/*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
*/
/*
* ClassifierPanel.java
* Copyright (C) 1999 Len Trigg
*
*/
package weka.gui.explorer;
import weka.core.Instances;
import weka.core.Instance;
import weka.core.FastVector;
import weka.core.OptionHandler;
import weka.core.Attribute;
import weka.core.Utils;
import weka.core.Drawable;
import weka.core.SerializedObject;
import weka.classifiers.Classifier;
import weka.classifiers.DistributionClassifier;
import weka.classifiers.Evaluation;
import weka.classifiers.CostMatrix;
import weka.classifiers.evaluation.NominalPrediction;
import weka.classifiers.evaluation.MarginCurve;
import weka.classifiers.evaluation.ThresholdCurve;
import weka.classifiers.evaluation.CostCurve;
import weka.filters.Filter;
import weka.gui.Logger;
import weka.gui.TaskLogger;
import weka.gui.SysErrLog;
import weka.gui.GenericObjectEditor;
import weka.gui.PropertyPanel;
import weka.gui.ResultHistoryPanel;
import weka.gui.SetInstancesPanel;
import weka.gui.CostMatrixEditor;
import weka.gui.PropertyDialog;
import weka.gui.InstancesSummaryPanel;
import weka.gui.SaveBuffer;
import weka.gui.visualize.VisualizePanel;
import weka.gui.visualize.PlotData2D;
import weka.gui.visualize.Plot2D;
import weka.gui.ExtensionFileFilter;
import weka.gui.treevisualizer.*;
import java.util.Random;
import java.util.Date;
import java.text.SimpleDateFormat;
import java.awt.FlowLayout;
import java.awt.BorderLayout;
import java.awt.GridLayout;
import java.awt.GridBagLayout;
import java.awt.GridBagConstraints;
import java.awt.Insets;
import java.awt.Font;
import java.awt.Point;
import java.awt.event.ActionListener;
import java.awt.event.ActionEvent;
import java.awt.event.InputEvent;
import java.awt.event.MouseAdapter;
import java.awt.event.MouseEvent;
import java.awt.Window;
import java.awt.Dimension;
import java.beans.PropertyChangeListener;
import java.beans.PropertyChangeEvent;
import java.beans.PropertyChangeSupport;
import java.io.File;
import java.io.FileWriter;
import java.io.Writer;
import java.io.BufferedWriter;
import java.io.PrintWriter;
import java.io.OutputStream;
import java.io.ObjectOutputStream;
import java.io.FileOutputStream;
import java.util.zip.GZIPOutputStream;
import java.io.InputStream;
import java.io.ObjectInputStream;
import java.io.FileInputStream;
import java.util.zip.GZIPInputStream;
import javax.swing.JFileChooser;
import javax.swing.JPanel;
import javax.swing.JLabel;
import javax.swing.JButton;
import javax.swing.BorderFactory;
import javax.swing.JTextArea;
import javax.swing.JScrollPane;
import javax.swing.JRadioButton;
import javax.swing.ButtonGroup;
import javax.swing.JOptionPane;
import javax.swing.JComboBox;
import javax.swing.DefaultComboBoxModel;
import javax.swing.JTextField;
import javax.swing.SwingConstants;
import javax.swing.JFrame;
import javax.swing.event.ChangeListener;
import javax.swing.event.ChangeEvent;
import javax.swing.JViewport;
import javax.swing.JCheckBox;
import javax.swing.ListSelectionModel;
import javax.swing.event.ListSelectionEvent;
import javax.swing.event.ListSelectionListener;
import javax.swing.JPopupMenu;
import javax.swing.JMenu;
import javax.swing.JMenuItem;
import javax.swing.filechooser.FileFilter;
/**
* This panel allows the user to select and configure a classifier, set the
* attribute of the current dataset to be used as the class, and evaluate
* the classifier using a number of testing modes (test on the training data,
* train/test on a percentage split, n-fold cross-validation, test on a
* separate split). The results of classification runs are stored in a result
* history so that previous results are accessible.
*
* @author Len Trigg (trigg@cs.waikato.ac.nz)
* @author Mark Hall (mhall@cs.waikato.ac.nz)
* @author Richard Kirkby (rkirkby@cs.waikato.ac.nz)
* @version $Revision: 1.1.1.1 $
*/
public class ClassifierPanel extends JPanel {
/** Lets the user configure the classifier */
protected GenericObjectEditor m_ClassifierEditor =
new GenericObjectEditor();
/** The panel showing the current classifier selection */
protected PropertyPanel m_CEPanel = new PropertyPanel(m_ClassifierEditor);
/** The output area for classification results */
protected JTextArea m_OutText = new JTextArea(20, 40);
/** The destination for log/status messages */
protected Logger m_Log = new SysErrLog();
/** The buffer saving object for saving output */
SaveBuffer m_SaveOut = new SaveBuffer(m_Log, this);
/** A panel controlling results viewing */
protected ResultHistoryPanel m_History = new ResultHistoryPanel(m_OutText);
/** Lets the user select the class column */
protected JComboBox m_ClassCombo = new JComboBox();
/** Click to set test mode to cross-validation */
protected JRadioButton m_CVBut = new JRadioButton("Cross-validation");
/** Click to set test mode to generate a % split */
protected JRadioButton m_PercentBut = new JRadioButton("Percentage split");
/** Click to set test mode to test on training data */
protected JRadioButton m_TrainBut = new JRadioButton("Use training set");
/** Click to set test mode to a user-specified test set */
protected JRadioButton m_TestSplitBut =
new JRadioButton("Supplied test set");
/** Check to save the predictions in the results list for visualizing
later on */
protected JCheckBox m_StorePredictionsBut =
new JCheckBox("Store predictions for visualization");
/** Check to output the model built from the training data */
protected JCheckBox m_OutputModelBut = new JCheckBox("Output model");
/** Check to output true/false positives, precision/recall for each class */
protected JCheckBox m_OutputPerClassBut =
new JCheckBox("Output per-class stats");
/** Check to output a confusion matrix */
protected JCheckBox m_OutputConfusionBut =
new JCheckBox("Output confusion matrix");
/** Check to output entropy statistics */
protected JCheckBox m_OutputEntropyBut =
new JCheckBox("Output entropy evaluation measures");
/** Check to output text predictions */
protected JCheckBox m_OutputPredictionsTextBut =
new JCheckBox("Output text predictions on test set");
/** Check to evaluate w.r.t a cost matrix */
protected JCheckBox m_EvalWRTCostsBut =
new JCheckBox("Cost-sensitive evaluation");
protected JButton m_SetCostsBut = new JButton("Set...");
/** Label by where the cv folds are entered */
protected JLabel m_CVLab = new JLabel("Folds", SwingConstants.RIGHT);
/** The field where the cv folds are entered */
protected JTextField m_CVText = new JTextField("10");
/** Label by where the % split is entered */
protected JLabel m_PercentLab = new JLabel("%", SwingConstants.RIGHT);
/** The field where the % split is entered */
protected JTextField m_PercentText = new JTextField("66");
/** The button used to open a separate test dataset */
protected JButton m_SetTestBut = new JButton("Set...");
/** The frame used to show the test set selection panel */
protected JFrame m_SetTestFrame;
/** The frame used to show the cost matrix editing panel */
protected PropertyDialog m_SetCostsFrame;
/**
* Alters the enabled/disabled status of elements associated with each
* radio button
*/
ActionListener m_RadioListener = new ActionListener() {
public void actionPerformed(ActionEvent e) {
updateRadioLinks();
}
};
/** Button for further output/visualize options */
