Package microsim.alignment.probability

Class MultiplicativeScalingAlignment<T>

java.lang.Object

microsim.alignment.probability.MultiplicativeScalingAlignment<T>

Type Parameters:

T - A class usually representing an agent.

All Implemented Interfaces:

AlignmentUtils<T>

public class MultiplicativeScalingAlignment<T>
extends Object
implements AlignmentUtils<T>

A class for multiplicative scaling alignment method. All probabilities are extracted from the collection agents using provided filter. Further steps involve calculation of the ration between the desired transition rate and the actual transition to be used as a scaling factor for probabilities.

See Also:

• Jinjing Li and Cathal O'Donoghue, Evaluating Binary Alignment Methods in Microsimulation Models, Journal of Artificial Societies and Social Simulation 17 (1) 15
• Li, Jinjing and O'Donoghue, Cathal (2014) 'Evaluating Binary Alignment Methods in Microsimulation Models' Journal of Artificial Societies and Social Simulation 17 (1) 15

Constructor Summary

Constructors

Constructor	Description
MultiplicativeScalingAlignment()

Method Summary

Modifier and Type	Method	Description
double @NonNull []	align(@NonNull Collection<T> agents, @Nullable org.apache.commons.collections4.Predicate<T> filter, @NonNull AlignmentProbabilityClosure<T> closure, double targetProbability)	The main alignment method.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Methods inherited from interface microsim.alignment.AlignmentUtils

extractAgentList, validateProbabilityValue

Constructor Details

MultiplicativeScalingAlignment

public MultiplicativeScalingAlignment()

Method Details

align

public double @NonNull [] align(@NonNull
 @NonNull Collection<T> agents,
 @Nullable
 @Nullable org.apache.commons.collections4.Predicate<T> filter,
 @NonNull
 @NonNull AlignmentProbabilityClosure<T> closure,
 double targetProbability)

The main alignment method. It involves scaling the existing probabilities by multiplying by a factor. Its value is the target probability divided by the sum of all initial probabilities. The resulting values then are compared to a random number from the (0, 1) range to decide if the transition happens.

Parameters:

agents - An unfiltered collection of agents.

filter - A logical filter for agents, can be null.

closure - A closure to handle probabilities of agents.

targetProbability - The target probability value.

Returns:

The corrected set of outcomes.

Throws:

NullPointerException - when agents, or closure, or both are null.