Package weka.gui.explorer
Class ClassifierErrorsPlotInstances
java.lang.Object
weka.gui.explorer.AbstractPlotInstances
weka.gui.explorer.ClassifierErrorsPlotInstances
- All Implemented Interfaces:
Serializable
,OptionHandler
A class for generating plottable visualization errors.
Example usage:
Instances train = ... // from somewhere Instances test = ... // from somewhere Classifier cls = ... // from somewhere // build classifier cls.buildClassifier(train); // evaluate classifier and generate plot instances ClassifierPlotInstances plotInstances = new ClassifierPlotInstances(); plotInstances.setClassifier(cls); plotInstances.setInstances(train); plotInstances.setClassIndex(train.classIndex()); plotInstances.setUp(); Evaluation eval = new Evaluation(train); for (int i = 0; i < test.numInstances(); i++) plotInstances.process(test.instance(i), cls, eval); // generate visualization VisualizePanel visPanel = new VisualizePanel(); visPanel.addPlot(plotInstances.getPlotData("plot name")); visPanel.setColourIndex(plotInstances.getPlotInstances().classIndex()+1); // clean up plotInstances.cleanUp();
- Version:
- $Revision: 10220 $
- Author:
- fracpete (fracpete at waikato dot ac dot nz)
- See Also:
-
Constructor Summary
-
Method Summary
Modifier and TypeMethodDescriptionvoid
cleanUp()
For freeing up memory.Returns the currently set classifier.int
Returns the 0-based class index.Returns the Evaluation object in use.Get the vector of plot shapes (see weka.gui.visualize.Plot2D).Get the vector of plot sizes (see weka.gui.visualize.Plot2D).boolean
Get whether the point size should be proportional to the prediction margin (classification only).boolean
Returns whether the instances are saved for visualization for only evaluation of the prediction is to happen.void
process
(Instance toPredict, Classifier classifier, Evaluation eval) Process a classifier's prediction for an instance and update a set of plotting instances and additional plotting info.void
process
(Instances batch, double[][] predictions, Evaluation eval) void
setClassifier
(Classifier value) Sets the classifier used for making the predictions.void
setClassIndex
(int index) Sets the 0-based class index.void
setEvaluation
(Evaluation value) Sets the Evaluation object to use.void
setPlotShapes
(ArrayList<Integer> plotShapes) Set the vector of plot shapes to use;void
setPlotSizes
(ArrayList<Object> plotSizes) Set the vector of plot sizes to usevoid
setPointSizeProportionalToMargin
(boolean b) Set whether the point size should be proportional to the prediction margin (classification only).void
setSaveForVisualization
(boolean value) Sets whether the instances are saved for visualization or only evaluation of the prediction is to happen.Methods inherited from class weka.gui.explorer.AbstractPlotInstances
canPlot, getInstances, getOptions, getPlotData, getPlotInstances, listOptions, setInstances, setOptions, setUp
-
Constructor Details
-
ClassifierErrorsPlotInstances
public ClassifierErrorsPlotInstances()
-
-
Method Details
-
getPlotShapes
Get the vector of plot shapes (see weka.gui.visualize.Plot2D).- Returns:
- the vector of plot shapes.
-
getPlotSizes
Get the vector of plot sizes (see weka.gui.visualize.Plot2D).- Returns:
- the vector of plot sizes.
-
setPlotShapes
Set the vector of plot shapes to use;- Parameters:
plotShapes
-
-
setPlotSizes
Set the vector of plot sizes to use- Parameters:
plotSizes
- the plot sizes to use
-
setClassifier
Sets the classifier used for making the predictions.- Parameters:
value
- the classifier to use
-
getClassifier
Returns the currently set classifier.- Returns:
- the classifier in use
-
setClassIndex
public void setClassIndex(int index) Sets the 0-based class index.- Parameters:
index
- the class index
-
getClassIndex
public int getClassIndex()Returns the 0-based class index.- Returns:
- the class index
-
setEvaluation
Sets the Evaluation object to use.- Parameters:
value
- the evaluation to use
-
getEvaluation
Returns the Evaluation object in use.- Returns:
- the evaluation object
-
setSaveForVisualization
public void setSaveForVisualization(boolean value) Sets whether the instances are saved for visualization or only evaluation of the prediction is to happen.- Parameters:
value
- if true then the instances will be saved
-
getSaveForVisualization
public boolean getSaveForVisualization()Returns whether the instances are saved for visualization for only evaluation of the prediction is to happen.- Returns:
- true if the instances are saved
-
setPointSizeProportionalToMargin
public void setPointSizeProportionalToMargin(boolean b) Set whether the point size should be proportional to the prediction margin (classification only).- Parameters:
b
- true if the point size should be proportional to the margin
-
getPointSizeProportionalToMargin
public boolean getPointSizeProportionalToMargin()Get whether the point size should be proportional to the prediction margin (classification only).- Returns:
- true if the point size should be proportional to the margin
-
process
-
process
Process a classifier's prediction for an instance and update a set of plotting instances and additional plotting info. m_PlotShape for nominal class datasets holds shape types (actual data points have automatic shape type assignment; classifier error data points have box shape type). For numeric class datasets, the actual data points are stored in m_PlotInstances and m_PlotSize stores the error (which is later converted to shape size values).- Parameters:
toPredict
- the actual data pointclassifier
- the classifiereval
- the evaluation object to use for evaluating the classifier on the instance to predict- See Also:
-
m_PlotShapes
m_PlotSizes
AbstractPlotInstances.m_PlotInstances
-
cleanUp
public void cleanUp()For freeing up memory. Plot data cannot be generated after this call!- Overrides:
cleanUp
in classAbstractPlotInstances
-