Adaptive rejection sampling in F#

Adaptive rejection sampling (derivative-free). Based on matlab implementation from PMTK3 library.

Compared to the original implementation in matlab, this version computes probabilities in log space which avoids numerical problems.

• func - Function which computes logarithm of a likelihood function
• a, b - Starting points, a < b.
• domain - Tuple of func boundaries. If the left boundary is -infinity, derivative of f at a must be positive. If the right boundary is infinity, derivative of f at b must be negative.
• nSamples - Total number of samples that the sampler should return.

• Gilks W. 1992. Derivative-free adaptive rejection sampling for Gibbs sampling. In: Bernardo J.M., Berger J.O., Dawid A.P., and Smith A.F.M. (Eds.), Bayesian Statistics 4. Oxford University Press, xford, pp. 641–665.
• Gilks W.R. and Wild P. 1992. Adaptive rejection sampling for Gibbs sampling. Applied Statistics 41: 337–348.