Searcharxiv⌕ Search
Home/Archive/arXiv

arXiv · 1810.04744

Generalized Ziggurat Algorithm for Unimodal and Unbounded Probability Density Functions with Zest

Morteza Jalalvand·Mohammad A. Charsooghi
View original source ↗Read source document ↗Export metadata ↓

Abstract

We present a modified Ziggurat algorithm that could generate a random number from all unimodal and unbounded PDFs. For PDFs that have unbounded density and/or unbounded support we use a combination of nonlinear mapping function and rejection sampling to generate a random number from the peak and/or the tail distribution. A family of mapping functions and their corresponding acceptance probability functions are presented (along with the criteria for their use and their efficiency) that could be used to generate random numbers from infinite tails and unbounded densities. The Zest library which is a C++ implementation of this algorithm is also presented. Zest can efficiently generate normal, exponential, cauchy, gamma, Weibull, log-normal, chi-squared, student's t and Fisher's f variates. The user can also define their custom PDF as a class and supply it as a template argument to our library's class without modifying any part of the library. Performance of Zest is compared against performance of random modules of (GCC's implementation of) Standard Template Library (STL) and Boost. The presented results show that Zest is faster than both in most cases, sometimes by a factor of more than 10. We also present a C++ implementation of a uniform floating-point random number generator (RNG) which is capable of producing all representable floating-point numbers in $[0,1)$ which will be used in the Ziggurat algorithm near unbounded peaks and tails. The common method of dividing a random integer by the range of the RNG can not produce random floating-point numbers with fully random fraction bits and very small random numbers. The presented uniform floating-point RNG is very efficient and in the case of producing double precision floating-point numbers it's even faster than simply multiplying a 64-bit integer by $2^{-64}$.

Explore related subjects

stat.COphysics.comp-ph

Keep this discovery

BibTeXRIS

Morteza Jalalvand, Mohammad A. Charsooghi. 2018-10-10. Generalized Ziggurat Algorithm for Unimodal and Unbounded Probability Density Functions with Zest. https://arxiv.org/abs/1810.04744

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Estimating Hierarchically Rank Structured Covariance Matrices

We consider the problem of estimating a high-dimensional covariance matrix from a very limited number of samples. This problem is ubiquitous in computational fluid dynamics, where a small number of fluid snapshots must be used to construct a Gramian matrix determining a reduced-order model, as well as in computational geoscience, where a small ensemble of Earth system forecasts must be used to estimate the covariance matrix associated with the forecast uncertainty. It is common practice to regularize the small-sample covariance by imposing a "localization" structure that enforces a physically realistic correlation length scale, imposing a sparsity constraint, "shrinking" towards a prescribed target, or attenuating small correlations. We propose an alternate technique that regularizes the small-sample covariance by imposing hierarchical rank structure. Compared to regularization methods that assume sparsity such as spatial localization, hierarchical rank structure accommodates a wider range of covariance matrices, roughly corresponding to situations where long-range correlations vary more smoothly than short-range ones. It also results in a data-sparse matrix format that permits highly efficient matrix-vector products. We present theory and algorithms which show how to efficiently estimate a high-dimensional, hierarchically rank structured covariance matrix from limited samples. Through an error analysis and numerical experiments with a variety of model problems, we demonstrate that these techniques are effective at reducing sampling errors, and that in many cases they achieve smaller estimation error than conventional techniques.

stat.CO↗
arXiv2026

Optimal Slice-Adaptive Tuning of Hybrid Slice Sampling

Slice sampling is a Markov chain Monte Carlo algorithm that draws its next state uniformly from a "slice"---a super-level set of the target density function---at each iteration, thereby providing automatic local adaptivity to the scale of the target. In practice the exact slice is not known, so general-purpose implementations use an approximate slice that is grown from a starting interval of length $w>0$, with a computational cost that depends on $w$. This work presents an analysis of the average per-iteration number of target density evaluations, as a function of $w$, of hybrid slice sampling with various slice-finding schemes for targets with contiguous slices. The paper uses the results of the analysis to develop automated, slice-adaptive tuning schemes along with suboptimality bounds and asymptotic convergence guarantees. Simulations demonstrate that the tuning schemes reliably yield near-optimal slice-adaptive tuning with essentially no dependence on the initial setting of $w$.

stat.CO↗