Package qmwi.kseg.som

Class CSV_SOM_dataEntry

java.lang.Object

qmwi.kseg.som.CSV_SOM_dataEntry

All Implemented Interfaces:

SOMdataEntry

public class CSV_SOM_dataEntry
extends Object
implements SOMdataEntry

Constructor Summary

Constructors

Constructor	Description
CSV_SOM_dataEntry(int size)
CSV_SOM_dataEntry(int columnCount, Object userData)

Method Summary

Modifier and Type	Method	Description
SOMdataEntry	addValues(String inputLine)
SOMdataEntry	addValues(String inputLine, boolean normalized)
String[]	getColumnData()
String	getColumnData(int i)
ArrayList<Double>	getDifferencesToCentroids()
double	getMinDiff()
Object	getUserData()
boolean	isAlreadyNormalized()
void	setDifferences(ArrayList<Double> differences)
void	setUserData(Object data)
String	toString()

Methods inherited from class java.lang.Object

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

Constructor Details

CSV_SOM_dataEntry

public CSV_SOM_dataEntry(int size)

CSV_SOM_dataEntry

public CSV_SOM_dataEntry(int columnCount,
Object userData)

Parameters:

columnCount -

n -

Method Details

toString

public String toString()

Overrides:

toString in class Object

getColumnData

public String[] getColumnData()

Specified by:

getColumnData in interface SOMdataEntry

getUserData

public Object getUserData()

Specified by:

getUserData in interface SOMdataEntry

setUserData

public void setUserData(Object data)

Specified by:

setUserData in interface SOMdataEntry

addValues

public SOMdataEntry addValues(String inputLine,
boolean normalized)

Specified by:

addValues in interface SOMdataEntry

addValues

public SOMdataEntry addValues(String inputLine)

Specified by:

addValues in interface SOMdataEntry

getColumnData

public String getColumnData(int i)

Specified by:

getColumnData in interface SOMdataEntry

isAlreadyNormalized

public boolean isAlreadyNormalized()

Specified by:

isAlreadyNormalized in interface SOMdataEntry

Returns:

True, if the supplied data is already normalized in the range of -1..1. If this value is False, The SOM train and recall method will normalize the data column wise! This means, e.g. all values for time point 1 will be collected and then normalized, thus from day to day a different normalization is performed. This is useful while analyzing independent data points. For time series data this is clearly not a good idea, in this case the data should be supplied normalized and this flag should return true.

getDifferencesToCentroids

public ArrayList<Double> getDifferencesToCentroids()

Specified by:

getDifferencesToCentroids in interface SOMdataEntry

setDifferences

public void setDifferences(ArrayList<Double> differences)

Specified by:

setDifferences in interface SOMdataEntry

getMinDiff

public double getMinDiff()

Specified by:

getMinDiff in interface SOMdataEntry