Skip to content

petehague/BASC

Repository files navigation

README

BaSC is short for Bayesian Source Characterisation. It is an MCMC process that performs source detection and characterisation on dirty maps, taking into account the properties of the beam in a more rigorous way that CLEAN does.

It is based on a method developed by Steve Gull, and uses the BayeSys MCMC driver.

Installation

BaSC requires a c compiler, Python 3, and the numpy, astropy and sklearn libraries. On Linux, type make to compile the extension, or if using Mac, type 'make mac'. Currently Windows is not supported - you will have to run it in an Ubuntu terminal on a Windows machine.

Usage

To try out BaSC, run the example script example.py in the BaSC folder. This should locate a single source in the centre of the map, and return the models from the burned in chain to chain.txt

To run BaSC on your own data, use

basc.py <dirty map file> <dirty psf file> <primary beam flux file>

In general, import basc into your Python program and use it as shown in example.py

Contact

For more information, please email Peter Hague at peterhague AT protonmail.com

About

A Bayesian source finding algorithm for interferometers

Resources

License

Stars

Watchers

Forks

Packages

No packages published