The Art of Randomness

12SAMPLING

We’ve powered the majority of our experiments by extracting samples from the uniform distribution. While we’ve also worked with the normal distribution (Chapter 1), beta distribution (Chapter 3), and binomial distribution (Chapter 9), the tried-and-true uniform distribution is our oldest and dearest friend.

In this chapter, we’ll sample from arbitrary probability distributions, be they discrete or continuous. This ability is critical to simulation and fundamental to Bayesian inference.

First, we’ll discuss terminology and unpack the term Bayesian inference. Following that, we’ll dive into sampling from arbitrary discrete probability ...