Probabilistic programming from scratch

Let's use it to finish our A/B test, starting with layout A. We need to prepare three arguments: data, prior_sampler and simulate. We already have the data for our A/B test. Let's now write a function that simulates the conversion of n_visitors visitors to a website with known probability p: Here's what happens when we run this function a few times to simulate 100 visitors converting with probability 0.1: Effectively, this function runs a fake A/B test in which we already know the conversion fraction.