Thomas Schweich, GEE Lab at Washington University in St. Louis. MIT License.
The repository serves two purposes, as:
- A clean and efficient Python interface to files created by Dan Morris' Voxelizer.
- A massively parallel gravitational potential calculator for arbitrary 3d point matrices (which can be easily modified to serve as an electric potential calculator) using OpenCL and PyOpenCL.
Source code relating to the former is stored in the voxel directory, while source relating to the latter is stored in the main directory.
As a Python Interface to Dan Morris' Voxelizer
Running the script from the command line provides options for loading a single .voxels file or multiple files, indicated
by a file glob (using * as a wildcard), into a single Voxels Python object.
The script retains all information in the file(s) except the file header(s).
The script can be run interactively, in which case the Voxels object resulting from the loaded file(s) can be accessed
through a variable called 'voxels', such as in the following example:
$ python -i load_voxels.py myvoxels_*.voxels
>>> voxels.to_set()
{(0, 0, 0), (0, 1, 0) ... }
>>> first_voxel = voxels.get_all_voxels()[0]
>>> first_voxel
<__main__.VoxelfileVoxel object at 0x0000000006A3FEC8>
>>> first_voxel.i
0
>>> first_voxel.has_normal
1
>>> voxels.plot_graph()
# Graph appears
# etc...
where 'myvoxels_1.voxels', 'myvoxels_2.voxels'... are the names of your .voxels files.
The set resulting from the call to voxels.to_set() would contain the positions of each voxel from the file.
The VoxelfileVoxel object resulting from voxels.voxels[0] would be the first voxel in the file.
ipython could also be used in place of python. The dollar-sign ($) is simply meant to indicate that the proceeding
line is entered into a console (whose working directory is the one containing the script).
Use $ python load_voxels.py -h for help. The module may also, of course, simply be imported by other scripts.
The Voxels object also contains methods for dealing with the concatenation of multiple .voxels files into a single
object. This is due to the fact that the script was originally written for use with the voxelization of a complex CAD
file which was divided into many small components. As a result, the script would theoretically work with multiple
objects as well.
The script has been tested to run relatively fluently with multi-million-voxel files, including plotting. The test was run on a laptop with a 2 core, 4 thread Intel i5 processor with 8GB RAM. Performance likely scales quite well with improved hardware.
The module can also be imported and used in other programs.
Scripts can be found in the voxels directory which serve the following purposes:
- Breaking apart a blender object into its component parts
- Combining a group of voxel files within a directory
- Repeatedly invoking the Voxelizer program on a directory of objects, and
- Detecting suspicious voxelizations
These are more one-off style scripts that will need to be modified to suit your needs.
The important function contained in this file is calculate_potentials_opencl, along with its helper function generate_opencl_objects.
The generate_opencl_objects takes the following parameters:
- A 2d array of 3-vectors representing each point at which you would like the gravitational potential
- A 2d array of 3-vectors representing each point of mass you would like to be accounted for in the calculations
The calculate_potentials_opencl function takes the following parameters:
- The result of your call to
generate_opencl_objects - The mass density of the volume matrix material
- The mass density of the mass matrix material
A function plot_results uses PyQTGraph to plot two resulting matrices, with the resulting potentials used as a color scale from
blue (low) to red (high). The module also contains functions for generating example spaces. By default, running the file will generate
an expansive universe of point masses with a cube matrix in the middle for which potential values are calculated. With reasonably new
hardware, despite being an O(n^2) operation, many points can be calculated very quickly.
This file simply proves the accuracy of the calculation when compared to a similar calculation made using NumPy as well as rote Python. Equality assertions are made, so if the script runs successfully your GPU (or other OpenCL device) has arrived at the same conclusion as your CPU.
The file contains the actual kernel used to calculate the gravitational potentials with OpenCL. The value of G could be changed to Coulomb's Constant and electric charge could be used in place of mass to calculate electric potentials.
An arbitrary experimental apparatus and volume may be input into the voxelizing module, with the output of a call to to_ndarray()
fed into the potential calculation module in order to obtain the potential analysis of a scientific experiment. This was the original
intention of the project.
Copyright 2017 Thomas Schweich and the GEE Lab at Washington University in St. Louis
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.