Skip to content

Latest commit

 

History

History
40 lines (26 loc) · 1.28 KB

README.md

File metadata and controls

40 lines (26 loc) · 1.28 KB
Raw

README

This version of EightBitTransit is an MIT-licensed python code redesigned to make use of GPU parallelization with Numba:

  1. Can calculate the model of any pixelated image transiting a star
  2. Can inject the model into a light curve with a given depth (negative allowed, i.e. a spike) and duration
  3. Can invert a light curve to recover the "shadow image" that produced it. (from original code unmodified).

Installation

To install EightBitTransit, run:

pip install eightbittransit

Dependencies

  • Numpy
  • Scipy
  • Numba
  • imageio
  • itertools
  • matplotlib

Optionally (for the example notebook):

  • jupyterlab

Examples

See ./examples/examples-gpu.ipynb for examples of both the forward and inverse shadow imaging problem, including for dip 5 of Boyajian's Star. This code reproduces figure 12 of Sandford & Kipping 2018 (https://arxiv.org/abs/1812.01618).

See /examples /eightbit-injection.ipynb for examples on how to inject a signal into an existing lighr curve.

Note on memory

Repeated TransitingImage() calls can cause python to run out of memory in certain cases (thanks to textadactyl for pointing this out!)--when you are done with a TransitingImage object, deallocate it explicitly, i.e.:

ti = TransitingImage(...)

some operations on ti, etc.

ti = None