org.deidentifier.arx.aggregates.quality

Class QualityModelRowOrientedAECS

java.lang.Object

org.deidentifier.arx.aggregates.quality.QualityModelRowOrientedAECS

public class QualityModelRowOrientedAECS
extends java.lang.Object

Implementation of the AECS measure, as proposed in:

K. LeFevre, D. DeWitt, R. Ramakrishnan: "Mondrian multidimensional k-anonymity" Proc Int Conf Data Engineering, 2006.

Constructor Summary

Constructors

Constructor and Description
QualityModelRowOrientedAECS(WrappedBoolean interrupt, WrappedInteger progress, int totalWorkload, DataHandle input, DataHandle output, int suppressedInput, int suppressedOutput, Groupify<TupleWrapper> groupedInput, Groupify<TupleWrapper> groupedOutput, java.lang.String[][][] hierarchies, QualityDomainShare[] shares, int[] indices, QualityConfiguration config) Creates a new instance

Method Summary

Modifier and Type	Method and Description
protected void	checkInterrupt() Checks whether an interruption happened.
QualityMeasureRowOriented	evaluate() Evaluates the utility measure
protected double[][]	getColumnsAsNumbers(DataHandle input, DataHandle output, java.lang.String[][] hierarchy, int column) Returns a columns from the input and output dataset converted to numbers
protected QualityDomainShare[]	getDomainShares() Returns the domain shares
protected Groupify<TupleWrapper>	getGroupedInput() Returns grouped input
protected Groupify<TupleWrapper>	getGroupedOutput() Returns grouped output
protected java.lang.String[][][]	getHierarchies() Returns the hierarchies
protected int[]	getIndices() Returns relevant indices
protected DataHandle	getInput() Returns input
protected double[]	getMinMax(double[] column) Returns minimum, maximum for the given column
protected double[]	getNumbersFromHierarchy(DataHandle handle, int column, java.lang.String[][] hierarchy) Returns a numeric representation, relying on the hierarchy
protected double[]	getNumbersFromNumericColumn(DataHandle input, int column) Parses numbers from a numeric input column
protected double[]	getNumbersFromNumericColumn(double[] inputAsNumbers, DataHandle output, int column) Parses numbers from a numeric output column
protected double[]	getNumbersFromNumericColumnAndHierarchy(DataHandle input, double[] inputAsNumbers, DataHandle output, int column, java.lang.String[][] hierarchy) Uses numeric input and a hierarchy to construct ranges
protected DataHandle	getOutput() Returns output
protected double[]	getRangeFromNumericColumn(double[] inputNumbers, DataHandle output, int column) Tries to parse numbers from output when there is a numeric input column
protected int	getSuppressedRecordsInInput() Returns the number of records suppressed in input
protected int	getSuppressedRecordsInOutput() Returns the number of records suppressed in output
protected java.lang.String	getSuppressionString() Returns the suppression string
protected boolean	isSuppressed(DataHandle handle, int[] indices, int row) We assume that an entry is suppressed, if all values are equal
protected boolean	isSuppressed(DataHandle handle, int row, int column) Returns whether a value is suppressed
protected boolean	isSuppressed(int column, java.lang.String value) Returns whether a value is suppressed
protected double	log2(double d) Log base-2
protected void	setStepPerformed() One step performed
protected void	setSteps(int steps) Total number of steps
protected void	setStepsDone() All steps performed

Methods inherited from class java.lang.Object

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

Constructor Detail

QualityModelRowOrientedAECS

public QualityModelRowOrientedAECS(WrappedBoolean interrupt,
                                   WrappedInteger progress,
                                   int totalWorkload,
                                   DataHandle input,
                                   DataHandle output,
                                   int suppressedInput,
                                   int suppressedOutput,
                                   Groupify<TupleWrapper> groupedInput,
                                   Groupify<TupleWrapper> groupedOutput,
                                   java.lang.String[][][] hierarchies,
                                   QualityDomainShare[] shares,
                                   int[] indices,
                                   QualityConfiguration config)

