Class ThresholdVisualizePanel

All Implemented Interfaces:
ImageObserver, MenuContainer, Serializable, Accessible, PrintableHandler

public class ThresholdVisualizePanel extends VisualizePanel
This panel is a VisualizePanel, with the added ablility to display the area under the ROC curve if an ROC curve is chosen.
Version:
$Revision: 10222 $
Author:
Dale Fletcher (dale@cs.waikato.ac.nz), FracPete (fracpete at waikato dot ac dot nz)
See Also:
  • Constructor Details

    • ThresholdVisualizePanel

      public ThresholdVisualizePanel()
      default constructor
  • Method Details

    • setROCString

      public void setROCString(String str)
      Set the string with ROC area
      Parameters:
      str - ROC area string to add to border
    • getROCString

      public String getROCString()
      This extracts the ROC area string
      Returns:
      ROC area string
    • setUpComboBoxes

      public void setUpComboBoxes(Instances inst)
      This overloads VisualizePanel's setUpComboBoxes to add ActionListeners to watch for when the X/Y Axis comboboxes are changed.
      Overrides:
      setUpComboBoxes in class VisualizePanel
      Parameters:
      inst - a set of instances with data for plotting
    • main

      public static void main(String[] args)
      Starts the ThresholdVisualizationPanel with parameters from the command line.

      Valid options are:

      -h
      lists all the commandline parameters

      -t file
      Dataset to process with given classifier.

      -W classname
      Full classname of classifier to run.
      Options after '--' are passed to the classifier.
      (default weka.classifiers.functions.Logistic)

      -r number
      The number of runs to perform (default 2).

      -x number
      The number of Cross-validation folds (default 10).

      -l file
      Previously saved threshold curve ARFF file.

      Parameters:
      args - optional commandline parameters