@ Copyright 2003, Jeff Miller.
This program and its documentation may be duplicated and used without charge for any educational or noncommercial purposes. For commercial use, please contact the author.
This program was designed for use in generating random numbers for use in randomization procedures, statistical simulation, etc.
The user specifies the number of cases (i.e., sample size), the number of variables per case, the underlying distribution of each variable (e.g., normal, with specified mean and standard deviation), and the matrix of correlations between variables. The program then generates a random sample from a population satisfying these conditions. The random values are saved to a file for subsequent analysis by a simulation program, statistical package, etc.
RandGen can be used to generate random variables from many common probability distributions, including the normal, exponential, gamma, Weibull, uniform, and so on. In addition, it can generate random variables from various transformed versions of these distributions (e.g., convolutions, mixtures, order statistics, censored or truncated distributions, etc.). A further strength of RandGen is that the different variables need not all have the same type of probability distribution. For example, it can generate random triples in which the first variable has a normal distribution, the second has a uniform distribution, and the third has an exponential distribution, with user-specified correlations among these variables.