backtracks is a python package to fit relative astrometry with background helical motion tracks, to discern directly imaged planets 🪐 from contaminant sources 💫 ⭐

The code is written and developed by Gilles Otten (@gotten), William Balmer (@wbalmer), and Tomas Stolker (@tomasstolker).

Documentation can be found at http://backtracks.readthedocs.io.

A Jupyter notebook will show you how to use backtracks by reproducing the result in Nielsen et al. (2017) and Wagner et al. (2022) for the case of the former exoplanet candidate around HD 131339 A.

If you use backtracks in your published work, please cite our Zenodo entry (button above, and here), and provide a footnote/acknowledgement linking to our package. Thank you!

• eDR3 Distance prior summary file from Bailer-Jones et al. (2021).
• Example of HD 131399Ab uses data from Wagner et al. (2022) and Nielsen et al. (2017). Thank you to Kevin Wagner for providing the latest astrometry!
• Log-likelihood and some utility functions borrowed heavily from orbitize! (BSD 3-clause).
• PPF of multivariate normal borrowed from pints (BSD 3-clause).

Currently requires and python 3.10-3.13 and astropy, corner, dynesty, matplotlib, numpy, novas, novas_de405, and their dependencies. Note that novas is not supported on Windows. You can create a working environment using conda+pip via a few lines of code:

$ conda create python=3.11 -n backtracks
$ conda activate backtracks
$ conda install pip
$ pip install backtracks

Or, to clone the repo and install in development mode (we recommend this, as the code is a work in progress and you can easily fix bugs you will likely encounter this way):

$ conda create python=3.11 -n backtracks
$ conda activate backtracks
$ conda install pip
$ git clone https://github.com/wbalmer/backtracks.git
$ cd backtracks
$ pip install -e .

Then, test your installation:

>>> from backtracks import System