Probability Distributions

You may need to model non-deterministic processes, e.g. weather changing, product demand changing, or a random choice. Stochastic processes can be modeled using probability distributions.

AnyLogic comes with a large set of probability distributions. To obtain a random value, call the corresponding function, e.g. exponential( 0.6 ).

All AnyLogic distribution classes are implemented on the basis of the standard Java java.util.Random random number generator. However, if you wish to use your own one by default, you should call the setDefaultRandomGenerator( java.util.Random gen ) static function of the Utilities class.

Using the following demo model you can explore different probability distributions supported by AnyLogic:

AnyLogic supports the following probability distributions:

• bernoulli
• beta
• beta (truncated)
  
• binomial
• binomial (truncated)
  
• cauchy
• chi2
• erlang
• exponential
• exponential (truncated)
  
• gamma
• gamma (truncated)
  
• geometric
• gumbel1
• gumbel2
• hypergeometric
• laplace
• logarithmic
• logistic
• lognormal
• negativeBinomial
• negativeBinomial (truncated)
  
• normal
• normal (truncated)
  
• pareto
• pert
• poisson
• poisson (truncated)
  
• randomFalse
• randomTrue
• rayleigh
• triangular
• triangular (truncated)
  
• triangularAV
• uniform
• uniform_discr
• uniform_pos
• weibull
• weibull (truncated)
  
• Custom (empirical) distribution