Kepler's Goat Herd
Code for solving Kepler's equation using contour integration, following Philcox et al. (2021, arXiv). This uses a method originally proposed by Ullisch (2020) to solve the "geometric goat problem".
The C++ code contains implementations of a variety of solution methods:
- Newton-Raphson: The quadratic Newton-Raphson root finder.
- Danby: The quartic root finder described in Danby (1988).
- Series: An elliptical series method, as described in Murray & Dermott.
- Contour: A new method based on contour integration.
Given an array of mean anomalies, an eccentricity and a desired precision, the code will estimate the eccentricity using each method. The accuracy of each approach is increased until the desired precision is reached, and timing is performed using the C++ chrono package.
To compile the code using g++, simply run g++ -o kepler keplers_goat_herd.cpp -std=c++17 -ffast-math -Wall -O3. The code can be run using ./kepler. The individual functions, e.g. compute_contour can also be used outside of this script, given an input array of mean anomalies and an eccentricity.
We also provide a pure numpy version of the contour integration function in keplers_goat_herd.py. This is around 9 times slower than the C++ code. Python bindings for the C++ code can be added if these would be of use.
Authors:
- Oliver Philcox (Princeton, [email protected])
37 Oct 31, 2022
390 Dec 8, 2022
28 Dec 7, 2022
1 Nov 1, 2021
6 Oct 14, 2022
2 Sep 3, 2022
15 Sep 3, 2022
2 Mar 22, 2022
3 Apr 25, 2022
34 Aug 2, 2022
10 Dec 30, 2021
4 Oct 24, 2022
1 Dec 11, 2021
7 Dec 31, 2021
19 Jun 19, 2022
136 Dec 27, 2022
14 Sep 11, 2022
403 Jan 9, 2023