JButton m_MoreOptions = new JButton("More options...");
/**
* User specified random seed for cross validation or % split
*/
protected JTextField m_RandomSeedText = new JTextField("1 ");
protected JLabel m_RandomLab = new JLabel("Random seed for XVal / % Split",
SwingConstants.RIGHT);
/** Click to start running the classifier */
protected JButton m_StartBut = new JButton("Start");
/** Click to stop a running classifier */
protected JButton m_StopBut = new JButton("Stop");
/** Stop the class combo from taking up to much space */
private Dimension COMBO_SIZE = new Dimension(150, m_StartBut
.getPreferredSize().height);
/** The cost matrix editor for evaluation costs */
protected CostMatrixEditor m_CostMatrixEditor = new CostMatrixEditor();
/** The main set of instances we're playing with */
protected Instances m_Instances;
/** The user-supplied test set (if any) */
protected Instances m_TestInstances;
/** The user supplied test set after preprocess filters have been applied */
protected Instances m_TestInstancesCopy;
/** A thread that classification runs in */
protected Thread m_RunThread;
/** default x index for visualizing */
protected int m_visXIndex;
/** default y index for visualizing */
protected int m_visYIndex;
/** The current visualization object */
protected VisualizePanel m_CurrentVis = null;
/** The instances summary panel displayed by m_SetTestFrame */
protected InstancesSummaryPanel m_Summary = null;
/** Filter to ensure only model files are selected */
protected FileFilter m_ModelFilter =
new ExtensionFileFilter("model", "Model object files");
/** The file chooser for selecting model files */
protected JFileChooser m_FileChooser
= new JFileChooser(new File(System.getProperty("user.dir")));
/* Register the property editors we need */
static {
java.beans.PropertyEditorManager
.registerEditor(weka.core.SelectedTag.class,
weka.gui.SelectedTagEditor.class);
java.beans.PropertyEditorManager
.registerEditor(weka.filters.Filter.class,
weka.gui.GenericObjectEditor.class);
java.beans.PropertyEditorManager
.registerEditor(weka.classifiers.Classifier [].class,
weka.gui.GenericArrayEditor.class);
java.beans.PropertyEditorManager
.registerEditor(weka.classifiers.DistributionClassifier.class,
weka.gui.GenericObjectEditor.class);
java.beans.PropertyEditorManager
.registerEditor(weka.classifiers.Classifier.class,
weka.gui.GenericObjectEditor.class);
java.beans.PropertyEditorManager
.registerEditor(weka.classifiers.CostMatrix.class,
weka.gui.CostMatrixEditor.class);
}
/**
* Creates the classifier panel
*/
public ClassifierPanel() {
// Connect / configure the components
m_OutText.setEditable(false);
m_OutText.setFont(new Font("Monospaced", Font.PLAIN, 12));
m_OutText.setBorder(BorderFactory.createEmptyBorder(5, 5, 5, 5));
m_OutText.addMouseListener(new MouseAdapter() {
public void mouseClicked(MouseEvent e) {
if ((e.getModifiers() & InputEvent.BUTTON1_MASK)
!= InputEvent.BUTTON1_MASK) {
m_OutText.selectAll();
}
}
});
m_History.setBorder(BorderFactory.createTitledBorder("Result list (right-click for options)"));
m_ClassifierEditor.setClassType(Classifier.class);
m_ClassifierEditor.setValue(new weka.classifiers.rules.ZeroR());
m_ClassifierEditor.addPropertyChangeListener(new PropertyChangeListener() {
public void propertyChange(PropertyChangeEvent e) {
repaint();
}
});
m_ClassCombo.setToolTipText("Select the attribute to use as the class");
m_TrainBut.setToolTipText("Test on the same set that the classifier"
+ " is trained on");
m_CVBut.setToolTipText("Perform a n-fold cross-validation");
m_PercentBut.setToolTipText("Train on a percentage of the data and"
+ " test on the remainder");
m_TestSplitBut.setToolTipText("Test on a user-specified dataset");
m_StartBut.setToolTipText("Starts the classification");
m_StopBut.setToolTipText("Stops a running classification");
m_StorePredictionsBut.
setToolTipText("Store predictions in the result list for later "
+"visualization");
m_OutputModelBut
.setToolTipText("Output the model obtained from the full training set");
m_OutputPerClassBut.setToolTipText("Output precision/recall & true/false"
+ " positives for each class");
m_OutputConfusionBut
.setToolTipText("Output the matrix displaying class confusions");
m_OutputEntropyBut
.setToolTipText("Output entropy-based evaluation measures");
m_EvalWRTCostsBut
.setToolTipText("Evaluate errors with respect to a cost matrix");
m_OutputPredictionsTextBut
.setToolTipText("Include the predictions on the test set in the output buffer");
m_FileChooser.setFileFilter(m_ModelFilter);
m_FileChooser.setFileSelectionMode(JFileChooser.FILES_ONLY);
m_StorePredictionsBut.setSelected(true);
m_OutputModelBut.setSelected(true);
m_OutputPerClassBut.setSelected(true);
m_OutputConfusionBut.setSelected(true);
m_ClassCombo.setEnabled(false);
m_ClassCombo.setPreferredSize(COMBO_SIZE);
m_ClassCombo.setMaximumSize(COMBO_SIZE);
m_ClassCombo.setMinimumSize(COMBO_SIZE);
m_CVBut.setSelected(true);
updateRadioLinks();
ButtonGroup bg = new ButtonGroup();
bg.add(m_TrainBut);
bg.add(m_CVBut);
bg.add(m_PercentBut);
bg.add(m_TestSplitBut);
m_TrainBut.addActionListener(m_RadioListener);
m_CVBut.addActionListener(m_RadioListener);
m_PercentBut.addActionListener(m_RadioListener);
m_TestSplitBut.addActionListener(m_RadioListener);
m_SetTestBut.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent e) {
setTestSet();
}
});
m_EvalWRTCostsBut.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent e) {
m_SetCostsBut.setEnabled(m_EvalWRTCostsBut.isSelected());
if ((m_SetCostsFrame != null)
&& (!m_EvalWRTCostsBut.isSelected())) {
m_SetCostsFrame.setVisible(false);
}
}
});
m_CostMatrixEditor.setValue(new CostMatrix(1));
m_SetCostsBut.setEnabled(m_EvalWRTCostsBut.isSelected());
m_SetCostsBut.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent e) {
m_SetCostsBut.setEnabled(false);
if (m_SetCostsFrame == null) {
m_SetCostsFrame = new PropertyDialog(m_CostMatrixEditor, 100, 100);
// pd.setSize(250,150);
m_SetCostsFrame.addWindowListener(new java.awt.event.WindowAdapter() {
public void windowClosing(java.awt.event.WindowEvent p) {
m_SetCostsBut.setEnabled(m_EvalWRTCostsBut.isSelected());
if ((m_SetCostsFrame != null)
&& (!m_EvalWRTCostsBut.isSelected())) {
m_SetCostsFrame.setVisible(false);
}
}
});
}
m_SetCostsFrame.setVisible(true);
}
});
m_StartBut.setEnabled(false);
m_StopBut.setEnabled(false);
m_StartBut.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent e) {
startClassifier();
}
});
m_StopBut.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent e) {
stopClassifier();
}
});
m_ClassCombo.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent e) {
int selected = m_ClassCombo.getSelectedIndex();
if (selected != -1) {
boolean isNominal = m_Instances.attribute(selected).isNominal();
m_OutputPerClassBut.setEnabled(isNominal);
m_OutputConfusionBut.setEnabled(isNominal);
}
}
});
m_History.setHandleRightClicks(false);
// see if we can popup a menu for the selected result
m_History.getList().addMouseListener(new MouseAdapter() {
public void mouseClicked(MouseEvent e) {
if ((e.getModifiers() & InputEvent.BUTTON1_MASK)
== InputEvent.BUTTON1_MASK) {
} else {
int index = m_History.getList().locationToIndex(e.getPoint());
if (index != -1) {
String name = m_History.getNameAtIndex(index);
visualize(name, e.getX(), e.getY());
