com.sun.faban.driver.util.timermeter
Class MeasurementGroupsCluster

java.lang.Object
  com.sun.faban.driver.util.timermeter.MeasurementGroupsCluster

All Implemented Interfaces:

Serializable

public class MeasurementGroupsCluster
extends Object
implements Serializable

A cluster which consists of key-value pairs that "aggregate" same-valued cluster elements. That is, instead of containing two items with the value 10, this cluster contains one entry with key 10 and value 2. Elements cannot be removed from the cluster. TODO optimise through "lazy computation" of median value and slot size

Author:

Michael Kuperberg TODO test me TODO integrate with WEKA and similar?

See Also:

Serialized Form

Constructor Summary

MeasurementGroupsCluster()

MeasurementGroupsCluster(Long key, Integer value)

MeasurementGroupsCluster(SortedMap<Long,Integer> clusterElements)

Method Summary

double	addData(Long key, Integer value) TODO check for negatives?
double	addData(Map<Long,Integer> keyValueMap)
int	getClassesNumberInCluster()
long	getClusterGroupSizeMaximum()
double	getClusterGroupSizeMean()
long	getClusterGroupSizeMedian()
long	getClusterGroupSizeMinimum()
long	getClusterGroupSizeSum()
long	getClusterValueMaximum() //TODO test me
double	getClusterValueMean()
double	getClusterValueMedian() TODO more effective median *re*computation? --> think about it! TODO test me
long	getClusterValueMinimum() TODO test me
long	getClusterValueSum()
String	toString()

Methods inherited from class java.lang.Object

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

Constructor Detail

MeasurementGroupsCluster

public MeasurementGroupsCluster()

MeasurementGroupsCluster

public MeasurementGroupsCluster(Long key,
                                Integer value)

MeasurementGroupsCluster

public MeasurementGroupsCluster(SortedMap<Long,Integer> clusterElements)

Method Detail

addData

public double addData(Long key,
                      Integer value)

TODO check for negatives?

Parameters:

key -

value -

Returns:

addData

public double addData(Map<Long,Integer> keyValueMap)

getClassesNumberInCluster

public int getClassesNumberInCluster()

getClusterGroupSizeMaximum

public long getClusterGroupSizeMaximum()

getClusterGroupSizeMean

public double getClusterGroupSizeMean()

getClusterGroupSizeMedian

public long getClusterGroupSizeMedian()

getClusterGroupSizeMinimum

public long getClusterGroupSizeMinimum()

getClusterGroupSizeSum

public long getClusterGroupSizeSum()

getClusterValueMaximum

public long getClusterValueMaximum()

//TODO test me

Returns:

getClusterValueMean

public double getClusterValueMean()

getClusterValueMedian

public double getClusterValueMedian()

TODO more effective median *re*computation? --> think about it! TODO test me

Returns:

getClusterValueMinimum

public long getClusterValueMinimum()

TODO test me

Returns:

getClusterValueSum

public long getClusterValueSum()

toString

public String toString()

Overrides:

toString in class Object