Creates a new instance

Parameters:

interrupt -

progress -

totalWorkload -

input -

output -

suppressedInput -

suppressedOutput -

groupedInput -

groupedOutput -

hierarchies -

shares -

indices -

config -

Method Detail

evaluate

public QualityMeasureRowOriented evaluate()

Evaluates the utility measure

Returns:

checkInterrupt

protected void checkInterrupt()

Checks whether an interruption happened.

getColumnsAsNumbers

protected double[][] getColumnsAsNumbers(DataHandle input,
                                         DataHandle output,
                                         java.lang.String[][] hierarchy,
                                         int column)

Returns a columns from the input and output dataset converted to numbers

Parameters:

input -

output -

hierarchy -

column -

Returns:

getDomainShares

protected QualityDomainShare[] getDomainShares()

Returns the domain shares

getGroupedInput

protected Groupify<TupleWrapper> getGroupedInput()

Returns grouped input

getGroupedOutput

protected Groupify<TupleWrapper> getGroupedOutput()

Returns grouped output

getHierarchies

protected java.lang.String[][][] getHierarchies()

Returns the hierarchies

getIndices

protected int[] getIndices()

Returns relevant indices

getInput

protected DataHandle getInput()

Returns input

Returns:

getMinMax

protected double[] getMinMax(double[] column)

Returns minimum, maximum for the given column

Parameters:

column -

Returns:

getNumbersFromHierarchy

protected double[] getNumbersFromHierarchy(DataHandle handle,
                                           int column,
                                           java.lang.String[][] hierarchy)

Returns a numeric representation, relying on the hierarchy

Parameters:

handle -

column -

hierarchy -

Returns:

getNumbersFromNumericColumn

protected double[] getNumbersFromNumericColumn(DataHandle input,
                                               int column)

Parses numbers from a numeric input column

Parameters:

input -

column -

Returns:

getNumbersFromNumericColumn

protected double[] getNumbersFromNumericColumn(double[] inputAsNumbers,
                                               DataHandle output,
                                               int column)

Parses numbers from a numeric output column

Parameters:

inputAsNumbers -

output -

column -

Returns:

getNumbersFromNumericColumnAndHierarchy

protected double[] getNumbersFromNumericColumnAndHierarchy(DataHandle input,
                                                           double[] inputAsNumbers,
                                                           DataHandle output,
                                                           int column,
                                                           java.lang.String[][] hierarchy)

Uses numeric input and a hierarchy to construct ranges

Parameters:

input -

inputAsNumbers -

output -

column -

hierarchy -

Returns:

getOutput

protected DataHandle getOutput()

Returns output

Returns:

getRangeFromNumericColumn

protected double[] getRangeFromNumericColumn(double[] inputNumbers,
                                             DataHandle output,
                                             int column)

Tries to parse numbers from output when there is a numeric input column

Parameters:

inputNumbers -

output -

column -

Returns:

getSuppressedRecordsInInput

protected int getSuppressedRecordsInInput()

Returns the number of records suppressed in input

Returns:

getSuppressedRecordsInOutput

protected int getSuppressedRecordsInOutput()

Returns the number of records suppressed in output

Returns:

getSuppressionString

protected java.lang.String getSuppressionString()

Returns the suppression string

Returns:

isSuppressed

protected boolean isSuppressed(DataHandle handle,
                               int row,
                               int column)

Returns whether a value is suppressed

Parameters:

handle -

row -

column -

Returns:

isSuppressed

protected boolean isSuppressed(DataHandle handle,
                               int[] indices,
                               int row)

We assume that an entry is suppressed, if all values are equal

Parameters:

entry -

Returns:

isSuppressed

protected boolean isSuppressed(int column,
                               java.lang.String value)

Returns whether a value is suppressed

Parameters:

column -

value -

Returns:

log2

protected double log2(double d)

Log base-2

Parameters:

d -

Returns:

setStepPerformed

protected void setStepPerformed()

One step performed

setSteps

protected void setSteps(int steps)

Total number of steps

Parameters:

steps -

setStepsDone

protected void setStepsDone()

All steps performed