Generation of random variates from statistical distributions

Work has been done in generating random variates of some distributions for simulation purposes. A subroutine which gives a random number of any common statistical distribution desired, and a compilation of the backgrounds, and derivations for the generating methods gives convenience and understanding to anyone who does simulation work. This thesis presents a brief history on the generation of random variables, a compilation and investigation of old and new generating methods, and comparisons and evaluations of different methods by chi-square and t-tests. The best method of each distribution investigated is selected to incorporate into subroutines CTIME or DTIME, depending on whether the distribution is continuous or discrete.