X4.1 Random value

Random value generation is governed by a common algorithm, internally implemented as an extension of the standard Random class. In RCNet, random values are used very often and it is usually necessary to specify parameters for the random value generation.
The common algorithm has parameters that match the "RandomValueType" type in RCNetTypes.xds defined as follows:

  <!-- Random value -->
  <xs:complexType name="RandomValueType">
    <xs:choice minOccurs="0" maxOccurs="1">
      <xs:element type="UniformDistrType" name="uniformDistr" minOccurs="1" maxOccurs="1"/>
      <xs:element type="GaussianDistrType" name="gaussianDistr" minOccurs="1" maxOccurs="1"/>
      <xs:element type="ExponentialDistrType" name="exponentialDistr" minOccurs="1" maxOccurs="1"/>
      <xs:element type="GammaDistrType" name="gammaDistr" minOccurs="1" maxOccurs="1"/>
    </xs:choice>
    <xs:attribute type="xs:double" name="min" use="required"/>
    <xs:attribute type="xs:double" name="max" use="required"/>
    <xs:attribute type="xs:boolean" name="randomSign" use="optional" default="false">
      <xs:annotation>
        <xs:documentation xml:lang="en">Default value is false</xs:documentation>
      </xs:annotation>
    </xs:attribute>
  </xs:complexType>

Xml element of a "RandomValueType" type

Example:

  <exampleRandomValue min="-1" max="1" randomSign="false">
    <uniformDistr/>
  </exampleRandomValue>

Attribute	Description
min,max	The base range of the random value is defined by the "min" and "max" borders. Parameters are real numbers, so they can be negative. The condition that "min" is less than or equal to "max" must be met
randomSign	Parameter may be true or false and determines whether the sign of the selected value will be additionally randomly changed

In RCNet are implemented following random distributions:

• Uniform
• Gaussian (normal)
• Exponential
• Gamma

Uniform distribution

Gaussian distribution

Exponential distribution

Gamma distribution

Random value algorithm pseudocode

1. Generate random value within the range "min" and "max", using specified "distribution"
2. If the "randomSign" is set to true
   • randomly generate a sign (-1 or 1)
   • multiply random value by the generated sign

Hints

• If you want to generate random values from continuous range, specify appropriate "min" and "max" and set "randomSign" to "false". For example specify "min"=-2.5, "max"=3.1 to generate random values within the range <-2.5, 3.1>
• If you want to generate random values within the positive and negative range, specify positive "min" and "max" and set "randomSign" to "true". For example specify "min"=1, "max"=2 to generate random values within the ranges <-2, -1> and <1, 2>

Random class extension methods

--> Next page