Construct a Dirichlet Process from any random sampler.
In [1]: from dp import DirichletProcess
In [2]: from random import normalvariate
In [3]: from collections import Counter
In [4]: dp = DirichletProcess(lambda: normalvariate(0, 1), alpha=1)
In [7]: Counter(dp() for _ in range(1000))
Out[7]: Counter({-0.029073686274754998: 679, 0.7332646430104509: 301, 0.23296111210571477: 12, -0.7042044460111074: 4, -2.2504937687943922: 4})
Install with pip install dirichletprocess or conda install -c tdhopper dirichletprocess.