} else {
visualize(null, e.getX(), e.getY());
}
}
}
});
m_MoreOptions.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent e) {
m_MoreOptions.setEnabled(false);
JPanel moreOptionsPanel = new JPanel();
moreOptionsPanel.setBorder(BorderFactory.createEmptyBorder(0, 5, 5, 5));
moreOptionsPanel.setLayout(new GridLayout(8, 1));
moreOptionsPanel.add(m_OutputModelBut);
moreOptionsPanel.add(m_OutputPerClassBut);
moreOptionsPanel.add(m_OutputEntropyBut);
moreOptionsPanel.add(m_OutputConfusionBut);
moreOptionsPanel.add(m_StorePredictionsBut);
moreOptionsPanel.add(m_OutputPredictionsTextBut);
JPanel costMatrixOption = new JPanel();
costMatrixOption.setLayout(new BorderLayout());
costMatrixOption.add(m_EvalWRTCostsBut, BorderLayout.WEST);
costMatrixOption.add(m_SetCostsBut, BorderLayout.EAST);
moreOptionsPanel.add(costMatrixOption);
JPanel seedPanel = new JPanel();
seedPanel.setLayout(new BorderLayout());
seedPanel.add(m_RandomLab, BorderLayout.WEST);
seedPanel.add(m_RandomSeedText, BorderLayout.EAST);
moreOptionsPanel.add(seedPanel);
JPanel all = new JPanel();
all.setLayout(new BorderLayout());
JButton oK = new JButton("OK");
JPanel okP = new JPanel();
okP.setBorder(BorderFactory.createEmptyBorder(5, 5, 5, 5));
okP.setLayout(new GridLayout(1,1,5,5));
okP.add(oK);
all.add(moreOptionsPanel, BorderLayout.CENTER);
all.add(okP, BorderLayout.SOUTH);
final javax.swing.JFrame jf =
new javax.swing.JFrame("Classifier evaluation options");
jf.getContentPane().setLayout(new BorderLayout());
jf.getContentPane().add(all, BorderLayout.CENTER);
jf.addWindowListener(new java.awt.event.WindowAdapter() {
public void windowClosing(java.awt.event.WindowEvent w) {
jf.dispose();
m_MoreOptions.setEnabled(true);
}
});
oK.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent a) {
m_MoreOptions.setEnabled(true);
jf.dispose();
}
});
jf.pack();
jf.setLocation(m_MoreOptions.getLocationOnScreen());
jf.setVisible(true);
}
});
// Layout the GUI
JPanel p1 = new JPanel();
p1.setBorder(BorderFactory.createCompoundBorder(
BorderFactory.createTitledBorder("Classifier"),
BorderFactory.createEmptyBorder(0, 5, 5, 5)
));
p1.setLayout(new BorderLayout());
p1.add(m_CEPanel, BorderLayout.NORTH);
JPanel p2 = new JPanel();
GridBagLayout gbL = new GridBagLayout();
p2.setLayout(gbL);
p2.setBorder(BorderFactory.createCompoundBorder(
BorderFactory.createTitledBorder("Test options"),
BorderFactory.createEmptyBorder(0, 5, 5, 5)
));
GridBagConstraints gbC = new GridBagConstraints();
gbC.anchor = GridBagConstraints.WEST;
gbC.gridy = 0; gbC.gridx = 0;
gbL.setConstraints(m_TrainBut, gbC);
p2.add(m_TrainBut);
gbC = new GridBagConstraints();
gbC.anchor = GridBagConstraints.WEST;
gbC.gridy = 1; gbC.gridx = 0;
gbL.setConstraints(m_TestSplitBut, gbC);
p2.add(m_TestSplitBut);
gbC = new GridBagConstraints();
gbC.anchor = GridBagConstraints.EAST;
gbC.fill = GridBagConstraints.HORIZONTAL;
gbC.gridy = 1; gbC.gridx = 1; gbC.gridwidth = 2;
gbC.insets = new Insets(2, 10, 2, 0);
gbL.setConstraints(m_SetTestBut, gbC);
p2.add(m_SetTestBut);
gbC = new GridBagConstraints();
gbC.anchor = GridBagConstraints.WEST;
gbC.gridy = 2; gbC.gridx = 0;
gbL.setConstraints(m_CVBut, gbC);
p2.add(m_CVBut);
gbC = new GridBagConstraints();
gbC.anchor = GridBagConstraints.EAST;
gbC.fill = GridBagConstraints.HORIZONTAL;
gbC.gridy = 2; gbC.gridx = 1;
gbC.insets = new Insets(2, 10, 2, 10);
gbL.setConstraints(m_CVLab, gbC);
p2.add(m_CVLab);
gbC = new GridBagConstraints();
gbC.anchor = GridBagConstraints.EAST;
gbC.fill = GridBagConstraints.HORIZONTAL;
gbC.gridy = 2; gbC.gridx = 2; gbC.weightx = 100;
gbC.ipadx = 20;
gbL.setConstraints(m_CVText, gbC);
p2.add(m_CVText);
gbC = new GridBagConstraints();
gbC.anchor = GridBagConstraints.WEST;
gbC.gridy = 3; gbC.gridx = 0;
gbL.setConstraints(m_PercentBut, gbC);
p2.add(m_PercentBut);
gbC = new GridBagConstraints();
gbC.anchor = GridBagConstraints.EAST;
gbC.fill = GridBagConstraints.HORIZONTAL;
gbC.gridy = 3; gbC.gridx = 1;
gbC.insets = new Insets(2, 10, 2, 10);
gbL.setConstraints(m_PercentLab, gbC);
p2.add(m_PercentLab);
gbC = new GridBagConstraints();
gbC.anchor = GridBagConstraints.EAST;
gbC.fill = GridBagConstraints.HORIZONTAL;
gbC.gridy = 3; gbC.gridx = 2; gbC.weightx = 100;
gbC.ipadx = 20;
gbL.setConstraints(m_PercentText, gbC);
p2.add(m_PercentText);
gbC = new GridBagConstraints();
gbC.anchor = GridBagConstraints.WEST;
gbC.fill = GridBagConstraints.HORIZONTAL;
gbC.gridy = 4; gbC.gridx = 0; gbC.weightx = 100;
gbC.gridwidth = 3;
gbC.insets = new Insets(3, 0, 1, 0);
gbL.setConstraints(m_MoreOptions, gbC);
p2.add(m_MoreOptions);
JPanel buttons = new JPanel();
buttons.setLayout(new GridLayout(2, 2));
buttons.add(m_ClassCombo);
m_ClassCombo.setBorder(BorderFactory.createEmptyBorder(5, 5, 5, 5));
JPanel ssButs = new JPanel();
ssButs.setBorder(BorderFactory.createEmptyBorder(5, 5, 5, 5));
ssButs.setLayout(new GridLayout(1, 2, 5, 5));
ssButs.add(m_StartBut);
ssButs.add(m_StopBut);
buttons.add(ssButs);
JPanel p3 = new JPanel();
p3.setBorder(BorderFactory.createTitledBorder("Classifier output"));
p3.setLayout(new BorderLayout());
final JScrollPane js = new JScrollPane(m_OutText);
p3.add(js, BorderLayout.CENTER);
js.getViewport().addChangeListener(new ChangeListener() {
private int lastHeight;
public void stateChanged(ChangeEvent e) {
JViewport vp = (JViewport)e.getSource();
int h = vp.getViewSize().height;
if (h != lastHeight) { // i.e. an addition not just a user scrolling
lastHeight = h;
int x = h - vp.getExtentSize().height;
vp.setViewPosition(new Point(0, x));
}
}
});
JPanel mondo = new JPanel();
gbL = new GridBagLayout();
mondo.setLayout(gbL);
gbC = new GridBagConstraints();
// gbC.anchor = GridBagConstraints.WEST;
gbC.fill = GridBagConstraints.HORIZONTAL;
gbC.gridy = 0; gbC.gridx = 0;
gbL.setConstraints(p2, gbC);
mondo.add(p2);
gbC = new GridBagConstraints();
gbC.anchor = GridBagConstraints.NORTH;
gbC.fill = GridBagConstraints.HORIZONTAL;
gbC.gridy = 1; gbC.gridx = 0;
gbL.setConstraints(buttons, gbC);
mondo.add(buttons);
gbC = new GridBagConstraints();
//gbC.anchor = GridBagConstraints.NORTH;
gbC.fill = GridBagConstraints.BOTH;
gbC.gridy = 2; gbC.gridx = 0; gbC.weightx = 0;
gbL.setConstraints(m_History, gbC);
mondo.add(m_History);
gbC = new GridBagConstraints();
gbC.fill = GridBagConstraints.BOTH;
gbC.gridy = 0; gbC.gridx = 1;
gbC.gridheight = 3;
gbC.weightx = 100; gbC.weighty = 100;
gbL.setConstraints(p3, gbC);
mondo.add(p3);
setLayout(new BorderLayout());
add(p1, BorderLayout.NORTH);
add(mondo, BorderLayout.CENTER);
}
/**
* Updates the enabled status of the input fields and labels.
*/
protected void updateRadioLinks() {
m_SetTestBut.setEnabled(m_TestSplitBut.isSelected());
m_OutputPredictionsTextBut.setEnabled(m_TestSplitBut.isSelected());
if ((m_SetTestFrame != null) && (!m_TestSplitBut.isSelected())) {
m_SetTestFrame.setVisible(false);
}
m_CVText.setEnabled(m_CVBut.isSelected());
m_CVLab.setEnabled(m_CVBut.isSelected());
m_PercentText.setEnabled(m_PercentBut.isSelected());
m_PercentLab.setEnabled(m_PercentBut.isSelected());
}
/**
* Sets the Logger to receive informational messages
*
* @param newLog the Logger that will now get info messages
*/
public void setLog(Logger newLog) {
m_Log = newLog;
}
/**
* Set the default attributes to use on the x and y axis
* of a new visualization object.
