Add Bayesian bootstrap

[timmens]

What would you like to enhance and why? Is it related to an issue/problem?

Currently, estimagic only supports the standard bootstrap. The Bayesian bootstrap (Rubin, 1980) provides an alternative with good theoretical properties in both a Bayesian and frequentist world. A good summary is given in this Twitter thread.

Describe the solution you'd like

Translate this R script to Python and integrate the functionality into the bootstrap function. Possibly use a bayesian=True flag or a type="bayesian" argument.