A flexible distribution family for testing MCMC implementations

We propose a flexible, extensible family of distributions for testing Markov Chain Monte Carlo implementations. Distributions are created by nesting simple transformations, which allow various shapes, including multiple modes and fat tails. The resulting distributions can be sampled with high
precision using quasi-random sequences, and have closed form (log) density and gradient at each point, making it possible to test gradient-based samplers without automatic differentiation.