Fix an integer and parameters , commonly written as a vector . has the Dirichlet distribution (we write ) if it has support on the probability simplex and its probability density function (PDF) satisfies
The distribution is special because its PDF is constant over its support:
The following lemma demonstrates one way to generate random draws from the distribution:
Lemma. Let , and let denote the order statistics for . Then
(See Section 2 of Reference 1 for various ways to generate samples from the Dirichlet distribution for arbitrary .)
The proof of the lemma starts with the following theorem, which gives the joint density of all the order statistics:
Theorem. Let be i.i.d. random variables with PDF . Then the joint density of the order statistics is
(This is Theorem 6.1 of Reference 2; you can find a proof for this theorem there.) Applying this theorem to , we have
The lemma then follows by applying a change of variables to the PDF (see this link for the steps one needs to carry out for the change of variables).
- Frigyik, B. A., et. al. (2010). Introduction to the Dirichlet Distribution and Related Processes.
- DasGupta, A. Finite Sample Theory of Order Statistics and Extremes.