Subject: CGAL users discussion list
List archive
- From: Monique Teillaud <>
- To:
- Subject: Re: [cgal-discuss] Re: Points, Lines, and Polygons on Spheres
- Date: Wed, 28 Apr 2010 18:06:50 +0200
infblnpf wrote:
Thank you for the information.
General circular arcs unfortunately help me a little only, since I have to
work with great circles (or segments of those) only.
The spherical kernel offers general arcs, it is not restricted to your specific needs, but you can still define arcs of great circles as a particular case. So, it seems that these circular arcs are helping you, even though you are not using them in their full generality.
What I meant by the geometric operations was the following: Lets say I would
like to draw a line from Seattle to Moscow. For each my intention was to
simply define the cities' coordinates and the line is drawn for me. As far
as I understand I would though have to define a plane (or other sphere) in
CGAL first. But this second object is based on Euclidean space, which means,
that I would first have to determine the Euclidean coordinates from my
geographic coordinates.
In any case, you already have to do something similar with the cities' coordinates, don't you?
If I understand well, a city is given in some kind of spherical coordinates (angles). This in fact defines a line in the 3D Euclidean space that contains the center of the sphere.
Each line defines a point of type Circular_arc_point_3 of the spherical kernel, through intersection with the sphere.
The two lines in R^3 corresponding to two cities define a plane in R^3. This is the plane than can be used to define the arc of great circles between the two cities.
(In fact a potential problem I see is that CGAL assumes the lines and planes to have a rational equation in R^3. (I am putting things in simple terms, more details are given in the manual.) But if the sines and cosines defining your city are given as doubles, then this is in fact already some kind of rational approximation, so, our rational framework should hopefully not be a problem.)
This would still leave me with more geometric operations necessary.
It depends what you call "necessary". Of course a software especially tuned at your specific needs would, by definition, be easier to use for you. Since the CGAL spherical kernel is more general, then it is "necessary" to perform some operations to define your specific objects.
That said, as I already mentioned, we might think of adding some functionality in the future, to offer a simpler interface for such a use, which is probably of general interest for geographic applications.
(As usual, this is a matter of manpower...)
--
Monique Teillaud
INRIA Sophia Antipolis - Méditerranée
http://www.inria.fr/sophia/members/Monique.Teillaud/
- [cgal-discuss] Points, Lines, and Polygons on Spheres, infblnpf, 04/28/2010
- Re: [cgal-discuss] Points, Lines, and Polygons on Spheres, Monique Teillaud, 04/28/2010
- [cgal-discuss] Re: Points, Lines, and Polygons on Spheres, infblnpf, 04/28/2010
- Re: [cgal-discuss] Re: Points, Lines, and Polygons on Spheres, Monique Teillaud, 04/28/2010
- Re: [cgal-discuss] Re: Points, Lines, and Polygons on Spheres, Sebastien Loriot (GeometryFactory), 04/28/2010
- Re: [cgal-discuss] Re: Points, Lines, and Polygons on Spheres, Monique Teillaud, 04/28/2010
- [cgal-discuss] Re: Points, Lines, and Polygons on Spheres, infblnpf, 04/28/2010
- Re: [cgal-discuss] Points, Lines, and Polygons on Spheres, Monique Teillaud, 04/28/2010
Archive powered by MHonArc 2.6.16.