CGAL users discussion list

[Sylvain Pion <>]

Hi Himanshu,

On this particular subject, here is another « mooc » I made, which you may 
find of interest :
https://www.youtube.com/watch?v=7cz7RNYoWow

Regards,
Sylvain

> Le 19 nov. 2020 à 10:06, Sebastien Loriot (GeometryFactory) 
> ( via cgal-discuss Mailing List) 
> <> a écrit :
> 
> CGAL relies on an multiprecision number type.
> They are several alternative but the most used is the third party GMP.
> Then, since exact arithmetic is costly, they are several filters applied
> to avoid relying on this multiprecision number type (static filters,
> interval arithmetic, ...).
> 
> You can watch the google tech talk starting from here:
> https://youtu.be/3DLfkWWw_Tg?t=2310
> to get more details.
> 
> Best,
> 
> Sebastien.
> 
> On 11/19/20 2:48 AM, Himanshu Sharma ( via 
> cgal-discuss Mailing List) wrote:
>> Hi Team,
>> I am trying to understand how CGAL goes about doing exact computation. You 
>> have a page https://www.cgal.org/exact.html where you explain it a bit. I 
>> need some clarification.
>> Quoting from the page:
>> In CGAL, we write the high-level algorithms in terms of a well-chosen set 
>> of basic questions (where is a point with respect to a line?) and basic 
>> objects (like a circle through three points). Doing this in the right way 
>> is not always easy, but once it is done, we have outsourced all the 
>> numerical issues, and we only have to make sure that the part of CGAL 
>> concerned with the basics returns correct results. Given this, the 
>> algorithms on top of it just work. Not in most cases, but always!
>> Can you tell me that how do you go about computing answers to set of 
>> 'basic questions' ? Are you using a third party library for it or you 
>> implement that part yourself ? If you do it yourself, can you guide me to 
>> where I can find it in the source code.
>> Regards,
>> Himanshu
>> -- 
>> You are currently subscribed to cgal-discuss.
>> To unsubscribe or access the archives, go to
>> https://sympa.inria.fr/sympa/info/cgal-discuss
> 
> -- 
> You are currently subscribed to cgal-discuss.
> To unsubscribe or access the archives, go to
> https://sympa.inria.fr/sympa/info/cgal-discuss
> 
>