CGAL users discussion list

[Dylan Lawrence <>]

After a lot of wrestling I was able to get my code working, but I think I'm missing something, below is my code and then the output

Line_2 l(Point_2(2, 4), Point_2(6, 8));
Circle_2 c(Point_2(3, 7),3);

std::vector<IntersectOut> output;

typedef CGAL::Dispatch_output_iterator < CGAL::cpp11::tuple<IntersectOut>, CGAL::cpp0x::tuple<std::back_insert_iterator<std::vector<IntersectOut>>>> Dispatcher;

Dispatcher disp = CGAL::dispatch_output<IntersectOut>(std::back_inserter(output));
CGAL::intersection(l, c, disp);

if (output.size() > 0)
{
std::cout << CGAL::to_double(output[0].first.x()) << ", " << CGAL::to_double(output[0].first.y()) << std::endl;
std::cout << output[0].second << std::endl;
if (output.size() > 1)
{
std::cout << CGAL::to_double(output[1].first.x()) << ", " << CGAL::to_double(output[1].first.y()) << std::endl;
}
}

Output

3.29289, 5.29289

1 

 4.70711, 6.70711

However on Wolfram using that circle and line it says the intersections should occur at 

((4-sqrt(7/2), 6-sqrt(7/2)) | (4+sqrt(7/2), 6+sqrt(7/2)))~~((2.12917, 4.12917) | (5.87083, 7.87083))

This seems to be giving a different answer. I checked a few other cases and some align with Wolfram but others are off like this and they seem to be off by more than a floating point error. 

On Tue, Jul 7, 2015 at 2:04 AM, Sebastien Loriot (GeometryFactory) <> wrote:

The type is boost::optional< boost::variant<T..> > where the types in the variant depends on what you call intersection with.
In you case, it can only be std::pair<Circular_arc_point_2<CircularKernel>, unsigned>

http://doc.cgal.org/latest/Kernel_23/group__intersection__circular__grp.html


Have a look at the example on this page how to get the type out of the boost one:
http://doc.cgal.org/latest/Kernel_23/group__intersection__linear__grp.html

Sebastien.

On 07/07/2015 08:56 AM, Dylan Lawrence wrote:

After quite a bit of rewrites I was able to actually get the
intersections running using CGAL::intersection() I know my vector
contains the intersections however what exactly is in that vector. It
seems to be an extension of a boost type but it's unclear what it is.
Essentially I have my intersections, but I do not understand how to
access them to get the point of incidence for an intersection.

On Tue, Jun 23, 2015 at 12:55 PM, Laurent Rineau (CGAL/GeometryFactory)

< <mailto:>> wrote:

    Le Tuesday 23 June 2015 16:22:27 Monique Teillaud a écrit :
    > a Line_arc_2 can be constructed from a Segment_2
    >http://doc.cgal.org/latest/Circular_kernel_2/classCGAL_1_1Line__arc__2.html
    >
    > then you can intersect it with a circle or a circular arc
    >http://doc.cgal.org/latest/Circular_kernel_2/classCircularKernel_1_1DoInters
    > ect__2.html
    > Le 23/06/15 04:22, dlawre14 a écrit :
    > > I've done some searching of the docs but cannot find any entry of a
    > > Circle_2 and Segment_2 intersecting. Can such a thing be done or do I
    > > need to convert the Segment_2 to another type?

    See also :

       - the Intersect_2 functor:

    http://doc.cgal.org/latest/Circular_kernel_2/classCircularKernel_1_1Intersect__2.html

       - the corresponding 'do_intersect' and 'intersection' global
    functions
    overloads:

    http://doc.cgal.org/latest/Kernel_23/group__do__intersect__circular__grp.html
    http://doc.cgal.org/latest/Kernel_23/group__intersection__circular__grp.html

    --
    Laurent Rineau, PhD
    R&D Engineer at GeometryFactory http://www.geometryfactory.com/
    Release Manager of the CGAL Project http://www.cgal.org/


    --
    You are currently subscribed to cgal-discuss.
    To unsubscribe or access the archives, go to
    https://sympa.inria.fr/sympa/info/cgal-discuss





--
Dylan Lawrence
Masters Student
Bioinformatics and Computational Biology

--
You are currently subscribed to cgal-discuss.
To unsubscribe or access the archives, go to
https://sympa.inria.fr/sympa/info/cgal-discuss

--

Dylan Lawrence

Masters Student

Bioinformatics and Computational Biology