Predefined Distributions

The KP op_dist_load() performs the initial set-up of a distribution, which can be used to generate a series of stochastic numeric values. There are two types of distributions: predefined distributions supplied by Riverbed and those created by the user using the PDF Editor. The two types of distributions are technologically different: predefined distributions are actually algorithms that are invoked to provide values when needed, whereas the PDF Editor creates distributions based on tabular collections of data. The exception to this is the fast_normal distribution, which is equivalent to the normal distribution, but read from a table, requiring less processing time. The way in which the two types of distribution are referenced varies as well. Predefined distributions can have one or two arguments, which are parameters that affect the distribution algorithm. These arguments specify the exact distribution needed via op_dist_load(). On the other hand, distributions created using the PDF Editor ignore the arguments of op_dist_load(). These types of distributions are simply model files that are read in by op_dist_load().

The list of the predefined distributions appears below, along with their supported arguments. For more information on the mathematical framework of these distributions, consult the textbook Fundamentals of Applied Probability Theory, by Dr. Alvin Drake, published in 1967 by McGraw-Hill.

Table 8: Predefined Distributions

Distribution Type Name Argument #0 Argument #1
Bernoulli PMF bernoulli mean outcome <ignored>
Binomial PMF binomial num_samples p_success
Chi-Square PDF chi_square mean outcome <ignored>
Constant PMF constant outcome <ignored>
Erlang PDF erlang scale shape
Exponential PDF exponential mean outcome <ignored>
Extreme Value (Gumbel) PDF extreme location mode_scale
Gamma PDF gamma scale shape
Geometric PMF geometric prob_success <ignored>
Laplace (Double Exponential) PDF laplace mean scale
Logistic PDF logistic mean scale
Lognormal PDF lognormal mean outcome outcome variance
Normal and Fast Normal PDF normal
fast_normal
mean outcome
mean outcome
outcome variance
outcome variance
Pareto PDF pareto location shape
Poisson PMF poisson mean outcome <ignored>
Power Function PDF power function shape scale
Rayleigh PDF rayleigh scale <ignored>
Triangular (Symmetrical) PDF triangular min max
Uniform PDF uniform minimum outcome (inclusive) maximum outcome (exclusive)
Uniform Integer PMF uniform_int minimum outcome (inclusive) maximum outcome (inclusive)
Weibull CDF Weibull shape parameter scale parameter