* @param x the index of the attribute to use on the x axis
* @param y the index of the attribute to use on the y axis
*/
public void setXY_VisualizeIndexes(int x, int y) {
m_visXIndex = x;
m_visYIndex = y;
}
/**
* Tells the panel to use a new set of instances.
*
* @param inst a set of Instances
*/
public void setInstances(Instances inst) {
m_Instances = inst;
setXY_VisualizeIndexes(0,0); // reset the default x and y indexes
String [] attribNames = new String [m_Instances.numAttributes()];
for (int i = 0; i < attribNames.length; i++) {
String type = "";
switch (m_Instances.attribute(i).type()) {
case Attribute.NOMINAL:
type = "(Nom) ";
break;
case Attribute.NUMERIC:
type = "(Num) ";
break;
case Attribute.STRING:
type = "(Str) ";
break;
default:
type = "(???) ";
}
attribNames[i] = type + m_Instances.attribute(i).name();
}
m_ClassCombo.setModel(new DefaultComboBoxModel(attribNames));
if (attribNames.length > 0) {
m_ClassCombo.setSelectedIndex(attribNames.length - 1);
m_ClassCombo.setEnabled(true);
m_StartBut.setEnabled(m_RunThread == null);
m_StopBut.setEnabled(m_RunThread != null);
} else {
m_StartBut.setEnabled(false);
m_StopBut.setEnabled(false);
}
}
/**
* Sets the user test set. Information about the current test set
* is displayed in an InstanceSummaryPanel and the user is given the
* ability to load another set from a file or url.
*
*/
protected void setTestSet() {
if (m_SetTestFrame == null) {
final SetInstancesPanel sp = new SetInstancesPanel();
m_Summary = sp.getSummary();
if (m_TestInstancesCopy != null) {
sp.setInstances(m_TestInstancesCopy);
}
sp.addPropertyChangeListener(new PropertyChangeListener() {
public void propertyChange(PropertyChangeEvent e) {
m_TestInstances = sp.getInstances();
}
});
// Add propertychangelistener to update m_TestInstances whenever
// it changes in the settestframe
m_SetTestFrame = new JFrame("Test Instances");
m_SetTestFrame.getContentPane().setLayout(new BorderLayout());
m_SetTestFrame.getContentPane().add(sp, BorderLayout.CENTER);
m_SetTestFrame.pack();
}
m_SetTestFrame.setVisible(true);
}
/**
* Process a classifier's prediction for an instance and update a
* set of plotting instances and additional plotting info. plotInfo
* for nominal class datasets holds shape types (actual data points have
* automatic shape type assignment; classifer error data points have
* box shape type). For numeric class datasets, the actual data points
* are stored in plotInstances and plotInfo stores the error (which is
* later converted to shape size values)
* @param toPredict the actual data point
* @param classifier the classifier
* @param eval the evaluation object to use for evaluating the classifer on
* the instance to predict
* @param predictions a fastvector to add the prediction to
* @param plotInstances a set of plottable instances
* @param plotShape additional plotting information (shape)
* @param plotSize additional plotting information (size)
*/
private void processClassifierPrediction(Instance toPredict,
Classifier classifier,
Evaluation eval,
FastVector predictions,
Instances plotInstances,
FastVector plotShape,
FastVector plotSize) {
try {
double pred;
// classifier is a distribution classifer and class is nominal
if (predictions != null) {
Instance classMissing = (Instance)toPredict.copy();
classMissing.setDataset(toPredict.dataset());
classMissing.setClassMissing();
DistributionClassifier dc =
(DistributionClassifier)classifier;
double [] dist =
dc.distributionForInstance(classMissing);
pred = eval.evaluateModelOnce(dist, toPredict);
int actual = (int)toPredict.classValue();
predictions.addElement(new
NominalPrediction(actual, dist, toPredict.weight()));
} else {
pred = eval.evaluateModelOnce(classifier,
toPredict);
}
double [] values = new double[plotInstances.numAttributes()];
for (int i = 0; i < plotInstances.numAttributes(); i++) {
if (i < toPredict.classIndex()) {
values[i] = toPredict.value(i);
} else if (i == toPredict.classIndex()) {
values[i] = pred;
values[i+1] = toPredict.value(i);
/* // if the class value of the instances to predict is missing then
// set it to the predicted value
if (toPredict.isMissing(i)) {
values[i+1] = pred;
} */
i++;
} else {
values[i] = toPredict.value(i-1);
}
}
plotInstances.add(new Instance(1.0, values));
if (toPredict.classAttribute().isNominal()) {
if (toPredict.isMissing(toPredict.classIndex())) {
plotShape.addElement(new Integer(Plot2D.MISSING_SHAPE));
} else if (pred != toPredict.classValue()) {
// set to default error point shape
plotShape.addElement(new Integer(Plot2D.ERROR_SHAPE));
} else {
// otherwise set to constant (automatically assigned) point shape
plotShape.addElement(new Integer(Plot2D.CONST_AUTOMATIC_SHAPE));
}
plotSize.addElement(new Integer(Plot2D.DEFAULT_SHAPE_SIZE));
} else {
// store the error (to be converted to a point size later)
Double errd = null;
if (!toPredict.isMissing(toPredict.classIndex())) {
errd = new Double(pred - toPredict.classValue());
plotShape.addElement(new Integer(Plot2D.CONST_AUTOMATIC_SHAPE));
} else {
// missing shape if actual class not present
plotShape.addElement(new Integer(Plot2D.MISSING_SHAPE));
}
plotSize.addElement(errd);
}
} catch (Exception ex) {
ex.printStackTrace();
}
}
/**
* Post processes numeric class errors into shape sizes for plotting
* in the visualize panel
* @param plotSize a FastVector of numeric class errors
*/
private void postProcessPlotInfo(FastVector plotSize) {
int maxpSize = 20;
double maxErr = Double.NEGATIVE_INFINITY;
double minErr = Double.POSITIVE_INFINITY;
double err;
for (int i = 0; i < plotSize.size(); i++) {
Double errd = (Double)plotSize.elementAt(i);
if (errd != null) {
err = Math.abs(errd.doubleValue());
if (err < minErr) {
minErr = err;
}
if (err > maxErr) {
maxErr = err;
}
}
}
for (int i = 0; i < plotSize.size(); i++) {
Double errd = (Double)plotSize.elementAt(i);
if (errd != null) {
err = Math.abs(errd.doubleValue());
if (maxErr - minErr > 0) {
double temp = (((err - minErr) / (maxErr - minErr))
* maxpSize);
plotSize.setElementAt(new Integer((int)temp), i);
} else {
plotSize.setElementAt(new Integer(1), i);
}
} else {
plotSize.setElementAt(new Integer(1), i);
}
}
}
/**
* Sets up the structure for the visualizable instances. This dataset
* contains the original attributes plus the classifier's predictions
* for the class as an attribute called "predicted+WhateverTheClassIsCalled".
* @param trainInstancs the instances that the classifier is trained on
* @return a new set of instances containing one more attribute (predicted
* class) than the trainInstances
*/
private Instances setUpVisualizableInstances(Instances trainInstances) {
FastVector hv = new FastVector();
Attribute predictedClass;
Attribute classAt = trainInstances.attribute(trainInstances.classIndex());
if (classAt.isNominal()) {
FastVector attVals = new FastVector();
for (int i = 0; i < classAt.numValues(); i++) {
attVals.addElement(classAt.value(i));
}
predictedClass = new Attribute("predicted"+classAt.name(), attVals);
} else {
predictedClass = new Attribute("predicted"+classAt.name());
}
for (int i = 0; i < trainInstances.numAttributes(); i++) {
if (i == trainInstances.classIndex()) {
hv.addElement(predictedClass);
}
hv.addElement(trainInstances.attribute(i).copy());
}
return new Instances(trainInstances.relationName()+"_predicted", hv,
trainInstances.numInstances());
}
/**
* Starts running the currently configured classifier with the current
* settings. This is run in a separate thread, and will only start if
* there is no classifier already running. The classifier output is sent
* to the results history panel.
