Bayesian nonparametric (BNP) models are prior models for infinite-dimensional parameters, such as an unknown probability measure 
or an unknown regression mean function 
We review some of the most widely used BNP priors, including the Dirichlet process (DP), DP mixture, the Polya tree (PT), and Gaussian process (GP) priors. We discuss how these models are used in typical inference problems. The examples include R code using available packages for inference under BNP priors.