Generating beta variates with nonintegral shape parameters

ABSTRACT

A new rejection method is described for generating beta variates. The method is compared with previously published methods both theoretically and through computer timings. It is suggested that the method has advantages in both speed and programming simplicity over previous methods, especially for “difficult” combinations of parameter values.

REFERENCES

• 1: Ahrens, J.H., and Dieter, U. Computer methods for sampling from gamma, beta, Poisson and binomial distribution. Computing (Vienna), 12 (1974), 223-246.
• 2: Atkinson, A.C., and Pearce, M.C. The computer generation of beta, gamma and normal random variables. J. Roy. Statist. Soc. Ser. A 139 (1976), 431-461.
• 3: Atkinson, A.C., and Whittaker, J. A switching algorithm for the generation of beta random variables with at least one parameter less than 1. J. Roy. Statist. Soc. Ser. A. 139 (1976), 462-467.
• 4: J~hnk, M.D. Erzeugung von Betaverteilten und Gammaverteilten Zuffallszahlen. Metrika 8 (1964), 5-15.
• 5: Marsaglia, G., and Bray, T.A. A convenient method for generating normal variables. SIAM Rev. 6 (1964), 260-264.
• 6: N. D. Wallace, Computer generation of gamma random variates with non-integral shape parameters, Communications of the ACM, v.17 n.12, p.691-695, Dec 1974 [doi>10.1145/361604.361621]
• 7: Whittaker, J. Generating gamma and beta random variables with non-integral shape parameters. Appl. Statist. 23 (1974), 210- 214.

INDEX TERMS

Primary Classification:
  G. Mathematics of Computing

Additional Classification:
  I. Computing Methodologies
  I.6 SIMULATION AND MODELING

General Terms:
Design, Performance, Theory

Keywords:
beta variates, random numbers, simulation