Bayes factor

Compute the Bayes factor using the Savage-Dickey ratio.

We can apply this function when the null model is nested within the alternative. In other words, when the null (ref_val) is a particular value of the model we are building (see here).

For other cases, computing the Bayes factor is not straightforward and requires more complex methods. Instead of Bayes factors, we usually recommend Pareto-smoothed importance sampling leave-one-out cross-validation (PSIS-LOO-CV). In ArviZ, you will find them as functions with loo in their names.

using ArviZPythonPlots, ArviZExampleData

use_style("arviz-variat")

data = load_example_data("centered_eight")
pc = plot_bf(data, ["mu"])

See plot_bf.