Pseudo Random Number Generators for Statistical Applications

Abstract

Pseudo random number generating programs coded in machine language for the IBM 360/65 computer have been prepared for producing random values from the normal, exponential, and gamma distributions as well as from any discrete probability distribution. The schemes are based largely on combinations of sophisticated techniques first suggested by Marsaglia. The distributions simulated are exact within the word size of the computer and average production time per number is very short, e.g. 10 to 40 microseconds depending upon the particular distribution.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Aug 01, 1971
Accession Number
AD0735144

Entities

People

  • Lovick E. Cannon Iii

Organizations

  • University of Georgia

Tags

Communities of Interest

  • Energy and Power Technologies

DTIC Thesaurus Topics

  • Computer Programs
  • Computers
  • Discrete Distribution
  • Distribution Functions
  • Goodness Of Fit Tests
  • Monte Carlo Method
  • Normal Distribution
  • Plastic Explosives
  • Precision
  • Probability
  • Probability Distributions
  • Random Number Generators
  • Random Variables
  • Statistical Sampling
  • Statistical Tests
  • Three Dimensional
  • Two Dimensional

Readers

  • Computer Programming and Software Development.
  • Computer Science.
  • Statistical inference.