/**
* Iterates over all models and returns the class with maximum likelihood.
*
* @param origExampleSet the set of examples to be classified
*/
@Override
public ExampleSet performPrediction(ExampleSet origExampleSet, Attribute predictedLabel)
throws OperatorException {
final String attributePrefix = "AdaBoostModelPrediction";
final int numLabels = predictedLabel.getMapping().size();
final Attribute[] specialAttributes = new Attribute[numLabels];
OperatorProgress progress = null;
if (getShowProgress() && getOperator() != null && getOperator().getProgress() != null) {
progress = getOperator().getProgress();
progress.setTotal(100);
}
for (int i = 0; i < numLabels; i++) {
specialAttributes[i] =
com.rapidminer.example.Tools.createSpecialAttribute(
origExampleSet, attributePrefix + i, Ontology.NUMERICAL);
if (progress != null) {
progress.setCompleted((int) (25.0 * (i + 1) / numLabels));
}
}
Iterator<Example> reader = origExampleSet.iterator();
int progressCounter = 0;
while (reader.hasNext()) {
Example example = reader.next();
for (int i = 0; i < specialAttributes.length; i++) {
example.setValue(specialAttributes[i], 0);
}
if (progress != null && ++progressCounter % OPERATOR_PROGRESS_STEPS == 0) {
progress.setCompleted((int) (25.0 * progressCounter / origExampleSet.size()) + 25);
}
}
reader = origExampleSet.iterator();
for (int modelNr = 0; modelNr < this.getNumberOfModels(); modelNr++) {
Model model = this.getModel(modelNr);
ExampleSet exampleSet = (ExampleSet) origExampleSet.clone();
exampleSet = model.apply(exampleSet);
this.updateEstimates(exampleSet, modelNr, specialAttributes);
PredictionModel.removePredictedLabel(exampleSet);
if (progress != null) {
progress.setCompleted((int) (25.0 * (modelNr + 1) / this.getNumberOfModels()) + 50);
}
}
// Turn prediction weights into confidences and a crisp predcition:
this.evaluateSpecialAttributes(origExampleSet, specialAttributes);
// Clean up attributes:
for (int i = 0; i < numLabels; i++) {
origExampleSet.getAttributes().remove(specialAttributes[i]);
origExampleSet.getExampleTable().removeAttribute(specialAttributes[i]);
if (progress != null) {
progress.setCompleted((int) (25.0 * (i + 1) / numLabels) + 75);
}
}
return origExampleSet;
}
public IOObject[] apply() throws OperatorException {
ExampleSet exampleSet = getInput(ExampleSet.class);
if (exampleSet.getAttributes().getLabel() == null) {
throw new UserError(this, 105);
}
if (!exampleSet.getAttributes().getLabel().isNominal()) {
throw new UserError(this, 101, "ROC Charts", exampleSet.getAttributes().getLabel());
}
if (exampleSet.getAttributes().getLabel().getMapping().getValues().size() != 2) {
throw new UserError(this, 114, "ROC Charts", exampleSet.getAttributes().getLabel());
}
if (exampleSet.getAttributes().getPredictedLabel() != null
&& getParameterAsBoolean(PARAMETER_USE_MODEL)) {
logWarning("Input example already has a predicted label which will be removed.");
PredictionModel.removePredictedLabel(exampleSet);
}
if (exampleSet.getAttributes().getPredictedLabel() == null
&& !getParameterAsBoolean(PARAMETER_USE_MODEL)) {
throw new UserError(this, 107);
}
Model model = null;
if (getParameterAsBoolean(PARAMETER_USE_MODEL)) {
model = getInput(Model.class);
exampleSet = model.apply(exampleSet);
}
if (exampleSet.getAttributes().getPredictedLabel() == null) {
throw new UserError(this, 107);
}
ROCDataGenerator rocDataGenerator = new ROCDataGenerator(1.0d, 1.0d);
ROCData rocPoints =
rocDataGenerator.createROCData(
exampleSet, getParameterAsBoolean(PARAMETER_USE_EXAMPLE_WEIGHTS));
rocDataGenerator.createROCPlotDialog(rocPoints);
PredictionModel.removePredictedLabel(exampleSet);
if (getParameterAsBoolean(PARAMETER_USE_MODEL)) {
return new IOObject[] {exampleSet, model};
} else return new IOObject[] {exampleSet};
}
/**
* Setting the parameter <code>MAX_MODEL_NUMBER</code> allows to discard all but the first n
* models for specified n.
*/
@Override
public void setParameter(String name, Object value) throws OperatorException {
if (name.equalsIgnoreCase(MAX_MODEL_NUMBER)) {
String stringValue = (String) value;
try {
this.maxModelNumber = Integer.parseInt(stringValue);
return;
} catch (NumberFormatException e) {
}
} else {
super.setParameter(name, value);
}
}
/**
* Applies the applier and evaluator (= second encapsulated inner operator). In order to reuse
* possibly created predicted label attributes, we do the following: We compare the predicted
* label of <code>testSet</code> before and after applying the inner operator. If it changed, the
* predicted label is removed again. No outer operator could ever see it. The same applies for the
* confidence attributes in case of classification learning.
*/
public IOContainer evaluate(ExampleSet testSet, IOContainer learnResult)
throws OperatorException {
if (learnResult == null) {
throw new RuntimeException(
"Wrong use of ValidationChain.evaluate(ExampleSet): "
+ "No preceding invocation of learn(ExampleSet)!");
}
Attribute predictedBefore = testSet.getAttributes().getPredictedLabel();
IOContainer evalInput = learnResult.append(new IOObject[] {testSet});
IOContainer result = getEvaluator().apply(evalInput);
Attribute predictedAfter = testSet.getAttributes().getPredictedLabel();
// remove predicted label and confidence attributes if there is a new prediction which is not
// equal to an old one
if ((predictedAfter != null)
&& ((predictedBefore == null)
|| (predictedBefore.getTableIndex() != predictedAfter.getTableIndex()))) {
PredictionModel.removePredictedLabel(testSet);
}
return result;
}