distrand

A Random Number Generation modules for users of Godot Engine.

Purpose

The goal of the project is to provide a convenient and efficient tool for generating random numbers with non-uniform distribution, which may then be used in games either through GDScript or by other modules. As a C++ modules it offers execution advantages over scripted solutions.

Using distrand

Types

Currently distrand exposes two object classes, RealNormal and IntNormal. Both create normal (or Gaussian) distributions of random numbers.

RealNormal generates random value floats and holds them in a RealArray.

IntNormal generates random value integers and holds them in an IntArray.

Public Methods

• setparameters
• generate
• getvalue
• getnext
• getsingle

setparameters( mean , deviation )

Takes two values, type-matched to the variant (ints for IntNormal, floats for RealNormal), and sets the parameters for the next set of numbers to be generated if those parameters are valid. Erases any set of values previously generated in the object instance.

generate( count )

Where count is an integer. Generates count random values according to the normal distribution defined by stored parameters (mean and deviation).

getvalue( i )

Where i is an integer lesser than the count last generated. Returns the value of the random number in position i of the internal array. RealNormal returns a float while IntNormal returns an integer.

getnext()

Starting at the first element, sequentially returns each value resultant from the previous generation. If getvalue has been called it continues from there.

getsingle()

Generates and returns a new random number under the currently-defined curve.

GDScript Instantiation

RealNormal and IntNormal objects may be created like other objects in GDScript:

var foo=RealNormal.new()

To define and generate the distribution:

foo.normal(5.0, 1.5)

foo.generate(50000)

To access the data:

n = foo.getvalue(42)

n = foo.getnext()

Future features

• Histogram-friendly data in Int* distributions, giving a count of each value.
• Heart-cut and dog-food subsets of normal distributions.
• Bi- and Multi-modal distributions.
• Weibull distributions.