Folks,
It’s a great pleasure to release the codes from my book on GitHub. Over the years, I’ve received many requests for this, and now seems like the right time to do it! A few important notes:
This code is released under the GNU Affero General Public License v3.0. You are free to use, modify, and distribute it under the terms of this license, provided that any modifications or derivative works are also made available under the same license.
We don’t have all the routines published yet, as we’ve been testing them in each language first. This is a big task, but very fruitful as different languages have peculiarities that are beneficial in finding small errors. The spreadsheet lists:
- Routines that are essentially complete
- Routines in progress
- Routines that still need some work
A few notable ones:
- Maneuver routines – These are simple magnitude-only routines that are often better served by Lambert’s solution.
- Orbit determination (simple Batch Least Squares) – Originally developed for SGP4DC - this needs some work to function correctly in C#. We have a beta version, but it’s not yet producing the correct outputs.
- Accelerations in coordinate transformations – These are partially implemented. The C# version includes them in the overall ECI-to-ECEF transformations, and we plan to include them in all languages shortly.
- STK scenarios – Many are from older versions but should still run in the latest versions of STK. We will eventually add Word documents to describe what is actually going on!
- SGP4 – 29 August 2024 Update -- Includes check for XP data. XP data cannot be propagated through SGP4, it requires the SGP4-XP propagator.
We are developing test programs but want to modify them to use the assert functionality available in each language, and this is still in progress. Our intent is to finalize these tests for the first set of functions before making them publicly available.
Some routines have known discrepancies between languages, which we are working on resolving. The C# version is generally the most authoritative, followed closely by MATLAB and Python.
Some key differences:
- Covariance conversions – Not all conversions work forward and backward as indicated in our paper (AAS 15-537). Additionally, the retrograde factor for equinoctial elements has been changed from applying to all retrograde orbits to only those near 180-degree inclinations. This was done to minimize differences near polar orbits, but of course, it changes a lot of the results (AAS 15-537 Rev 3)
- Angles-only routines – These are complex, and while some results match between languages, others do not. We are still working on these!
- Other discrepancies – If you notice any issues, please let us know and open a GitHub issue! We are working on integrating these checks into our
assert-based test framework.
We have included C++, but a full update is planned in the coming months. We have heard concerns about C++ memory leaks making it vulnerable to exploitation, so its' long-term role in this project may have some uncertainties.
We appreciate interest in contributing to this project! However, as we are still figuring this process out, we ask contributors to submit an issue as their first step. For more details, see the Contributing Guide.
As always, let us know what you think, what you’d like to add, and any feedback you have. Looking forward to this new adventure with you all! 🚀
Dave (and Samira)