*/
protected void startClassifier() {
if (m_RunThread == null) {
synchronized (this) {
m_StartBut.setEnabled(false);
m_StopBut.setEnabled(true);
}
m_RunThread = new Thread() {
public void run() {
// Copy the current state of things
m_Log.statusMessage("Setting up...");
CostMatrix costMatrix = null;
Instances inst = new Instances(m_Instances);
Instances userTest = null;
// additional vis info (either shape type or point size)
FastVector plotShape = new FastVector();
FastVector plotSize = new FastVector();
Instances predInstances = null;
// will hold the prediction objects if the class is nominal
FastVector predictions = null;
// for timing
long trainTimeStart = 0, trainTimeElapsed = 0;
if (m_TestInstances != null) {
userTest = new Instances(m_TestInstancesCopy);
}
if (m_EvalWRTCostsBut.isSelected()) {
costMatrix = new CostMatrix((CostMatrix) m_CostMatrixEditor
.getValue());
}
boolean outputModel = m_OutputModelBut.isSelected();
boolean outputConfusion = m_OutputConfusionBut.isSelected();
boolean outputPerClass = m_OutputPerClassBut.isSelected();
boolean outputSummary = true;
boolean outputEntropy = m_OutputEntropyBut.isSelected();
boolean saveVis = m_StorePredictionsBut.isSelected();
boolean outputPredictionsText = m_OutputPredictionsTextBut.isSelected();
String grph = null;
int testMode = 0;
int numFolds = 10, percent = 66;
int classIndex = m_ClassCombo.getSelectedIndex();
Classifier classifier = (Classifier) m_ClassifierEditor.getValue();
StringBuffer outBuff = new StringBuffer();
String name = (new SimpleDateFormat("HH:mm:ss - "))
.format(new Date());
String cname = classifier.getClass().getName();
if (cname.startsWith("weka.classifiers.")) {
name += cname.substring("weka.classifiers.".length());
} else {
name += cname;
}
try {
if (m_CVBut.isSelected()) {
testMode = 1;
numFolds = Integer.parseInt(m_CVText.getText());
if (numFolds <= 1) {
throw new Exception("Number of folds must be greater than 1");
}
} else if (m_PercentBut.isSelected()) {
testMode = 2;
percent = Integer.parseInt(m_PercentText.getText());
if ((percent <= 0) || (percent >= 100)) {
throw new Exception("Percentage must be between 0 and 100");
}
} else if (m_TrainBut.isSelected()) {
testMode = 3;
} else if (m_TestSplitBut.isSelected()) {
testMode = 4;
// Check the test instance compatibility
if (userTest == null) {
throw new Exception("No user test set has been opened");
}
if (!inst.equalHeaders(userTest)) {
throw new Exception("Train and test set are not compatible");
}
userTest.setClassIndex(classIndex);
} else {
throw new Exception("Unknown test mode");
}
inst.setClassIndex(classIndex);
// set up the structure of the plottable instances for
// visualization
predInstances = setUpVisualizableInstances(inst);
predInstances.setClassIndex(inst.classIndex()+1);
if (inst.classAttribute().isNominal() &&
classifier instanceof DistributionClassifier) {
predictions = new FastVector();
}
// Output some header information
m_Log.logMessage("Started " + cname);
if (m_Log instanceof TaskLogger) {
((TaskLogger)m_Log).taskStarted();
}
outBuff.append("=== Run information ===\n\n");
outBuff.append("Scheme: " + cname);
if (classifier instanceof OptionHandler) {
String [] o = ((OptionHandler) classifier).getOptions();
outBuff.append(" " + Utils.joinOptions(o));
}
outBuff.append("\n");
outBuff.append("Relation: " + inst.relationName() + '\n');
outBuff.append("Instances: " + inst.numInstances() + '\n');
outBuff.append("Attributes: " + inst.numAttributes() + '\n');
if (inst.numAttributes() < 100) {
for (int i = 0; i < inst.numAttributes(); i++) {
outBuff.append(" " + inst.attribute(i).name()
+ '\n');
}
} else {
outBuff.append(" [list of attributes omitted]\n");
}
outBuff.append("Test mode: ");
switch (testMode) {
case 3: // Test on training
outBuff.append("evaluate on training data\n");
break;
case 1: // CV mode
outBuff.append("" + numFolds + "-fold cross-validation\n");
break;
case 2: // Percent split
outBuff.append("split " + percent
+ "% train, remainder test\n");
break;
case 4: // Test on user split
outBuff.append("user supplied test set: "
+ userTest.numInstances() + " instances\n");
break;
}
if (costMatrix != null) {
outBuff.append("Evaluation cost matrix:\n")
.append(costMatrix.toString()).append("\n");
}
outBuff.append("\n");
m_History.addResult(name, outBuff);
m_History.setSingle(name);
// Build the model and output it.
if (outputModel || (testMode == 3) || (testMode == 4)) {
m_Log.statusMessage("Building model on training data...");
trainTimeStart = System.currentTimeMillis();
classifier.buildClassifier(inst);
trainTimeElapsed = System.currentTimeMillis() - trainTimeStart;
}
if (outputModel) {
outBuff.append("=== Classifier model (full training set) ===\n\n");
outBuff.append(classifier.toString() + "\n");
outBuff.append("\nTime taken to build model: " +
Utils.doubleToString(trainTimeElapsed / 1000.0,2)
+ " seconds\n\n");
m_History.updateResult(name);
if (classifier instanceof Drawable) {
grph = null;
try {
grph = ((Drawable)classifier).graph();
} catch (Exception ex) {
}
}
}
Evaluation eval = null;
switch (testMode) {
case 3: // Test on training
m_Log.statusMessage("Evaluating on training data...");
eval = new Evaluation(inst, costMatrix);
for (int jj=0;jj<inst.numInstances();jj++) {
processClassifierPrediction(inst.instance(jj), classifier,
eval, predictions,
predInstances, plotShape,
plotSize);
if ((jj % 100) == 0) {
m_Log.statusMessage("Evaluating on training data. Processed "
+jj+" instances...");
}
}
outBuff.append("=== Evaluation on training set ===\n");
break;
case 1: // CV mode
m_Log.statusMessage("Randomizing instances...");
int rnd = 1;
try {
rnd = Integer.parseInt(m_RandomSeedText.getText().trim());
System.err.println("Using random seed "+rnd);
} catch (Exception ex) {
m_Log.logMessage("Trouble parsing random seed value");
rnd = 1;
}
inst.randomize(new Random(rnd));
if (inst.attribute(classIndex).isNominal()) {
m_Log.statusMessage("Stratifying instances...");
inst.stratify(numFolds);
}
eval = new Evaluation(inst, costMatrix);
// Make some splits and do a CV
for (int fold = 0; fold < numFolds; fold++) {
m_Log.statusMessage("Creating splits for fold "
+ (fold + 1) + "...");
Instances train = inst.trainCV(numFolds, fold);
Instances test = inst.testCV(numFolds, fold);
m_Log.statusMessage("Building model for fold "
+ (fold + 1) + "...");
classifier.buildClassifier(train);
m_Log.statusMessage("Evaluating model for fold "
+ (fold + 1) + "...");
for (int jj=0;jj<test.numInstances();jj++) {
processClassifierPrediction(test.instance(jj), classifier,
eval, predictions,
predInstances, plotShape,
plotSize);
}
}
if (inst.attribute(classIndex).isNominal()) {
outBuff.append("=== Stratified cross-validation ===\n");
} else {
outBuff.append("=== Cross-validation ===\n");
}
break;
case 2: // Percent split
m_Log.statusMessage("Randomizing instances...");
try {
rnd = Integer.parseInt(m_RandomSeedText.getText().trim());
} catch (Exception ex) {
m_Log.logMessage("Trouble parsing random seed value");
rnd = 1;
}
inst.randomize(new Random(rnd));
int trainSize = inst.numInstances() * percent / 100;
int testSize = inst.numInstances() - trainSize;
Instances train = new Instances(inst, 0, trainSize);
Instances test = new Instances(inst, trainSize, testSize);
m_Log.statusMessage("Building model on training split...");
classifier.buildClassifier(train);
eval = new Evaluation(train, costMatrix);
m_Log.statusMessage("Evaluating on test split...");
for (int jj=0;jj<test.numInstances();jj++) {
processClassifierPrediction(test.instance(jj), classifier,
eval, predictions,
predInstances, plotShape,
plotSize);
if ((jj % 100) == 0) {
m_Log.statusMessage("Evaluating on test split. Processed "
+jj+" instances...");
}
}
outBuff.append("=== Evaluation on test split ===\n");
break;
case 4: // Test on user split
m_Log.statusMessage("Evaluating on test data...");
eval = new Evaluation(inst, costMatrix);
if (outputPredictionsText) {
outBuff.append("=== Predictions on test set ===\n\n");
outBuff.append(" inst#, actual, predicted, error");
if (inst.classAttribute().isNominal()
&& classifier instanceof DistributionClassifier) {
outBuff.append(", probability distribution");
}
outBuff.append("\n");
}
for (int jj=0;jj<userTest.numInstances();jj++) {
processClassifierPrediction(userTest.instance(jj), classifier,
eval, predictions,
predInstances, plotShape,
plotSize);
if (outputPredictionsText) {
outBuff.append(predictionText(classifier, userTest.instance(jj), jj+1));
}
if ((jj % 100) == 0) {
m_Log.statusMessage("Evaluating on test data. Processed "
+jj+" instances...");
}
}
if (outputPredictionsText) {
outBuff.append("\n");
}
outBuff.append("=== Evaluation on test set ===\n");
break;
default:
throw new Exception("Test mode not implemented");
}
if (outputSummary) {
outBuff.append(eval.toSummaryString(outputEntropy) + "\n");
}
if (inst.attribute(classIndex).isNominal()) {
if (outputPerClass) {
outBuff.append(eval.toClassDetailsString() + "\n");
}
if (outputConfusion) {
outBuff.append(eval.toMatrixString() + "\n");
}
}
m_History.updateResult(name);
m_Log.logMessage("Finished " + cname);
m_Log.statusMessage("OK");
} catch (Exception ex) {
ex.printStackTrace();
m_Log.logMessage(ex.getMessage());
m_Log.statusMessage("See error log");
} finally {
try {
if (predInstances != null && predInstances.numInstances() > 0) {
if (predInstances.attribute(predInstances.classIndex())
.isNumeric()) {
postProcessPlotInfo(plotSize);
}
m_CurrentVis = new VisualizePanel();
m_CurrentVis.setName(name+" ("+inst.relationName()+")");
m_CurrentVis.setLog(m_Log);
PlotData2D tempd = new PlotData2D(predInstances);
tempd.setShapeSize(plotSize);
tempd.setShapeType(plotShape);
tempd.setPlotName(name+" ("+inst.relationName()+")");
tempd.addInstanceNumberAttribute();
m_CurrentVis.addPlot(tempd);
m_CurrentVis.setColourIndex(predInstances.classIndex()+1);
m_CurrentVis.setXIndex(m_visXIndex);
m_CurrentVis.setYIndex(m_visYIndex);
m_CurrentVis.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent e) {
if (m_CurrentVis.getInstances().
