An efficient sampler of Polya-Gamma distribution in pure Java.

It is also compatible to other distributions of Apache Commons Math3.

// use default sampler
PolyaGammaDistribution pg1 = new PolyaGammaDistribution(2, 4);
double x = pg1.sample(); 
double mean = pg1.getNumericalMean();
double vari = pg1.getNumericalVariance();

// select sampler explicitly
PolyaGammaDistribution pg2 = new PolyaGammaDistribution(2, 4, PolyaGammaDistribution.AvailableSampler.Gaussian);

• Wrapper (default)
• SumGamma approx
• Gaussian approx
• Saddle Point approx (WIP)

• Polson, Nicholas G., James G. Scott, and Jesse Windle. "Bayesian inference for logistic models using Pólya–Gamma latent variables." Journal of the American statistical Association (2013)
• J. Windle, N. G. Polson, and J. G. Scott. "Improved Polya-gamma sampling". Technical Report, University of Texas at Austin (2013)