Conference site ยป Proceedings

Automatic random variate generation in Python

Christoph Baumgarten
Unaffiliated

Tirth Patel

Abstract

The generation of random variates is an important tool that is required in many applications. Various software programs or packages contain generators for standard distributions like the normal, exponential or Gamma, e.g., the programming language R and the packages SciPy and NumPy in Python. However, it is not uncommon that sampling from new/non-standard distributions is required. Instead of deriving specific generators in such situations, so-called automatic or black-box methods have been developed. These allow the user to generate random variates from fairly large classes of distributions by only specifying some properties of the distributions (e.g. the density and/or cumulative distribution function). In this note, we describe the implementation of such methods from the C library UNU.RAN in the Python package SciPy and provide a brief overview of the functionality.

Keywords

numerical inversion, generation of random variates

DOI

10.25080/majora-212e5952-007

Bibtex entry

Full text PDF