CGAL users discussion list

[Sebastien Loriot <>]

Sorry I thought you were using "inexact_constructions".

Then using
std::cout << CGAL::exact(p) << " " << CGAL::exact(q) << " " << 
CGAL::exact(r) << "\n";

You will see that you have different numbers. You can also
call q == CGAL::midpoint(p, r) ? "equal" : "not equal" if you want.

By default operator<< is giving a approximation but not the exact value
(it is a filtered number type).

Best,

Sebastien.

On 3/12/21 2:34 PM, Aomandeyi ( via cgal-discuss 
Mailing List) wrote:
> Aha! i can give you counter-example:
> ------------------------------
> #include <CGAL/Exact_predicates_exact_constructions_kernel.h>
> #include <sstream>
> typedef CGAL::Exact_predicates_exact_constructions_kernel Kernel;
> typedef Kernel::Point_2 Point_2;
> int main()
> {
>         Point_2 p(0, 0.4), q(1, 0.7), r(2, 1);
>         {
>                 std::cout << std::setprecision(17) << p << " " << q << " " << r << 
> "\n";
>                 std::cout << (CGAL::collinear(p, q, r) ? "collinear\n" : "not
> collinear\n");
>         }
>         
>         {
>                 q = CGAL::midpoint(p, r);
>                 std::cout << std::setprecision(17) << p << " " << q << " " << r << 
> "\n";
>                 std::cout << (CGAL::collinear(p, q, r) ? "collinear\n" : "not
> collinear\n");
>         }
>         
>
>         return 0;
> }
> --------------------
> the results:the same input but different output
> ----------------------------------------
> 0 0.40000000000000002 1 0.69999999999999996 2 1
> not collinear
> 0 0.40000000000000002 1 0.69999999999999996 2 1
> collinear
> ---------------------------------------------
>
>
>
> --
> Sent from: http://cgal-discuss.949826.n4.nabble.com/