blip-alpha

BLIP: Bayesian LISA Pipeline

This is a bayesian pipeline for detecting stochastic backgrounds with LISA. BLIP stands for Bayesian LIsa Pipeline fully written in python

1. It is easier to maintain and run python code in virtual environments. Make a new virtualenv by doing

python3 -m venv lisaenv

1. Source it on linux or Mac by doing

source lisaenv/bin/activate

For Windows, source it by

activate while in \lisawork\Scripts

1. We need numpy, scipy for running this and matplotlib and chainconsumer are required for plotting. Install them all by doing

pip install numpy scipy matplotlib chainconsumer

1. We also need the healpy, the skymap package

pip install healpy

1. The sampler dynesty is used for nested sampling. We get both the posteriors and bayesian evidence from it. The latter is the detection statistic. Install dynesty by doing

pip install dynesty

1. Some functionality also needs cython

pip install cython

1. You can change the parameters and the signal model in params.ini

To run do python run_blip.py params.ini

Posterior plots are automatically made in the output directory specified in params.ini

1. If you want to generate local documentation pages you also need sphinx

pip install sphinx

Note: The code is setup to work with python 3 and might not work with python2 More documentation at https://blip.readthedocs.io/en/latest/