Code for numerical computation in research on orbital dynamics in the Kepler-Heisenberg problem (solving for sun/planet dynamics in the Heisenberg group) -- a collaboration with Corey Shanbrom.
The publication "Numerical Methods and Closed Orbits in the Kepler-Heisenberg Problem" in which these results appear can be found in the Journal of Experimental Mathematics, and the preprint on the arXiv.
The publication "Self-Similarity in the Kepler-Heisenberg Problem" in which these results appear can be found in Journal of Nonlinear Science (publication forthcoming as of 2021.03.22), and the preprint on the arXiv.
- Victor Dods : programming, some math
- Corey Shanbrom : math, some programming
The principal aim in publishing this code repository in tandem with the peer-reviewed publication is to provide the ability to freely, easily, and exactly reproduce the the results therein. Ideally, this will contribute to a higher standard of what is expected from a publication having a computer-based element -- it should be as easy as is reasonably possible for an interested reader to reproduce the results.
Full details on how to reproduce the results found in "Numerical Methods and Closed Orbits in the Kepler-Heisenberg Problem" can be found here.
Full details on how to reproduce the results found in "Self-Similarity in the Kepler-Heisenberg Problem" can be found here.
This project is held under Copyright (2014-2018) by Victor Dods and is released as free, open-source software under the MIT License, without warranty.
If you use any code from this project, I'd love to hear about what you're applying it to! Send me an email about it at
victor <dot> dods <at-sign> gmail <dot> com
It is requested, though not required, that any use of the code from this project in a derivative work be cited/acknowledged in the relevant documentation for the derivative work. The citation should include:
- Author:
Victor Dods - The name of this project:
Kepler-Heisenberg Problem Computational Tool Suite - The public link to this github repository:
https://github.com/vdods/heisenberg
- heisenberg : Top-level program (and module) to explain and redirect to all the specific subprograms of this project.
- heisenberg.library : A module that contains mostly math-oriented code that is used in multiple subprograms.
- heisenberg.plot : Plots an integral curve of the system using given initial condition, optionally running an optimization method to find a nearby curve that closes back up on itself.
- heisenberg.plot_samples : Provides visualization of the data generated by the heisenberg.sample subprogram. In particular, this gives a colormapped scatterplot of the objective function on the fully reduced, 2-parameter initial condition space.
- heisenberg.sample : Samples a specified parameter space of initial conditions, computing the corresponding integral curves, and computing and storing relevant data about the each curve in a file. This is intended to sample various functions of the initial condition space, and can be used in later processing; see the heisenberg.plot_samples subprogram.
- heisenberg.search : Search a specified parameter space for initial conditions for curves. Once a curve is found that comes close enough to closing back up on itself, an optimization method is used to attempt to alter the initial conditions so that the curve closes back up on itself. The result is integral curves which approximate closed solutions to within numerical accuracy.
Additional content that is no longer an active part of the project:
- attic : Directory containing deprecated code.
The following command, when executed from this project's root directory, will print a directory of the available subprograms.
python3 -m heisenberg
Examples of invoking specific subprograms:
python3 -m heisenberg.plot --dt=0.003 --max-time=25.03042826445711 --initial-preimage=[0.2706994702908095] --embedding-dimension=1 --embedding-solution-sheet-index=0 --output-dir=generated-data --disable-plot-decoration --cut-off-initial-curve-tail --quantities-to-plot=x,y --plot-type=pdf
python3 -m heisenberg.search --dt=1.0e-2 --max-time=200 --seed=123456789 --embedding-dimension=2 --embedding-solution-sheet-index=1 --plot-type=pdf --output-dir=generated-data --exit-after-number-of-successes=1 --quantities-to-plot="x,y;t,z;sqd;objective" --use-terse-plot-titles --plot-size=3
Please see this and the scripts that it refers to for more specific examples.
Please see this and the scripts that it refers to for the specific examples for that paper.
There are various Python modules needed by this tool suite, including vorpy which has the
module for symplectic integration. Version 0.4.3 was the specific version used to produce the results in the "Numerical Methods..."
paper, and version 0.9.0 was the specific version used to produce the results in the "Self-Similarity..." paper. Python modules
can be installed easily using a command having the form
pip3 install <package-name>
In particular, to install vorpy, use
pip3 install vorpy
The pip3 command may either need to be run as the super-user (e.g. sudo pip3 install <package-name> which will install
the package globally for use by everyone on the system) or in the user scheme (e.g. pip3 install --user <package-name> which
will install the package under the user's home directory and therefore be available only to that user).
It was discovered that there was a request to be cited by the dill package which is used
indirectly by this codebase after the "Numerical Methods..." paper was published. The citation is as follows.
If you use dill to do research that leads to publication, we ask that you acknowledge use of dill by citing the following in your publication:
M.M. McKerns, L. Strand, T. Sullivan, A. Fang, M.A.G. Aivazis,
"Building a framework for predictive science", Proceedings of
the 10th Python in Science Conference, 2011;
http://arxiv.org/pdf/1202.1056
Michael McKerns and Michael Aivazis,
"pathos: a framework for heterogeneous computing", 2010- ;
http://trac.mystic.cacr.caltech.edu/project/pathos
Please see http://trac.mystic.cacr.caltech.edu/project/pathos or http://arxiv.org/pdf/1202.1056 for further information.
A software project is never truly finished. See the list of to-dos here.