[ascl:1806.030]
foxi: Forecast Observations and their eXpected Information

Using information theory and Bayesian inference, the foxi Python package computes a suite of expected utilities given futuristic observations in a flexible and user-friendly way. foxi requires a set of n-dim prior samples for each model and one set of n-dim samples from the current data, and can calculate the expected ln-Bayes factor between models, decisiveness between models and its maximum-likelihood averaged equivalent, the decisivity, and the expected Kullback-Leibler divergence (i.e., the expected information gain of the futuristic dataset). The package offers flexible inputs and is designed for all-in-one script calculation or an initial cluster run then local machine post-processing, which should make large jobs quite manageable subject to resources and includes features such as LaTeX tables and plot-making for post-data analysis visuals and convenience of presentation.

- Code site:
- https://github.com/umbralcalc/foxi
- Described in:
- http://adsabs.harvard.edu/abs/2018JCAP...05..070H

- Bibcode:
- 2018ascl.soft06030H

Preferred citation method:

Views: 1701