Copyright (c) 2016, Lawrence Livermore National Security, LLC.
Produced at the Lawrence Livermore National Laboratory.
Written by Harsh Bhatia (bhatia4@llnl.gov).
CODE-701040.
All rights reserved.
Robust Critical Point Detection is a command line utility to compute critical points in a 2D or 3D vector field robustly. The technique was published in 2014 [1], and uses a numerically robust approach for detecting whether a singularity exists in a simplex. The code assumes piecewise-linear interpolation in a simplex. Built upon Simulation of Simplicity [2], the code naturally resolves degenerate cases.
[1] H. Bhatia, A. Gyulassy, H. Wang, P.-T. Bremer, and V. Pascucci. Robust Detection of Singularities in Vector Fields. In Topological Methods in Data Analysis and Visualization III - Theory, Algorithms, and Applications, Mathematics and Visualization, pp. 3–18, P-T Bremer, I Hotz, V Pascucci, R Peikert (Eds.), Springer Berlin Heidelberg, 2014.
[2] H. Edelsbrunner, E.P. Mücke, Simulation of simplicity: a technique to cope with degenerate cases in geometric algorithms. ACM Trans. Graph. 9, (1990), pp. 66–104.
The software depends upon Simulation of Simplicity. The code for SoS was provided by Ernst Mücke (http://www.geom.uiuc.edu/software/cglist/GeomDir/) as a sub-module for a different software. The original SoS code was written in C in the 1990's, and requires some modifications for compilation with modern C++ compilers. A patch has been provided to fix the code and build it (patch_SOS.txt)
The SoS and the patch can be automatically installed using the install_deps.sh script.
The CP detection tool can be installed by first building SimulationOfSimplicity as a static library and then using the cmake system.
$ pwd
RobustCriticalPointDetection
$ sh install_deps.sh
$ mkdir build && cd build
$ cmake ../
$ make
The tool can be run in one of the following four ways (2,3,4, or 5 arguments).
$ ./CriticalPointDetection file1.vti [[NOTE: this mode works for both 2D and 3D]]
where,
file.vti is a XML-format VTK Image data file
or
$ ./CriticalPointDetection file1 file2 [[NOTE: this mode currently works for 3D only]]
where,
file1 is a text file where each line is: x y z vx vy vz (coordinates of points and corresponding vectors)
file2 is a text file where each line is: i1 i2 i3 i4 (indices of the 4 corners of a tet)
or
$ ./CriticalPointDetection file1 X Y [[NOTE: this mode is for 2D fields]]
where,
file1 is a text file where each line is: x y vx vy (coordinates of points and corresponding vectors)
X Y are the dimensions of the regular grid (program creates triangles automatically)
or
$ ./CriticalPointDetection file1 X Y Z [[NOTE: this mode is for 3D fields]]
where,
file1 is a text file where each line is: x y z vx vy vz (coordinates of points and corresponding vectors)
X Y Z are the dimensions of the regular grid (program creates tets automatically)
The program writes the critical points as a space-delimeted text file. The output filename is <file1>.cp.txt. Each line of the output file contains 4 numbers:
simplex_id x y z
where, simplex_id is the id of the simplex containing the critical point, and x y z represent the centroid of the simplex.
The code is easy to extend for a variety of data formats. The code only needs the vector field and tetrahedra. Please see the main function to write custom input/output formats.