relationName().
compareTo(m_Instances.relationName()) == 0) {
setXY_VisualizeIndexes(m_CurrentVis.getXIndex(),
m_CurrentVis.getYIndex());
}
}
});
if (saveVis) {
FastVector vv = new FastVector();
if (outputModel) {
vv.addElement(classifier);
Instances trainHeader = new Instances(m_Instances, 0);
trainHeader.setClassIndex(classIndex);
vv.addElement(trainHeader);
}
vv.addElement(m_CurrentVis);
if (grph != null) {
vv.addElement(grph);
}
if (predictions != null) {
vv.addElement(predictions);
vv.addElement(inst.classAttribute());
}
m_History.addObject(name, vv);
} else if (outputModel) {
FastVector vv = new FastVector();
vv.addElement(classifier);
Instances trainHeader = new Instances(m_Instances, 0);
trainHeader.setClassIndex(classIndex);
vv.addElement(trainHeader);
m_History.addObject(name, vv);
}
}
} catch (Exception ex) {
ex.printStackTrace();
}
if (isInterrupted()) {
m_Log.logMessage("Interrupted " + cname);
m_Log.statusMessage("See error log");
}
synchronized (this) {
m_StartBut.setEnabled(true);
m_StopBut.setEnabled(false);
m_RunThread = null;
}
if (m_Log instanceof TaskLogger) {
((TaskLogger)m_Log).taskFinished();
}
}
}
};
m_RunThread.setPriority(Thread.MIN_PRIORITY);
m_RunThread.start();
}
}
protected String predictionText(Classifier classifier, Instance inst, int instNum) throws Exception {
//> inst# actual predicted error probability distribution
StringBuffer text = new StringBuffer();
// inst #
text.append(Utils.padLeft("" + instNum, 6) + " ");
if (inst.classAttribute().isNominal()) {
// actual
if (inst.classIsMissing()) text.append(Utils.padLeft("?", 10) + " ");
else text.append(Utils.padLeft("" + ((int) inst.classValue()+1) + ":"
+ inst.stringValue(inst.classAttribute()), 10) + " ");
// predicted
double[] probdist = null;
double pred;
if (classifier instanceof DistributionClassifier) {
probdist = ((DistributionClassifier)classifier).distributionForInstance(inst);
pred = (double) Utils.maxIndex(probdist);
if (probdist[(int) pred] <= 0.0) pred = Instance.missingValue();
} else {
pred = classifier.classifyInstance(inst);
}
text.append(Utils.padLeft((Instance.isMissingValue(pred) ? "?" :
(((int) pred+1) + ":"
+ inst.classAttribute().value((int) pred))), 10) + " ");
// error
if (pred == inst.classValue()) text.append(Utils.padLeft(" ", 6) + " ");
else text.append(Utils.padLeft("+", 6) + " ");
// prob dist
if (classifier instanceof DistributionClassifier) {
for (int i=0; i<probdist.length; i++) {
if (i == (int) pred) text.append(" *");
else text.append(" ");
text.append(Utils.doubleToString(probdist[i], 5, 3));
}
}
} else {
// actual
if (inst.classIsMissing()) text.append(Utils.padLeft("?", 10) + " ");
else text.append(Utils.doubleToString(inst.classValue(), 10, 3) + " ");
// predicted
double pred = classifier.classifyInstance(inst);
if (Instance.isMissingValue(pred)) text.append(Utils.padLeft("?", 10) + " ");
else text.append(Utils.doubleToString(pred, 10, 3) + " ");
// err
if (!inst.classIsMissing() && !Instance.isMissingValue(pred))
text.append(Utils.doubleToString(pred - inst.classValue(), 10, 3));
}
text.append("\n");
return text.toString();
}
/**
* Handles constructing a popup menu with visualization options.
* @param name the name of the result history list entry clicked on by
* the user
* @param x the x coordinate for popping up the menu
* @param y the y coordinate for popping up the menu
*/
protected void visualize(String name, int x, int y) {
final String selectedName = name;
JPopupMenu resultListMenu = new JPopupMenu();
JMenuItem visMainBuffer = new JMenuItem("View in main window");
if (selectedName != null) {
visMainBuffer.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent e) {
m_History.setSingle(selectedName);
}
});
} else {
visMainBuffer.setEnabled(false);
}
resultListMenu.add(visMainBuffer);
JMenuItem visSepBuffer = new JMenuItem("View in separate window");
if (selectedName != null) {
visSepBuffer.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent e) {
m_History.openFrame(selectedName);
}
});
} else {
visSepBuffer.setEnabled(false);
}
resultListMenu.add(visSepBuffer);
JMenuItem saveOutput = new JMenuItem("Save result buffer");
if (selectedName != null) {
saveOutput.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent e) {
saveBuffer(selectedName);
}
});
} else {
saveOutput.setEnabled(false);
}
resultListMenu.add(saveOutput);
resultListMenu.addSeparator();
JMenuItem loadModel = new JMenuItem("Load model");
loadModel.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent e) {
loadClassifier();
}
});
resultListMenu.add(loadModel);
FastVector o = null;
if (selectedName != null) {
o = (FastVector)m_History.getNamedObject(selectedName);
}
VisualizePanel temp_vp = null;
String temp_grph = null;
FastVector temp_preds = null;
Attribute temp_classAtt = null;
Classifier temp_classifier = null;
Instances temp_trainHeader = null;
if (o != null) {
for (int i = 0; i < o.size(); i++) {
Object temp = o.elementAt(i);
if (temp instanceof Classifier) {
temp_classifier = (Classifier)temp;
} else if (temp instanceof Instances) { // training header
temp_trainHeader = (Instances)temp;
} else if (temp instanceof VisualizePanel) { // normal errors
temp_vp = (VisualizePanel)temp;
} else if (temp instanceof String) { // graphable output
temp_grph = (String)temp;
} else if (temp instanceof FastVector) { // predictions
temp_preds = (FastVector)temp;
} else if (temp instanceof Attribute) { // class attribute
temp_classAtt = (Attribute)temp;
}
}
}
final VisualizePanel vp = temp_vp;
final String grph = temp_grph;
final FastVector preds = temp_preds;
final Attribute classAtt = temp_classAtt;
final Classifier classifier = temp_classifier;
final Instances trainHeader = temp_trainHeader;
JMenuItem saveModel = new JMenuItem("Save model");
if (classifier != null) {
saveModel.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent e) {
saveClassifier(selectedName, classifier, trainHeader);
}
});
} else {
saveModel.setEnabled(false);
}
resultListMenu.add(saveModel);
JMenuItem reEvaluate =
new JMenuItem("Re-evaluate model on current test set");
if (classifier != null && m_TestInstances != null) {
reEvaluate.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent e) {
reevaluateModel(selectedName, classifier, trainHeader);
}
});
} else {
reEvaluate.setEnabled(false);
}
resultListMenu.add(reEvaluate);
resultListMenu.addSeparator();
JMenuItem visErrors = new JMenuItem("Visualize classifer errors");
if (vp != null) {
visErrors.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent e) {
visualizeClassifierErrors(vp);
}
});
} else {
visErrors.setEnabled(false);
}
resultListMenu.add(visErrors);
JMenuItem visTree = new JMenuItem("Visualize tree");
if (grph != null) {
visTree.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent e) {
String title;
if (vp != null) title = vp.getName();
