/*
 *  RapidMiner
 *
 *  Copyright (C) 2001-2011 by Rapid-I and the contributors
 *
 *  Complete list of developers available at our web site:
 *
 *       http://rapid-i.com
 *
 *  This program is free software: you can redistribute it and/or modify
 *  it under the terms of the GNU Affero General Public License as published by
 *  the Free Software Foundation, either version 3 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 Affero General Public License for more details.
 *
 *  You should have received a copy of the GNU Affero General Public License
 *  along with this program.  If not, see http://www.gnu.org/licenses/.
 */
package com.rapidminer.operator.postprocessing;

import java.util.HashMap;
import java.util.List;

import com.rapidminer.example.Attribute;
import com.rapidminer.example.Attributes;
import com.rapidminer.example.Example;
import com.rapidminer.example.ExampleSet;
import com.rapidminer.operator.OperatorDescription;
import com.rapidminer.operator.OperatorException;
import com.rapidminer.operator.UserError;
import com.rapidminer.operator.ports.metadata.ExampleSetPrecondition;
import com.rapidminer.operator.preprocessing.AbstractDataProcessing;
import com.rapidminer.parameter.ParameterType;
import com.rapidminer.parameter.ParameterTypeCategory;
import com.rapidminer.parameter.ParameterTypeDouble;
import com.rapidminer.parameter.ParameterTypeList;
import com.rapidminer.parameter.ParameterTypeString;
import com.rapidminer.parameter.conditions.EqualTypeCondition;
import com.rapidminer.tools.Ontology;

/**
 * This operator sets all predictions which do not have a higher confidence than the 
 * specified one to "unknown" (missing value). This operator is a quite simple
 * version of the CostBasedThresholdLearner which might be useful in simple binominal
 * classification settings (although it does also work for polynominal classifications).
 *  
 * @author Ingo Mierswa
 */
public class SimpleUncertainPredictionsTransformation extends AbstractDataProcessing {

	public static final String PARAMETER_CLASS_HANDLING = "class_handling";
	
	public static final String[] CLASS_HANDLING_MODES = { "balanced" , "unbalanced" };
	
	public static final int CLASS_HANDLING_BALANCED = 0;
	
	public static final int CLASS_HANDLING_UNBALANCED = 1;
	
	public static final String PARAMETER_MIN_CONFIDENCE = "min_confidence";
	
	public static final String PARAMETER_MIN_CONFIDENCES = "min_confidences";
	
	public static final String PARAMETER_CLASS_VALUE = "class";


	public SimpleUncertainPredictionsTransformation(OperatorDescription description) {
		super(description);

		getExampleSetInputPort().addPrecondition(new ExampleSetPrecondition(getExampleSetInputPort(), Ontology.VALUE_TYPE, Attributes.PREDICTION_NAME, Attributes.CONFIDENCE_NAME));
	}

	@Override
	public ExampleSet apply(ExampleSet exampleSet) throws OperatorException {		
		// checks
		Attribute predictedLabel = exampleSet.getAttributes().getPredictedLabel(); 
		if (predictedLabel == null) {
			throw new UserError(this, 107);
		}
		if (!predictedLabel.isNominal()) {
			throw new UserError(this, 119, predictedLabel, getName());			
		}

		switch (getParameterAsInt(PARAMETER_CLASS_HANDLING)) {
		case CLASS_HANDLING_BALANCED:
			double minConfidence = getParameterAsDouble(PARAMETER_MIN_CONFIDENCE);
			for (Example example : exampleSet) {
				double predictionValue = example.getValue(predictedLabel);
				String predictionClass = predictedLabel.getMapping().mapIndex((int) predictionValue);
				double confidence = example.getConfidence(predictionClass);
				if (!Double.isNaN(confidence)) {
					if (confidence < minConfidence) {
						example.setValue(predictedLabel, Double.NaN);
					}
				}
			}
			break;
		case CLASS_HANDLING_UNBALANCED:
			HashMap<String, Double> thresholdMap = new HashMap<String, Double>();
			for (String[] threshold : getParameterList(PARAMETER_MIN_CONFIDENCES)) {
				thresholdMap.put(threshold[0], Double.valueOf(threshold[1]));
			}
			
			for (Example example : exampleSet) {
				double predictionValue = example.getValue(predictedLabel);
				String predictionClass = predictedLabel.getMapping().mapIndex((int) predictionValue);
				double confidence = example.getConfidence(predictionClass);
				Double threshold  = thresholdMap.get(predictionClass);
				if (!Double.isNaN(confidence) && threshold != null) {
					if (confidence < threshold.doubleValue()) {
						example.setValue(predictedLabel, Double.NaN);
					}
				}
			}
			break;
		}

		return exampleSet;
	}
	
	@Override
	public boolean writesIntoExistingData() {
		return true;
	}
	
	@Override
	public List<ParameterType> getParameterTypes() {
		List<ParameterType> list = super.getParameterTypes();
		list.add(new ParameterTypeCategory(PARAMETER_CLASS_HANDLING, "The mode which defines if all classes are handled equally or if class individual thresholds are set.", CLASS_HANDLING_MODES, CLASS_HANDLING_BALANCED, false));
		ParameterType type = new ParameterTypeDouble(PARAMETER_MIN_CONFIDENCE, "The minimal confidence necessary for not setting the prediction to 'unknown'.", 0.0d, 1.0d, 0.5d);
		type.registerDependencyCondition(new EqualTypeCondition(this, PARAMETER_CLASS_HANDLING, CLASS_HANDLING_MODES, true, CLASS_HANDLING_BALANCED));
		type.setExpert(false);
		list.add(type);
		type = new ParameterTypeList(PARAMETER_MIN_CONFIDENCES, "A list which defines individual thresholds for classes.",
									 new ParameterTypeString(PARAMETER_CLASS_VALUE, "The class for which the confidence threshold should be set."),
									 new ParameterTypeDouble(PARAMETER_MIN_CONFIDENCE, "The minimal confidence necessary for not setting the prediction to 'unknown'.", 0.0d, 1.0d, 0.5d),
									 false);
		type.registerDependencyCondition(new EqualTypeCondition(this, PARAMETER_CLASS_HANDLING, CLASS_HANDLING_MODES, true, CLASS_HANDLING_UNBALANCED));
		list.add(type);
		return list;
	}
}