wushichao / galacticdnsmass

A demonstration of how bayesian inference and nested sampling allows us to explore the mass distributions of Galactic Double Neutron Star systems.

The mass distribution of Galactic double neutron stars

We highly recommend reading The Mass Distribution of Galactic Double Neutron Stars; (Farrow, Zhu, & Thrane 2019) for details along with this demonstraion.

Here we provide code which performs Bayesian inference on a sample of 17 Galactic double neutron stars (DNS) in order to investigate their mass distribution. Each DNS is comprised of two neutron stars (NS), a recycled NS and a non-recycled (slow) NS. We compare two hypotheses: A - recycled NS and non-recycled NS follow an identical mass distribution, and B - they are drawn from two distinct populations. Within each hypothesis we also explore three possible functional models: gaussian, two-gaussian (mixture model), and uniform mass distributions.

You can take a look at the demo here or you can download the git repository with:

git clone https://github.com/NicholasFarrow/GalacticDNSMass.

Requirements

Without running inference (just demonstration & data analysis):

• Jupyter or Ipython
• numpy, scipy

Additional requirements if performing own inference:

• PyMultiNest (see https://johannesbuchner.github.io/PyMultiNest/install.html)

Full code

A more detailed version of the code can be found here under mainCode.

Citations

Thank you Buchner et al. 2014, A&A for their python interface of MultiNest F. Feroz, M.P. Hobson, M. Bridges. 2008