Note that the discrete distributions, including the binomial, negative binomial, Poisson & Bernoulli, are all mathematically defined as discrete functions: only integral values of the random variate are envisaged and the functions are only defined at these integral values. However because the method of calculation often uses continuous functions, it is convenient to treat them as if they were continuous functions, and permit non-integral values of their parameters. To enforce a strict mathematical model, users may use floor or ceil functions on the random variate, prior to calling the distribution function, to enforce integral values. For similar reasons, in continuous distributions, parameters like degrees of freedom that might appear to be integral, are treated as real values (and are promoted from integer to floating-point if necessary). In this case however, that there are a small number of situations where non-integral degrees of freedom do have a genuine meaning. Generally speaking there is no loss of performance from allowing real-values parameters: the underlying special functions contain optimizations for integer-valued arguments when applicable.
|