Parallel Delaunay and Voronoi Tessellation and Density Estimation
C++ C CMake Python
Switch branches/tags
Nothing to show
Clone or download

Parallel Delaunay and Voronoi Tessellation and Density Estimation


Tess is released as open source software under a BSD style license.


  • a C++11 compiler
  • the DIY block parallel library
  • either the Qhull or CGAL computational geometry library


The following instructions assume Qhull is the serial geometry engine.

Build Dependencies

a. DIY

git clone

b. Qhull

tar -xvf qhull-2012.1-src.tgz
cd qhull-2012.1-src

Build Tess

git clone

cmake /path/to/tess \
-DCMAKE_INSTALL_PREFIX=/path/to/tess2/install \
-Dserial=QHull \
-DDIY_INCLUDE_DIRS=/path/to/diy/include \
-DQHull_INCLUDE_DIRS=/path/to/qhull/include \

make install

Optionally, Tess can use CGAL instead of Qhull. To do so, pass -Dserial=CGAL as the choice to cmake.


  1. Test tessellation only
cd path/to/tess2/install/examples/tess

Edit TESS_TEST: select ARCH, num_procs, dsize (number of particles)

path/to/tess2/install/tools/draw del.out

Mouse interaction with drawing: mouse move to rotate, ‘z’ + mouse up, down to zoom, ‘t’ to toggle voronoi tessellation, ‘y’ to toggle delaunay tessellation, ‘f’ to toggle shaded rendering

  1. Test tessellation + density estimator

(from tess top level directory)

cd path/to/tess2/install/examples/tess-dense

Edit TESS_DENSE_TEST; select ARCH, num_procs, dsize (number of particles), gsize (number of grid points)

path/to/tess2/install/tools/ --raw=dense.raw --numpts=512

(assuming outfile was dense.raw and gsize was 512 512 512 in TESS_DENSE_TEST) is a python script using numpy and matplotlib, but you can use your favorite visualization/plotting tool (VisIt, ParaView, R, Octave, Matlab, etc.) to plot the output. It is just an array of 32-bit floating-point density values listed in C-order (x changes fastest).