Statistical distributions

Statistical Distributions are used to model non-deterministic processes in the model, e.g. processing times. Depending on the process there might be different distributions to choose from. This topic lists and describes all distributions that can be used in FACTS-Analyzer.

In the distribution editor the desired distribution is selected in a drop-down box and once a distribution is selected input-fields for its parameters appear below the drop-down box. Parameters marked with a red star are required whereas the ones without are optional. At the bottom of the distribution editor the distribution is shown as a string, this is also how it will be shown in the Property panel.

Constant

The simplest of distributions as it always returns a specified fixed value.

Parameter

Description

Value

Desired positive real number. Range [0,∞).


Triangular

The triangular distribution returns a value within certain (user defined) bounds, but the values within those bounds aren't equally likely. One value (the Mode) are more likely than the rest and the probability of the other values decreases linearly from that value towards the borders of the bound.

See wikipedia for a more detailed description of the Triangular distribution.

Parameter

Description

Mode

The most common value, i.e. the value with the highest probability. Range [Min,Max].

Note: Not the same as the mean value.

Min

The lower bound for the range of allowed values. Range [0,Mode], where Min ≠ Max.

Max

The upper bound for the range of allowed values. Range [Mode,∞), where Max ≠ Min.


Exponential

The exponential distribution returns values representing the time between events that belongs to a process in which events occur independently of each other and at a constant average rate equal to the Mean of the distribution.

See wikipedia for a more detailed description of the Exponential distribution.

Parameter

Description

Mean

The mean value of the distribution. Range (Min,Max).

Note: Not the same as the mean value.

Min

The lower bound for the range of allowed values, works as a positive offest. Range [0,Mean), where Min ≠ Max.

Note: The mean value of the distribution must fall between Min and Max.

Max

The upper bound for the range of allowed values, truncates the distribution. Range (Mean,∞), where Max ≠ Min.

Note: The mean value of the distribution must fall between Min and Max.


Uniform

Returns a value within certain (user defined) bounds and all values within those bounds are equally likely.

See wikipedia for a more detailed description of the Uniform distribution.

Parameter

Description

Min

The lower bound for the range of allowed values. Range [0,Max), where Min ≠ Max.

Max

The upper bound for the range of allowed values. Range (Min,∞), where Max ≠ Min.


Normal

The normal distribution returns values that represent a process that have values that tend to cluster around a single mean value.

See wikipedia for a more detailed description of the Normal distribution.

Parameter

Description

Mean

The mean value of the distribution. Range [Min,Max].

Sigma

The standard deviation of the distribution. Range [0,∞).

Min

The lower bound for the range of allowed values, truncates the distribution by resampling until allowed value. Range [0,Max), where Min ≠ Max.

Note: The mean value of the distribution must fall between Min and Max.

Max

The upper bound for the range of allowed values, truncates the distribution by resampling until allowed value. Range (Min,∞), where Max ≠ Min.

Note: The mean value of the distribution must fall between Min and Max.


Lognormal

The lognormal distribution returns values that represent a process that can be described by the product of many independent positive factors whose values are close to 1.

See wikipedia for a more detailed description of the Lognormal distribution.

Parameter

Description

Mean

The mean value of the distribution. Range [Min,Max].

Sigma

The standard deviation of the distribution. Range [0,∞).

Min

The lower bound for the range of allowed values, truncates the distribution by resampling until allowed value. Range [0,Max), where Min ≠ Max.

Note: The mean value of the distribution must fall between Min and Max.

Max

The upper bound for the range of allowed values, truncates the distribution by resampling until allowed value. Range (Min,∞), where Max ≠ Min.

Note: The mean value of the distribution must fall between Min and Max.


Beta

With the two shape parameters Alpha, and Beta of the beta distribution the shape of the distribution can take a wide range of forms. This together with the possibility to linearly transform the distribution from its default support interval [0,1] makes it a a very versatile distribution.

See wikipedia for a more detailed description of the Beta distribution.

Parameter

Description

Alpha

Shape parameter of the destribution. Range (0,∞).

Beta

Shape parameter of the destribution. Range (0,∞).

Min

The lower bound for the range of allowed values. Used to linearly transform the distribution from default [0,1] to [Min,Max]. Range [0,Max), default value 0.

Max

The upper bound for the range of allowed values. Used to linearly transform the distribution from default [0,1] to [Min,Max]. Range (Min,∞), default value 1.


Gamma

The Gamma distribution is defined by a Shape and a Scale parameter and can also be truncated using the Min and Max parameters.

See wikipedia for a more detailed description of the Gamma distribution.

Parameter

Description

Shape

Shape parameter of the destribution. Range (0,∞).

Scale

Shape parameter of the destribution. Range (0,∞).

Min

The lower bound for the range of allowed values, truncates the distribution by resampling until allowed value. Range [0,Max), where Min ≠ Max.

Note: The mean value of the distribution must fall between Min and Max.

Max

The upper bound for the range of allowed values, truncates the distribution by resampling until allowed value. Range (Min,∞), where Max ≠ Min.

Note: The mean value of the distribution must fall between Min and Max.


Weibull

The Weibull distribution is defined by a Shape and a Scale parameter and can also be truncated using the Min and Max parameters.

See wikipedia for a more detailed description of the Weibull distribution.

Parameter

Description

Shape

Shape parameter of the destribution. Range (0,∞).

Scale

Shape parameter of the destribution. Range (0,∞).

Min

The lower bound for the range of allowed values, truncates the distribution by resampling until allowed value. Range [0,Max), where Min ≠ Max.

Note: The mean value of the distribution must fall between Min and Max.

Max

The upper bound for the range of allowed values, truncates the distribution by resampling until allowed value. Range (Min,∞), where Max ≠ Min.

Note: The mean value of the distribution must fall between Min and Max.


Discrete Uniform

Discrete version of the uniform distribution that returns integer values within certain (user defined) bounds and all integer values within those bounds are equally likely.

See wikipedia for a more detailed description of the Discrete uniform distribution.

Parameter

Description

Min

The lower bound for the range of allowed values. Range [0,1,2,...,Max), where Min ≠ Max.

Max

The upper bound for the range of allowed values. Range (Min, Min + 1, Min + 2,...,∞), where Max ≠ Min.


Binomial

A discrete probability distribution used to model the number of successes n*p from n independent experiments.

See wikipedia for a more detailed description of the Binomial distribution.

Parameter

Description

n

Number of independent experiements. Range [0,1,2,...,∞).

p

Probability of success in an experiment. Range [0,1].


Negative Binomial

A discrete probability distribution used to model the number of failures before r successes are observed in a series of independent experiments.

See wikipedia for a more detailed description of the Negative binomial distribution.

Parameter

Description

r

Number of desired successes. Range [0,1,2,...,∞).

p

Probability of success in an experiment. Range [0,1].