desmoj.core.dist

Class DiscreteDistEmpirical<N extends java.lang.Number>

java.lang.Object

desmoj.core.simulator.NamedObject

desmoj.core.simulator.ModelComponent

desmoj.core.simulator.Reportable

desmoj.core.dist.Distribution

desmoj.core.dist.NumericalDist<N>

desmoj.core.dist.DiscreteDist<N>

desmoj.core.dist.DiscreteDistEmpirical<N>

public class DiscreteDistEmpirical<N extends java.lang.Number>
extends DiscreteDist<N>

Empirically distributed stream of pseudo random numbers of a custom type. Values produced by this distribution follow an empirical distribution which is specified by entries consisting of the observed value and the frequency (probability) this value has been observed to occur. These entries are made by using the addEntry() method.

Version:

DESMO-J, Ver. 2.5.1c copyright (c) 2015

Author:

Tim Lechler Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

See Also:

Distribution

Constructor Summary

Constructors

Constructor and Description
DiscreteDistEmpirical(Model owner, java.lang.String name, boolean showInReport, boolean showInTrace) Constructs an empirical distribution producing double values.

Method Summary

Modifier and Type	Method and Description
void	addEntry(N value, double frequency) Adds a new entry of an empirical value and its associated frequency.
Reporter	createDefaultReporter() Creates the default reporter for the EmpiricalDiscreteDist distribution.
N	getInverseOfCumulativeProbabilityFunction(double p) Abstract method to map a double p from 0...1 to the distribution's domain by determining the value x that satisfies P(X < x) = p.
boolean	isInitialized() Shows if the EmpiricalDiscreteDist distribution already is initialized.
void	removeEntry(N value) Removes all entries of an empirical value.
void	removeEntry(N value, double tolerance) Removes all entries of all empirical values close to a given value.
N	sample() Returns the next sample specified by the empirical distribution.

Methods inherited from class desmoj.core.dist.NumericalDist

sampleObject, sampleTimeSpan, sampleTimeSpan

Methods inherited from class desmoj.core.dist.Distribution

changeRandomGenerator, getInitialSeed, getNonNegative, getNumSamples, isAntithetic, reset, reset, setAntithetic, setNonNegative, setSeed

Methods inherited from class desmoj.core.simulator.Reportable

getCorrespondingSchedulable, getDescription, getObservations, getReporter, incrementObservations, incrementObservations, reportIsOn, reportOff, reportOn, resetAt, setCorrespondingSchedulable, setDescription, setReporter

Methods inherited from class desmoj.core.simulator.ModelComponent

current, currentEntity, currentEntityAll, currentEvent, currentModel, currentSimProcess, debugIsOn, debugOff, debugOn, getModel, isExperimentCompatible, isModelCompatible, presentTime, sendDebugNote, sendMessage, sendTraceNote, sendWarning, skipTraceNote, skipTraceNote, traceIsOn, traceOff, traceOn

Methods inherited from class desmoj.core.simulator.NamedObject

getName, getQuotedName, toString

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, wait, wait, wait

Constructor Detail

DiscreteDistEmpirical

public DiscreteDistEmpirical(Model owner,
                             java.lang.String name,
                             boolean showInReport,
                             boolean showInTrace)

Constructs an empirical distribution producing double values. Empirical distributions have to be initialized manually before use. This is done by calling the addEntry(double, double) method to add values defining the behaviour of the desired distribution.s

Parameters:

owner - Model : The distribution's owner

name - java.lang.String : The distribution's name

showInReport - boolean : Flag for producing reports

showInTrace - boolean : Flag for producing trace output

Method Detail

addEntry

public void addEntry(N value,
                     double frequency)

Adds a new entry of an empirical value and its associated frequency.

Parameters:

value - N : The empirical value observed

frequency - double : The corresponding frequency of the empirical value

removeEntry

public void removeEntry(N value)

Removes all entries of an empirical value.

Parameters:

value - N : The value to be removed

removeEntry

public void removeEntry(N value,
                        double tolerance)

Removes all entries of all empirical values close to a given value. This method is useful for removing floating point values whose internal precision may be subject to rounding errors and depend on the JVM in use.
Example: In a DiscreteDistEmprirical after calling addEntriy(42.42, 5), you might find the value actually stored is 42.42000000000001. To remove this entry independently of its precise internal representation, use e.g. removeEntry(42.42, 0.0001). (This assumes all other values stored in this distribution differ by more than 0.001 from 42.42.)

Parameters:

value - N : The value to be removed

tolerance - double : The value to be removed

createDefaultReporter

public Reporter createDefaultReporter()

Creates the default reporter for the EmpiricalDiscreteDist distribution.

Overrides:

createDefaultReporter in class Distribution

Returns:

Reporter : The reporter for the EmpiricalDiscreteDist distribution

See Also:

DiscreteDistEmpiricalReporter

isInitialized

public boolean isInitialized()

Shows if the EmpiricalDiscreteDist distribution already is initialized. Being initialized means that at least one value has already been added via the addEntry(double, double) method.

Returns:

boolean

sample

public N sample()

Returns the next sample specified by the empirical distribution. In contrast to RealDistEmpirical here is no interpolation needed.

Overrides:

sample in class NumericalDist<N extends java.lang.Number>

Returns:

N : The next sample for this empirical distribution or returns zero (0) with a warning if the distribution has not been properly initialized yet

getInverseOfCumulativeProbabilityFunction

public N getInverseOfCumulativeProbabilityFunction(double p)

Abstract method to map a double p from 0...1 to the distribution's domain by determining the value x that satisfies P(X < x) = p.

Specified by:

getInverseOfCumulativeProbabilityFunction in class NumericalDist<N extends java.lang.Number>

Parameters:

p - double: A value between 0 and 1

Returns:

N : The value x that satisfies P(X < x) = p