Class %DeepSee.extensions.stat.MixedDistribution

class %DeepSee.extensions.stat.MixedDistribution extends %RegisteredObject, AbstractContinuousDistribution

Represents multicomponent continues distribution. For example let the user to create a distribution consisting of mixed Gaussian components or a mixed Normal/Exponential distribution.

Here is an example of it's usage:

	set ndist = ##class(%DeepSee.extensions.stat.NormalDistribution).create(m, (sigma * sigma))
	set edist = ##class(%DeepSee.extensions.stat.ExpDistribution).create(lambda)
	set mdist = ##class(%DeepSee.extensions.stat.MixedDistribution).%New()
	set sc = mdist.AddComponent(ndist, w)
	Quit:$$$ISERR(sc)
	set sc = mdist.AddComponent(edist, (1-w))
	Quit:$$$ISERR(sc)
	set sc = mdist.balance()
	Quit:$$$ISERR(sc)
	w "Mean = "_mdist.Mean
	w "; Var = "_mdist.Variance
	w "; Entropy = "_mdist.Entropy, !

Inventory

Parameters	Properties	Methods	Queries	Indices	ForeignKeys	Triggers
	3	4

Summary

Properties
Entropy	Mean	N	P	Variance
w

Methods
%%OIDGet	%AddToSaveSet	%ClassIsLatestVersion	%ClassName
%ConstructClone	%DispatchClassMethod	%DispatchGetModified	%DispatchGetProperty
%DispatchMethod	%DispatchSetModified	%DispatchSetMultidimProperty	%DispatchSetProperty
%Extends	%GetParameter	%IsA	%IsContinuous
%IsModified	%New	%NormalizeObject	%ObjectModified
%OriginalNamespace	%PackageName	%RemoveFromSaveSet	%SerializeObject
%SetModified	%ValidateObject	AddComponent	GetDiscreteDistribution
GetKullbackLeiblerDivergence	GetLogLikelihood	balance	cdf
pdf

Properties

• property N as %Integer [ InitialExpression = 0 ];

• property P as AbstractContinuousDistribution [ MultiDimensional ];

• property w as %Double [ MultiDimensional ];

Methods

• method AddComponent(p As AbstractContinuousDistribution, weight As %Double) as %Status

Add a component p to the distribution with the given weight. After all components are added application must call balance

• method balance() as %Status

The method to call after all components of the distribution are specified.

• method cdf(x1 As %Double, x2 As %Double = "") as %Double

An interface method for cumulative distribution function (CDF). Optionally provides a cumulative probability of a value being between two values x1 and x2.
If called with 1 argument(x1) or if the value of second argument is empty string then returns CDF value for x1. If called with two arguments, x1 and x2 then returns a probability of a value being between x1 and x2.

• method pdf(x As %Double) as %Double

An interface method for probability distribution function (PDF). Returns PDF value for x.


[BASEXML] > [%DeepSee] > [extensions] > [stat] > [MixedDistribution]		Private Storage