else title = selectedName;
visualizeTree(grph, title);
}
});
} else {
visTree.setEnabled(false);
}
resultListMenu.add(visTree);
JMenuItem visMargin = new JMenuItem("Visualize margin curve");
if (preds != null) {
visMargin.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent e) {
try {
MarginCurve tc = new MarginCurve();
Instances result = tc.getCurve(preds);
VisualizePanel vmc = new VisualizePanel();
vmc.setName(result.relationName());
vmc.setLog(m_Log);
PlotData2D tempd = new PlotData2D(result);
tempd.setPlotName(result.relationName());
tempd.addInstanceNumberAttribute();
vmc.addPlot(tempd);
visualizeClassifierErrors(vmc);
} catch (Exception ex) {
ex.printStackTrace();
}
}
});
} else {
visMargin.setEnabled(false);
}
resultListMenu.add(visMargin);
JMenu visThreshold = new JMenu("Visualize threshold curve");
if (preds != null && classAtt != null) {
for (int i = 0; i < classAtt.numValues(); i++) {
JMenuItem clv = new JMenuItem(classAtt.value(i));
final int classValue = i;
clv.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent e) {
try {
ThresholdCurve tc = new ThresholdCurve();
Instances result = tc.getCurve(preds, classValue);
VisualizePanel vmc = new VisualizePanel();
vmc.setLog(m_Log);
vmc.setName(result.relationName()+". Class value "+
classAtt.value(classValue)+")");
PlotData2D tempd = new PlotData2D(result);
tempd.setPlotName(result.relationName());
tempd.addInstanceNumberAttribute();
vmc.addPlot(tempd);
visualizeClassifierErrors(vmc);
} catch (Exception ex) {
ex.printStackTrace();
}
}
});
visThreshold.add(clv);
}
} else {
visThreshold.setEnabled(false);
}
resultListMenu.add(visThreshold);
JMenu visCost = new JMenu("Visualize cost curve");
if (preds != null && classAtt != null) {
for (int i = 0; i < classAtt.numValues(); i++) {
JMenuItem clv = new JMenuItem(classAtt.value(i));
final int classValue = i;
clv.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent e) {
try {
CostCurve cc = new CostCurve();
Instances result = cc.getCurve(preds, classValue);
VisualizePanel vmc = new VisualizePanel();
vmc.setLog(m_Log);
vmc.setName(result.relationName()+". Class value "+
classAtt.value(classValue)+")");
PlotData2D tempd = new PlotData2D(result);
tempd.m_displayAllPoints = true;
tempd.setPlotName(result.relationName());
boolean [] connectPoints =
new boolean [result.numInstances()];
for (int jj = 1; jj < connectPoints.length; jj+=2) {
connectPoints[jj] = true;
}
tempd.setConnectPoints(connectPoints);
// tempd.addInstanceNumberAttribute();
vmc.addPlot(tempd);
visualizeClassifierErrors(vmc);
} catch (Exception ex) {
ex.printStackTrace();
}
}
});
visCost.add(clv);
}
} else {
visCost.setEnabled(false);
}
resultListMenu.add(visCost);
resultListMenu.show(m_History.getList(), x, y);
}
/**
* Pops up a TreeVisualizer for the classifier from the currently
* selected item in the results list
* @param dottyString the description of the tree in dotty format
* @param treeName the title to assign to the display
*/
protected void visualizeTree(String dottyString, String treeName) {
final javax.swing.JFrame jf =
new javax.swing.JFrame("Weka Classifier Tree Visualizer: "+treeName);
jf.setSize(500,400);
jf.getContentPane().setLayout(new BorderLayout());
TreeVisualizer tv = new TreeVisualizer(null,
dottyString,
new PlaceNode2());
jf.getContentPane().add(tv, BorderLayout.CENTER);
jf.addWindowListener(new java.awt.event.WindowAdapter() {
public void windowClosing(java.awt.event.WindowEvent e) {
jf.dispose();
}
});
jf.setVisible(true);
tv.fitToScreen();
}
/**
* Pops up a VisualizePanel for visualizing the data and errors for
* the classifier from the currently selected item in the results list
* @param sp the VisualizePanel to pop up.
*/
protected void visualizeClassifierErrors(VisualizePanel sp) {
if (sp != null) {
String plotName = sp.getName();
final javax.swing.JFrame jf =
new javax.swing.JFrame("Weka Classifier Visualize: "+plotName);
jf.setSize(500,400);
jf.getContentPane().setLayout(new BorderLayout());
jf.getContentPane().add(sp, BorderLayout.CENTER);
jf.addWindowListener(new java.awt.event.WindowAdapter() {
public void windowClosing(java.awt.event.WindowEvent e) {
jf.dispose();
}
});
jf.setVisible(true);
}
}
/**
* Save the currently selected classifier output to a file.
* @param name the name of the buffer to save
*/
protected void saveBuffer(String name) {
StringBuffer sb = m_History.getNamedBuffer(name);
if (sb != null) {
if (m_SaveOut.save(sb)) {
m_Log.logMessage("Save successful.");
}
}
}
/**
* Stops the currently running classifier (if any).
*/
protected void stopClassifier() {
if (m_RunThread != null) {
m_RunThread.interrupt();
// This is deprecated (and theoretically the interrupt should do).
m_RunThread.stop();
}
}
/**
* Saves the currently selected classifier
*/
protected void saveClassifier(String name, Classifier classifier,
Instances trainHeader) {
File sFile = null;
boolean saveOK = true;
int returnVal = m_FileChooser.showSaveDialog(this);
if (returnVal == JFileChooser.APPROVE_OPTION) {
sFile = m_FileChooser.getSelectedFile();
m_Log.statusMessage("Saving model to file...");
try {
OutputStream os = new FileOutputStream(sFile);
if (sFile.getName().endsWith(".gz")) {
os = new GZIPOutputStream(os);
}
ObjectOutputStream objectOutputStream = new ObjectOutputStream(os);
objectOutputStream.writeObject(classifier);
if (trainHeader != null) objectOutputStream.writeObject(trainHeader);
objectOutputStream.flush();
objectOutputStream.close();
} catch (Exception e) {
JOptionPane.showMessageDialog(null, e, "Save Failed",
JOptionPane.ERROR_MESSAGE);
saveOK = false;
}
if (saveOK)
m_Log.logMessage("Saved model (" + name
+ ") to file '" + sFile.getName() + "'");
m_Log.statusMessage("OK");
}
}
/**
* Loads a classifier
*/
protected void loadClassifier() {
int returnVal = m_FileChooser.showOpenDialog(this);
if (returnVal == JFileChooser.APPROVE_OPTION) {
File selected = m_FileChooser.getSelectedFile();
Classifier classifier = null;
Instances trainHeader = null;
m_Log.statusMessage("Loading model from file...");
try {
InputStream is = new FileInputStream(selected);
if (selected.getName().endsWith(".gz")) {
is = new GZIPInputStream(is);
}
ObjectInputStream objectInputStream = new ObjectInputStream(is);
classifier = (Classifier) objectInputStream.readObject();
try { // see if we can load the header
trainHeader = (Instances) objectInputStream.readObject();
} catch (Exception e) {} // don't fuss if we can't
objectInputStream.close();
} catch (Exception e) {
JOptionPane.showMessageDialog(null, e, "Load Failed",
JOptionPane.ERROR_MESSAGE);
}
m_Log.statusMessage("OK");
if (classifier != null) {
m_Log.logMessage("Loaded model from file '" + selected.getName()+ "'");
String name = (new SimpleDateFormat("HH:mm:ss - ")).format(new Date());
String cname = classifier.getClass().getName();
if (cname.startsWith("weka.classifiers."))
