderfully.  Now I wish to run through a
list of several thousand observation sites to find out in which division
each is located.  I figure that I can compute the winding number for each
site and be done.  However a search doesn't find any references and I
can't find a winding number function among the map/tools/stats.  I have
the code for an efficient C++ but was expecting that it would already be
implemented as an R function

Since I haven't used the map/tools/stats collection before, I suspect I'm 
overlooking the function and would be thankful for a pointer.

This is work in progress - where all good ideas and contributions will be
welcome. If you look on http://sourceforge.net/projects/r-spatial/, you
will see an "alpha" package called "sp", which already has a
point-in-polygon facility, but which may not scale up to the kinds of data
volumes you have, but which invites a spatial query (match polygon?)
function between a SpatialDataFrame with point coordinates and a
SpatialDataPolygons object (sometime). This is only as source packages so 
far, so Windows binaries are not yet available.

I have used the splancs package inout() function before, trying the points
coordinate matrix on each polygon in turn; splancs is available as a
Windows binary. This ought to be less "rough at the edges", and in time
will be. The immediate solution is to use splancs, but this will not work
if the Shapes have multiple polygons. There is a more specialised list for
these kinds of questions in addition to r-help:

https://www.stat.math.ethz.ch/mailman/listinfo/r-sig-geo

and since yesterday (thanks to Jonathan Baron), its archives are also 
searchable from:

http://finzi.psych.upenn.edu/search.html

This question hasn't come up there, but maybe we could move further 
discussion there - posting only for subscribers?

TIA

Clint

Roger Bivand
Economic Geography Section, Department of Economics, Norwegian School of
Economics and Business Administration, Breiviksveien 40, N-5045 Bergen,
Norway. voice: +47 55 95 93 55; fax +47 55 95 93 93
e-mail: Roger.Bivand at nhh.no
Members of this list may be interested in
http://www.vividsolutions.com/jts/jtshome.htm. There is a C++ port asw
well by the postgis folks.

Tim
Roger,

Thanks for your reference.  Since I can get the polygon coordinates (and 
have the coordinates of my sites, I can cobble together a function that 
will do the trick.

Again, thanks,

Clint

On Fri, 2 Apr 2004, Roger Bivand wrote:

On Fri, 2 Apr 2004, Clint Bowman wrote:

I have shapefiles for the state climatic divisions for the United States
and read.shape brings them in wonderfully.  Now I wish to run through a
list of several thousand observation sites to find out in which division
each is located.  I figure that I can compute the winding number for each
site and be done.  However a search doesn't find any references and I
can't find a winding number function among the map/tools/stats.  I have
the code for an efficient C++ but was expecting that it would already be
implemented as an R function

Since I haven't used the map/tools/stats collection before, I suspect I'm 
overlooking the function and would be thankful for a pointer.

This is work in progress - where all good ideas and contributions will be
welcome. If you look on http://sourceforge.net/projects/r-spatial/, you
will see an "alpha" package called "sp", which already has a
point-in-polygon facility, but which may not scale up to the kinds of data
volumes you have, but which invites a spatial query (match polygon?)
function between a SpatialDataFrame with point coordinates and a
SpatialDataPolygons object (sometime). This is only as source packages so 
far, so Windows binaries are not yet available.

I have used the splancs package inout() function before, trying the points
coordinate matrix on each polygon in turn; splancs is available as a
Windows binary. This ought to be less "rough at the edges", and in time
will be. The immediate solution is to use splancs, but this will not work
if the Shapes have multiple polygons. There is a more specialised list for
these kinds of questions in addition to r-help:

https://www.stat.math.ethz.ch/mailman/listinfo/r-sig-geo

and since yesterday (thanks to Jonathan Baron), its archives are also 
searchable from:

http://finzi.psych.upenn.edu/search.html

This question hasn't come up there, but maybe we could move further 
discussion there - posting only for subscribers?

TIA

Clint

Timothy H. Keitt
Section of Integrative Biology
University of Texas at Austin
http://www.keittlab.org/