Bayesian excess variance for Poisson data time series with backgrounds. Excess variance is over-dispersion beyond the observational poisson noise, caused by an astrophysical source.
- Introduction
- Method
- Tutorial
- Output plot and files
In high-energy astrophysics, the analysis of photon count time series is common. Examples include the detection of gamma-ray bursts, periodicity searches in pulsars, or the characterisation of damped random walk-like accretion in the X-ray emission of active galactic nuclei.
paper: https://arxiv.org/abs/2106.14529
This repository provides new statistical analysis methods for light curves. They can deal with
- very low count statistics (0 or a few counts per time bin)
- (potentially variable) instrument sensitivity
- (potentially variable) backgrounds, measured simultaneously in an 'off' region.
The tools can read eROSITA light curves. Contributions that can read other file formats are welcome.
The bexvar_ero.py tool computes posterior distributions on the Bayesian excess variance, and source count rate.
quick_ero.py computes simpler statistics, including Bayesian blocks, fraction variance, the normalised excess variance, and the amplitude maximum deviation statistics.
AGPLv3 (see COPYING file). Contact me if you need a different licence.
Install as usual:
$ pip3 install bexvar
This also installs the required ultranest python package.
Run with:
$ bexvar_ero.py 020_LightCurve_00001.fits
Run simpler variability analyses with:
$ quick_ero.py 020_LightCurve_*.fits.gz
Contributions are welcome. Please open pull requests with code contributions, or issues for bugs and questions.
Contributors include:
- Johannes Buchner
- David Bogensberger
If you use this software, please cite this paper: https://arxiv.org/abs/2106.14529
- https://github.com/rarcodia/eRebin for rebinning eROSITA light curves to eroDays