> Dear Roger Bivand
> Since I noticed a new version of maptools, I upgraded the maptools and then
> checked the changelog.
> > maptools(changes=T)
> maptools, version 0.4-11, 2005-01-17
> 0.4-10 Function write.linelistShape() added to output a list of lines to a
> 0.4-10 shapefile, as suggested by Patrick Giraudoux.
> So, no changelog of version 0.4.11 was given. It is nice if you show
> us the change.
Thanks for keeping me alert! The entry is: "Adding checks for length()
being zero", so no changes to functionality. The problem is that when
length(x) is 0, 1:0 produces indices for the loop for (i in 1:length(x)),
but should omit the loop if length(x) is 0. 0.4-12 will be coming soon,
and will have the required information in the changes file.
Can I ask whether I should continue to provide the current plot order
heurisitic as an option - I am considering dropping the present scheme,
which tries to guess the ring direction and plot order of polygons and
sub-polygons form their inclusion inside each other. The reason for
dropping it as default (for Map2poly(..., raw=FALSE)) is a user report of
an otherwise valid shapefile generating a plot order which was corrupted
(had one polygon of 1750 missing).
The new approach is to force all single polygons to be clockwise, and to
plot them from largest to smallest. For multipolygons, the largest
sub-polygon will be clockwise, and plotted first, and included
sub-polygons will have their ring directions switched to anti-clockwise if
included in another subpolygon of the multipolygon.
This scheme will also be used in the up-coming sp package, so please
express your views now!
> Susumu Tanimura
> Dept. of Socio-environmental Medicine
> Inst. of Tropical Medicine, Nagasaki University
> R-sig-Geo mailing list
> R-sig-Geo at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo >
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
Department of Economics
Norwegian School of Economics
N-5045 Bergen, Norway
> Thanks for keeping me alert! The entry is: "Adding checks for length()
> being zero", so no changes to functionality. The problem is that when
> length(x) is 0, 1:0 produces indices for the loop for (i in
> 1:length(x)), but should omit the loop if length(x) is 0. 0.4-12 will be
> coming soon, and will have the required information in the changes file.
As for the new plotting approach, since my knowledge of data
structure in polygons and shapefile is too poor to understand it
(there are so many unclear points), I cannot help you. Hope other
members in this list provide valuable suggestion.
Dept. of Socio-environmental Medicine
Inst. of Tropical Medicine, Nagasaki University