Simulation of random vectors from three-dimensional spherically symmetric stable distributions (Q1291199)

An approximate algorithm for generating random vectors \(\mathbb{R}\) from a three-dimensional spherically symmetric stable distribution based on the inverse transform method is indicated. Essentially a relation between a three-dimensional spherically symmetric stable probability density function and an one-dimensional symmetric probability density function, due to \textit{V. M. Zolotarev} [``One-dimensional stable distributions'' (1986; Zbl 0589.60015)], is used for approximating the distribution function of \(R\), \(\mathbb{R}= R\Omega\), where \(R\) is independent of the isotropic three-dimensional unit vector \(\Omega\). Explicit expressions are given for special values of the stability parameter. Reviewer's remark: A different algorithm for generating multivariate (any dimension) symmetric stable and Linnik random vectors is to be found in the paper of \textit{C. Jacques, B. Rémillard} and the reviewer [Stat. Decis. 17, No. 3, 213-235 (1999)].