ASCL.net

Astrophysics Source Code Library

Making codes discoverable since 1999

ASCL Code Record

[ascl:2306.011] margarine: Posterior sampling and marginal Bayesian statistics

Margarine computes marginal bayesian statistics given a set of samples from an MCMC or nested sampling run. Specifically, the code calculates marginal Kullback-Leibler divergences and Bayesian dimensionalities using Masked Autoregressive Flows and Kernel Density Estimators to learn and sample posterior distributions of signal subspaces in high dimensional data models, and determines the properties of cosmological subspaces, such as their log-probability densities and how well constrained they are, independent of nuisance parameters. Margarine thus allows for direct and specific comparison of the constraining ability of different experimental approaches, which can in turn lead to improvements in experimental design.

Code site:
https://github.com/htjb/margarine
Used in:
https://ui.adsabs.harvard.edu/abs/2023MNRAS.520..850A
Described in:
https://ui.adsabs.harvard.edu/abs/2022arXiv220512841B
Bibcode:
2023ascl.soft06011B
Preferred citation method:

Please see citation information here: https://github.com/htjb/margarine#licence-and-citation


Views: 723

ascl:2306.011
Add this shield to your page
Copy the above HTML to add this shield to your code's website.