Package adams.flow.sink

Class WekaClassifierErrors.DataGenerator

java.lang.Object

adams.flow.sink.WekaClassifierErrors.DataGenerator

Enclosing class:

WekaClassifierErrors

public static class WekaClassifierErrors.DataGenerator
extends Object

Helper class for generating visualization data.

Author:

fracpete (fracpete at waikato dot ac dot nz)

Field Summary

Fields

Modifier and Type	Field	Description
protected AbstractErrorScaler	m_ErrorScaler	the scaler scheme to use.
protected weka.classifiers.Evaluation	m_Evaluation	the underlying Evaluation object.
protected weka.core.Instances	m_PlotInstances	the underlying data.
protected ArrayList	m_PlotShapes	for storing the plot shapes.
protected ArrayList	m_PlotSizes	for storing the plot sizes.
protected boolean	m_Processed	whether the data has already been processed.

Constructor Summary

Constructors

Constructor	Description
DataGenerator(weka.classifiers.Evaluation eval, AbstractErrorScaler scaler)	Initializes the generator.

Method Summary

Modifier and Type	Method	Description
protected void	createDataset(weka.classifiers.Evaluation eval)	Generates a dataset, containing the predicted vs actual values.
AbstractErrorScaler	getErrorScaler()	Returns the scaling scheme.
weka.classifiers.Evaluation	getEvaluation()	Returns the underlying Evaluation object.
weka.gui.visualize.PlotData2D	getPlotData()	Assembles and returns the plot.
weka.core.Instances	getPlotInstances()	Returns the generated dataset that is plotted.
protected void	process()	Processes the data if necessary.

Methods inherited from class java.lang.Object

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

Field Detail

m_Evaluation

protected weka.classifiers.Evaluation m_Evaluation

the underlying Evaluation object.

m_PlotInstances

protected weka.core.Instances m_PlotInstances

the underlying data.

m_PlotShapes

protected ArrayList m_PlotShapes

for storing the plot shapes.

m_PlotSizes

protected ArrayList m_PlotSizes

for storing the plot sizes.

m_ErrorScaler

protected AbstractErrorScaler m_ErrorScaler

the scaler scheme to use.

m_Processed

protected boolean m_Processed

whether the data has already been processed.

Constructor Detail

DataGenerator

public DataGenerator(weka.classifiers.Evaluation eval,
                     AbstractErrorScaler scaler)

Initializes the generator.

Parameters:

eval - the Evaluation object to use

scaler - the scaler scheme to use for the errors

Method Detail

process

protected void process()

Processes the data if necessary.

getEvaluation

public weka.classifiers.Evaluation getEvaluation()

Returns the underlying Evaluation object.

Returns:

the Evaluation object

getErrorScaler

public AbstractErrorScaler getErrorScaler()

Returns the scaling scheme.

Returns:

the scaler

getPlotInstances

public weka.core.Instances getPlotInstances()

Returns the generated dataset that is plotted.

Returns:

the dataset

createDataset

protected void createDataset(weka.classifiers.Evaluation eval)

Generates a dataset, containing the predicted vs actual values.

Parameters:

eval - for obtaining the dataset information and predictions

getPlotData

public weka.gui.visualize.PlotData2D getPlotData()
                                          throws Exception

Assembles and returns the plot. The relation name of the dataset gets added automatically.

Returns:

the plot

Throws:

Exception - if plot generation fails