/*
 *  RapidMiner
 *
 *  Copyright (C) 2001-2008 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.preprocessing.discretization;

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

import com.rapidminer.example.Attribute;
import com.rapidminer.example.Example;
import com.rapidminer.example.ExampleSet;
import com.rapidminer.example.Statistics;
import com.rapidminer.example.set.SortedExampleSet;
import com.rapidminer.operator.Model;
import com.rapidminer.operator.OperatorDescription;
import com.rapidminer.operator.OperatorException;
import com.rapidminer.operator.UserError;
import com.rapidminer.operator.preprocessing.PreprocessingOperator;
import com.rapidminer.parameter.ParameterType;
import com.rapidminer.parameter.ParameterTypeBoolean;
import com.rapidminer.parameter.ParameterTypeInt;

/**
 * This operator discretizes all numeric attributes in the dataset into nominal attributes. This discretization is performed by equal frequency binning, i.e. the thresholds of all bins is selected in a way that all bins contain the same number of
 * numerical values. The number of bins is specified by a parameter, or, alternatively, is calculated as the square root of the number of examples with non-missing values (calculated for every single attribute). Skips all special attributes including
 * the label.
 * 
 * @author Sebastian Land, Ingo Mierswa
 * @version $Id: FrequencyDiscretization.java,v 1.6 2008/07/07 11:18:55 ingomierswa Exp $
 */
public class FrequencyDiscretization extends PreprocessingOperator {

	/** The parameter name for "If true, the number of bins is instead determined by the square root of the number of non-missing values." */
	public static final String PARAMETER_USE_SQRT_OF_EXAMPLES = "use_sqrt_of_examples";

	/** The parameter for the number of bins. */
	public static final String PARAMETER_NUMBER_OF_BINS = "number_of_bins";

	/** Indicates if long range names should be used. */
	public static final String PARAMETER_USE_LONG_RANGE_NAMES = "use_long_range_names";
	
	public FrequencyDiscretization(OperatorDescription description) {
		super(description);
	}

	public Model createPreprocessingModel(ExampleSet exampleSet) throws OperatorException {
		HashMap<Attribute, double[]> ranges = new HashMap<Attribute, double[]>();
		// Get and check parametervalues
		boolean useSqrt = getParameterAsBoolean(PARAMETER_USE_SQRT_OF_EXAMPLES);
		int numberOfBins = 0;
		if (!useSqrt) {
			// if not automatic sizing of bins, use parametervalue
			numberOfBins = getParameterAsInt(PARAMETER_NUMBER_OF_BINS);
			if (numberOfBins >= (exampleSet.size() - 1)) {
				throw new UserError(this, 116, PARAMETER_NUMBER_OF_BINS, "number of bins must be smaller than number of examples (here: " + exampleSet.size() + ")");
			}
		} else {
			exampleSet.recalculateAllAttributeStatistics();
		}

		for (Attribute currentAttribute : exampleSet.getAttributes()) {
			if (useSqrt) {
				numberOfBins = (int)Math.round(Math.sqrt(exampleSet.size() - (int) exampleSet.getStatistics(currentAttribute, Statistics.UNKNOWN)));
			}
			double[] attributeRanges = new double[numberOfBins];
			ExampleSet sortedSet = new SortedExampleSet(exampleSet, currentAttribute, SortedExampleSet.INCREASING);

			// finding ranges
			double examplesPerBin = exampleSet.size() / (double) numberOfBins;
			double currentBinSpace = examplesPerBin;
			double lastValue = Double.NaN;
			int currentBin = 0;

			for (Example example : sortedSet) {
				double value = example.getValue(currentAttribute);
				if (!Double.isNaN(value)) {
					// change bin if full and not last
					if (currentBinSpace < 1 && currentBin < numberOfBins && value != lastValue) {
						if (!Double.isNaN(lastValue)) {
							attributeRanges[currentBin] = (lastValue + value) / 2;
							currentBin++;
							currentBinSpace += examplesPerBin;
						}
					}
					currentBinSpace--;
					lastValue = value;
				}
			}
			attributeRanges[numberOfBins - 1] = Double.POSITIVE_INFINITY;
			ranges.put(currentAttribute, attributeRanges);
		}
		DiscretizationModel model = new DiscretizationModel(exampleSet);
		model.setRanges(ranges, "range", getParameterAsBoolean(PARAMETER_USE_LONG_RANGE_NAMES));
		return model;
	}

	public List<ParameterType> getParameterTypes() {
		List<ParameterType> types = super.getParameterTypes();
		ParameterType type = new ParameterTypeInt(PARAMETER_NUMBER_OF_BINS, "Defines the number of bins which should be used for each attribute.", 2, Integer.MAX_VALUE, 2);
		type.setExpert(false);
		types.add(type);
		types.add(new ParameterTypeBoolean(PARAMETER_USE_SQRT_OF_EXAMPLES, "If true, the number of bins is instead determined by the square root of the number of non-missing values.", false));
		types.add(new ParameterTypeBoolean(PARAMETER_USE_LONG_RANGE_NAMES, "Indicates if long range names including the limits should be used.", true));
		return types;
	}
}