Accuracy in random number generation

Author:
John F. Monahan

Journal:
Math. Comp. **45** (1985), 559-568

MSC:
Primary 65C10; Secondary 68U20

MathSciNet review:
804945

Full-text PDF Free Access

Abstract | References | Similar Articles | Additional Information

Abstract: The generation of continuous random variables on a digital computer encounters a problem of accuracy caused by approximations and discretization error. These in turn impose a bias on the simulation results. An ideal discrete approximation of a continuous distribution and a measure of error are proposed. Heuristic analysis of common methods for transforming uniform deviates to other continuous random variables is discussed. Comments and recommendations are made for the design of algorithms to reduce the bias and avoid overflow problems.

**[1]**Ȧke Björck and Germund Dahlquist,*Numerical methods*, Prentice-Hall, Inc., Englewood Cliffs, N.J., 1974. Translated from the Swedish by Ned Anderson; Prentice-Hall Series in Automatic Computation. MR**0368379****[2]**Philip J. Davis and Philip Rabinowitz,*Methods of numerical integration*, Academic Press [A subsidiary of Harcourt Brace Jovanovich, Publishers] New York-London, 1975. Computer Science and Applied Mathematics. MR**0448814****[3]**Luc Devroye,*A note on approximation in random variate generation*, J. Statist. Comput. Simulation**14**(1981/82), no. 2, 149–158. MR**651481**, 10.1080/00949658208810536**[4]**U. Dieter and J. H. Ahrens,*A combinatorial method for the generation of normally distributed random numbers*, Computing (Arch. Elektron. Rechnen)**11**(1973), no. 2, 137–146 (English, with German summary). MR**0388727****[5]**George E. Forsythe,*von Neumann’s comparison method for random sampling from the normal and other distributions*, Math. Comp.**26**(1972), 817–826. MR**0315863**, 10.1090/S0025-5718-1972-0315863-9**[6]**J. E. Gentle, "Portability considerations for random number generation,"*Comp. Science and Statist.*:*Proc.*13*th Sympos. Interface*(W. F. Eddy, ed.), Springer-Verlag, Berlin and New York, 1981, pp. 158-164.**[7]**A. J. Kinderman & J. F. Monahan, "Computer generation of random variables using the ratio of uniform deviates,"*ACM Trans. Math. Software*, v. 3, 1977, pp. 257-260.**[8]**Donald E. Knuth and Andrew C. Yao,*The complexity of nonuniform random number generation*, Algorithms and complexity (Proc. Sympos., Carnegie-Mellon Univ., Pittsburgh, Pa., 1976) Academic Press, New York, 1976, pp. 357–428. MR**0431601****[9]**P. A. W. Lewis, A. S. Goodman & J. M. Miller, "A pseudorandom number generator for the System/360,"*IBM Systems J.*, v. 8, 1969, pp. 136-146.**[10]**Michel Loève,*Probability theory. I*, 4th ed., Springer-Verlag, New York-Heidelberg, 1977. Graduate Texts in Mathematics, Vol. 45. MR**0651017****[11]**John F. Monahan,*Extensions of von Neumann’s method for generating random variables*, Math. Comp.**33**(1979), no. 147, 1065–1069. MR**528058**, 10.1090/S0025-5718-1979-0528058-7**[12]**R. E. Odeh & J. O. Evans, "Algorithm AS 70: Percentage points of the normal distributions,"*Appl. Statist.*, v. 23, 1974, pp. 96-97.**[13]**L. Schrage, "A more portable Fortran random number generator,"*ACM Trans. Math. Software*, v. 5, 1979, pp. 132-138.**[14]**J. von Neumann, "Various techniques used in connection with random digits," in*Monte Carlo Method*, Nat. Bur. Standards Appl. Math. Series, v. 12, 1951, pp. 36-38.

Retrieve articles in *Mathematics of Computation*
with MSC:
65C10,
68U20

Retrieve articles in all journals with MSC: 65C10, 68U20

Additional Information

DOI:
https://doi.org/10.1090/S0025-5718-1985-0804945-X

Keywords:
Random number generation,
discretization error,
approximation error,
relative accuracy,
floating-point arithmetic

Article copyright:
© Copyright 1985
American Mathematical Society