SIMULATION · 1983 · 96 citations · 8 references
ComputingComputer ScienceComputer SystemsAlgorithmsSimulation-based DesignStatistical ComputingComputational StatisticsRandom Number GenerationSimulation MethodsSoftware TestingIndependenceRandom Number Generator TestingTesting
The implementation and testing of a portable FORTRAN routine is described which utilizes a well-known random number generator to produce Uniform (0,1) values. The implementa tion techniques illustrate the advantages of a well-tested por table algorithm and expose some problems which commonly occur in the generation of Uniform (0, 1) numbers for applica tion areas such as simulation. These problems include the pit falls of using machine-dependent random number generators with unknown characteristics typically supplied with computer system software. The excellent global statistical properties of the generator used in this routine are documented and local testing of the Uniform (0,1) values produced is provided. An additional routine is described which enables the user to obtain indepen dent streams of values. Fifteen streams are rated for their per formance on uniformity and independence tests. The provid ed seed values enable the user to select streams with the desired properties for a particular application.
8
The Art of Computer Programming
G.E. Whitesides · Nuclear Science and Engineering · 1970 · 6.1K citations
Statistical Principles in Experimental Design
Paul E. Green, B. J. WINER, Donald R. Brown et al. · Journal of Marketing Research · 1992 · 4.4K citations
Principles of Discrete Event Simulation
Jack Marshall · Electronics and Power · 1979 · 341 citations
T. E. Hull, A. R. Dobell · SIAM Review · 1962 · 335 citations