A python version of qrng library written for R [1]. This code is taken from the qrng GitHub repository with minor translations to remove dependencies on R utils in C.
The Uniform(0,1) iid rng from R utils was replaced with the MRG63k3a generator [2].
The directional numbers are found using methods described in [3].
- Clone this repository:
git clone https://github.com/alegresor/qrngpy.git - Ensure /qrngpy/ is on your python path
- Easiest with virtual environment
- Install requirements:
pip install -r requirements.txt - Compile C files into shared libraries:
make all - Test for successful setup:
python qrng.py
[1] Marius Hofert and Christiane Lemieux (2019). qrng: (Randomized) Quasi-Random Number Generators. R package version 0.0-7. https://CRAN.R-project.org/package=qrng.
[2] Good Parameter Sets for Combined Multiple Recursive Random Number Generators, Shorter version in Operations Research 47, 1 (1999), 159--164. https://doi.org/10.1287/opre.47.1.159.
[3] Faure, Henri, and Christiane Lemieux. “Implementation of Irreducible Sobol’ Sequences in Prime Power Bases.” Mathematics and Computers in Simulation 161 (2019): 13–22. Crossref. Web. https://arxiv.org/abs/1910.04084.