Skip to Content.
Sympa Menu

cgal-discuss - [cgal-discuss] Performance of Triangulation_3

Subject: CGAL users discussion list

List archive

[cgal-discuss] Performance of Triangulation_3


Chronological Thread 
  • From: mytien <>
  • To:
  • Subject: [cgal-discuss] Performance of Triangulation_3
  • Date: Tue, 15 Oct 2013 09:52:36 -0700 (PDT)

Hi,
In the chapter 39.6.1 Running Time of the 3D triangulation manual:
http://www.cgal.org/Manual/latest/doc_html/cgal_manual/Triangulation_3/Chapter_main.html
I have seen that the triangulation of 1000 points takes only 0.00724
seconds.
But when I have a vector v with 1000 random points and insert them into the
triangulation t via t.insert(v.begin(), v.end()), it takes approximately
0.250 to 0.44 seconds.

I defined my triangulation types like this:
typedef CGAL::Exact_predicates_inexact_constructions_kernel K;
typedef CGAL::Triangulation_3<K> Triangulation;
typedef Triangulation::Point CGAL_Point;

Then I created random x,y,z float-values and added them to my vector like
this:
std::vectory<CGAL_Point> points;
CGAL_Point p = CGAL_Point(x, y, z);
points.push_back(p)

I start the timer before this statement and measure it directly afterwards:
t.insert(v.begin(), v.end());

In my CGAL CMakeCache.txt I found this configuration:
CMAKE_CXX_FLAGS_RELEASE:STRING=-O3 -DNDEBUG

When compiling my test application I use these flags:
-O3 -DCGAL_NDEBUG

I also added these defines at the very first two lines of my application:
#define _SECURE_SCL 0
#define DCGAL_NDEBUG

My computer has an i7 processor with 8GB RAM. Release mode and debug mode
yield almost identical results.

Do I need to make any additional adjustments?





--
View this message in context:
http://cgal-discuss.949826.n4.nabble.com/Performance-of-Triangulation-3-tp4658201.html
Sent from the cgal-discuss mailing list archive at Nabble.com.



Archive powered by MHonArc 2.6.18.

Top of Page