cname = cname.substring("weka.classifiers.".length());
name += cname + " from file '" + selected.getName() + "'";
StringBuffer outBuff = new StringBuffer();
outBuff.append("=== Model information ===\n\n");
outBuff.append("Filename: " + selected.getName() + "\n");
outBuff.append("Scheme: " + cname);
if (classifier instanceof OptionHandler) {
String [] o = ((OptionHandler) classifier).getOptions();
outBuff.append(" " + Utils.joinOptions(o));
}
outBuff.append("\n");
if (trainHeader != null) {
outBuff.append("Relation: " + trainHeader.relationName() + '\n');
outBuff.append("Attributes: " + trainHeader.numAttributes() + '\n');
if (trainHeader.numAttributes() < 100) {
for (int i = 0; i < trainHeader.numAttributes(); i++) {
outBuff.append(" " + trainHeader.attribute(i).name()
+ '\n');
}
} else {
outBuff.append(" [list of attributes omitted]\n");
}
} else {
outBuff.append("\nTraining data unknown\n");
}
outBuff.append("\n=== Classifier model ===\n\n");
outBuff.append(classifier.toString() + "\n");
m_History.addResult(name, outBuff);
m_History.setSingle(name);
FastVector vv = new FastVector();
vv.addElement(classifier);
if (trainHeader != null) vv.addElement(trainHeader);
// allow visualization of graphable classifiers
String grph = null;
if (classifier instanceof Drawable) {
try {
grph = ((Drawable)classifier).graph();
} catch (Exception ex) {
}
}
if (grph != null) vv.addElement(grph);
m_History.addObject(name, vv);
}
}
}
/**
* Re-evaluates the named classifier with the current test set. Unpredictable
* things will happen if the data set is not compatible with the classifier.
*
* @param name the name of the classifier entry
* @param classifier the classifier to evaluate
*/
protected void reevaluateModel(String name, Classifier classifier, Instances trainHeader) {
StringBuffer outBuff = m_History.getNamedBuffer(name);
Instances userTest = null;
// additional vis info (either shape type or point size)
FastVector plotShape = new FastVector();
FastVector plotSize = new FastVector();
Instances predInstances = null;
// will hold the prediction objects if the class is nominal
FastVector predictions = null;
CostMatrix costMatrix = null;
if (m_EvalWRTCostsBut.isSelected()) {
costMatrix = new CostMatrix((CostMatrix) m_CostMatrixEditor
.getValue());
}
boolean outputConfusion = m_OutputConfusionBut.isSelected();
boolean outputPerClass = m_OutputPerClassBut.isSelected();
boolean outputSummary = true;
boolean outputEntropy = m_OutputEntropyBut.isSelected();
boolean saveVis = m_StorePredictionsBut.isSelected();
String grph = null;
try {
if (m_TestInstances != null) {
userTest = new Instances(m_TestInstancesCopy);
}
// Check the test instance compatibility
if (userTest == null) {
throw new Exception("No user test set has been opened");
}
if (trainHeader != null) {
if (trainHeader.classIndex() > userTest.numAttributes()-1)
throw new Exception("Train and test set are not compatible");
userTest.setClassIndex(trainHeader.classIndex());
if (!trainHeader.equalHeaders(userTest)) {
throw new Exception("Train and test set are not compatible");
}
} else {
userTest.setClassIndex(userTest.numAttributes()-1);
}
m_Log.statusMessage("Evaluating on test data...");
m_Log.logMessage("Re-evaluating classifier (" + name + ") on test set");
Evaluation eval = new Evaluation(userTest, costMatrix);
// set up the structure of the plottable instances for
// visualization
predInstances = setUpVisualizableInstances(userTest);
predInstances.setClassIndex(userTest.classIndex()+1);
if (userTest.classAttribute().isNominal() &&
classifier instanceof DistributionClassifier) {
predictions = new FastVector();
}
for (int jj=0;jj<userTest.numInstances();jj++) {
processClassifierPrediction(userTest.instance(jj), classifier,
eval, predictions,
predInstances, plotShape,
plotSize);
if ((jj % 100) == 0) {
m_Log.statusMessage("Evaluating on test data. Processed "
+jj+" instances...");
}
}
outBuff.append("\n=== Re-evaluation on test set ===\n\n");
outBuff.append("User supplied test set\n");
outBuff.append("Relation: " + userTest.relationName() + '\n');
outBuff.append("Instances: " + userTest.numInstances() + '\n');
outBuff.append("Attributes: " + userTest.numAttributes() + "\n\n");
if (trainHeader == null)
outBuff.append("NOTE - if test set is not compatible then results are "
+ "unpredictable\n\n");
if (outputSummary) {
outBuff.append(eval.toSummaryString(outputEntropy) + "\n");
}
if (userTest.classAttribute().isNominal()) {
if (outputPerClass) {
outBuff.append(eval.toClassDetailsString() + "\n");
}
if (outputConfusion) {
outBuff.append(eval.toMatrixString() + "\n");
}
}
m_History.updateResult(name);
m_Log.logMessage("Finished re-evaluation");
m_Log.statusMessage("OK");
} catch (Exception ex) {
ex.printStackTrace();
m_Log.logMessage(ex.getMessage());
m_Log.statusMessage("See error log");
} finally {
try {
if (predInstances != null && predInstances.numInstances() > 0) {
if (predInstances.attribute(predInstances.classIndex())
.isNumeric()) {
postProcessPlotInfo(plotSize);
}
m_CurrentVis = new VisualizePanel();
m_CurrentVis.setName(name+" ("+userTest.relationName()+")");
m_CurrentVis.setLog(m_Log);
PlotData2D tempd = new PlotData2D(predInstances);
tempd.setShapeSize(plotSize);
tempd.setShapeType(plotShape);
tempd.setPlotName(name+" ("+userTest.relationName()+")");
tempd.addInstanceNumberAttribute();
m_CurrentVis.addPlot(tempd);
m_CurrentVis.setColourIndex(predInstances.classIndex()+1);
m_CurrentVis.setXIndex(m_visXIndex);
m_CurrentVis.setYIndex(m_visYIndex);
m_CurrentVis.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent e) {
if (m_Instances != null &&
m_CurrentVis.getInstances().
relationName().
compareTo(m_Instances.relationName()) == 0) {
setXY_VisualizeIndexes(m_CurrentVis.getXIndex(),
m_CurrentVis.getYIndex());
}
}
});
if (classifier instanceof Drawable) {
try {
grph = ((Drawable)classifier).graph();
} catch (Exception ex) {
}
}
if (saveVis) {
FastVector vv = new FastVector();
vv.addElement(classifier);
if (trainHeader != null) vv.addElement(trainHeader);
vv.addElement(m_CurrentVis);
if (grph != null) {
vv.addElement(grph);
}
if (predictions != null) {
vv.addElement(predictions);
vv.addElement(userTest.classAttribute());
}
m_History.addObject(name, vv);
} else {
FastVector vv = new FastVector();
vv.addElement(classifier);
if (trainHeader != null) vv.addElement(trainHeader);
m_History.addObject(name, vv);
}
}
} catch (Exception ex) {
ex.printStackTrace();
}
}
}
/**
* Tests out the classifier panel from the command line.
*
* @param args may optionally contain the name of a dataset to load.
*/
public static void main(String [] args) {
try {
final javax.swing.JFrame jf =
new javax.swing.JFrame("Weka Knowledge Explorer: Classifier");
jf.getContentPane().setLayout(new BorderLayout());
final ClassifierPanel sp = new ClassifierPanel();
jf.getContentPane().add(sp, BorderLayout.CENTER);
weka.gui.LogPanel lp = new weka.gui.LogPanel();
sp.setLog(lp);
jf.getContentPane().add(lp, BorderLayout.SOUTH);
jf.addWindowListener(new java.awt.event.WindowAdapter() {
public void windowClosing(java.awt.event.WindowEvent e) {
jf.dispose();
System.exit(0);
}
});
jf.pack();
jf.setSize(800, 600);
jf.setVisible(true);
if (args.length == 1) {
System.err.println("Loading instances from " + args[0]);
java.io.Reader r = new java.io.BufferedReader(
new java.io.FileReader(args[0]));
Instances i = new Instances(r);
sp.setInstances(i);
}
} catch (Exception ex) {
ex.printStackTrace();
System.err.println(ex.getMessage());
}
}
}