ABSTRACT

Within this definition of an inverse cdf, the standard uniform random Z plays dual roles. It not only appears within the arguments of the two functions, ln(2Z) and Zn(2 ( l — Z )), but it also provides the arguments of the indicator functions /(o,o.5](^) and /(o.5,i)(Z), that determine the choice of either ln(2Z) or Zn(2 (l — Z)). In these two ways, Z not only determines what should happen once the decision to change gears has been made, but it also determines when gears should be changed. Surely this is a lot to ask of the single simple random quantity Z.