Skip to content

GRB triangulation via non-stationary time-series models

License

Notifications You must be signed in to change notification settings

grburgess/nazgul

Repository files navigation

Travis CI w/ Logo codecov PyPi Downloads PyPI version fury.io Documentation Status GitHub contributors astropy

nazgul

alt text

What is this?

Nazgul is a framework for performing GRB localization via fitting non-parametric models to their data time-series and computing the the time delay between them. It is currentrly built upon the magic of Stan and implements a parallel version of non-stationary Random Fourier Features. The idea is get away from heuristic methods such as cross-correlation which do not have a self-consistent statitical model.

The idea is that satellites throughout the Sol system observe gamma-ray bursts at different times due to the finite speed of light. This creates a time delay in their observed light curves which can be used to triangulate the gamma-ray burst position on the sky. These triangulation create annuli or rings on the sky which Nazgul searchs for so that it, in the darkness, it can bind them to a location on the sky. (We are nerds, get over it).

alt text

The heriarchical model is shown below and details can be found in here. If you find the method and/or code useful in your research we ask that you please cite the paper.

alt text

The sister program to simulate time-delayed light curves is pyIPN and can be used to generate time-delayed light curves for algorithm testing.


This work is a joint effort by:

  • J. Michael Burgess
  • Ewan Cameron
  • Dmitry Svinkin

Installation

pip install nazgul

Note!

Nazgul is still under heavy development! Expect problems!