A Subroutine for Generating Pseudo-Random Numbers from a Variety of Distributions.

Abstract

In this paper, a method for generating pseudo-random numbers is proposed. The method produces good approximations to certain distributions which are difficult to approximate by means of previously published methods. The authors begin by discussing previously published algorithms. Then the authors describe the new algorithm which is given for the case where F, the c.d.f. of the desired distribution, is available as a subroutine but F sup(-1) is not, as is the case with the beta distribution. The FORTRAN 4 code is also provided. This new algorithm preserves qualities of 'randomness' and goodness of fit possessed by the uniform (0,1) random number generator contained in the code. Also presented are results of statistical tests on the generator and instructions for substituting another uniform (0,1) random number generator.

Document Details

Document Type
Technical Report
Publication Date
Apr 01, 1975
Accession Number
ADA009891

Entities

People

  • Charles B. Davis
  • William J. Zimmer

Organizations

  • University of New Mexico

Tags

DTIC Thesaurus Topics

  • Algorithms
  • Computer Programs
  • Data Science
  • Digital Information
  • Generators
  • Information Science
  • Instructions
  • Mathematics
  • Procedures (Computers)
  • Random Number Generators
  • Statistical Algorithms
  • Statistical Tests

Fields of Study

  • Computer science
  • Mathematics

Readers

  • Computational Modeling and Simulation
  • Computer Programming and Software Development.
  • Mathematical Modeling and Probability Theory.