001/*
002 * Licensed to the Apache Software Foundation (ASF) under one or more
003 * contributor license agreements.  See the NOTICE file distributed with
004 * this work for additional information regarding copyright ownership.
005 * The ASF licenses this file to You under the Apache License, Version 2.0
006 * (the "License"); you may not use this file except in compliance with
007 * the License.  You may obtain a copy of the License at
008 *
009 *      http://www.apache.org/licenses/LICENSE-2.0
010 *
011 * Unless required by applicable law or agreed to in writing, software
012 * distributed under the License is distributed on an "AS IS" BASIS,
013 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
014 * See the License for the specific language governing permissions and
015 * limitations under the License.
016 */
017
018package org.apache.commons.math3.ml.neuralnet.twod.util;
019
020import org.apache.commons.math3.ml.neuralnet.MapUtils;
021import org.apache.commons.math3.ml.neuralnet.Neuron;
022import org.apache.commons.math3.ml.neuralnet.twod.NeuronSquareMesh2D;
023import org.apache.commons.math3.ml.distance.DistanceMeasure;
024
025/**
026 * Computes the quantization error histogram.
027 * Each bin will contain the average of the distances between samples
028 * mapped to the corresponding unit and the weight vector of that unit.
029 * @since 3.6
030 */
031public class QuantizationError implements MapDataVisualization {
032    /** Distance. */
033    private final DistanceMeasure distance;
034
035    /**
036     * @param distance Distance.
037     */
038    public QuantizationError(DistanceMeasure distance) {
039        this.distance = distance;
040    }
041
042    /** {@inheritDoc} */
043    public double[][] computeImage(NeuronSquareMesh2D map,
044                                   Iterable<double[]> data) {
045        final int nR = map.getNumberOfRows();
046        final int nC = map.getNumberOfColumns();
047
048        final LocationFinder finder = new LocationFinder(map);
049
050        // Hit bins.
051        final int[][] hit = new int[nR][nC];
052        // Error bins.
053        final double[][] error = new double[nR][nC];
054
055        for (double[] sample : data) {
056            final Neuron best = MapUtils.findBest(sample, map, distance);
057
058            final LocationFinder.Location loc = finder.getLocation(best);
059            final int row = loc.getRow();
060            final int col = loc.getColumn();
061            hit[row][col] += 1;
062            error[row][col] += distance.compute(sample, best.getFeatures());
063        }
064
065        for (int r = 0; r < nR; r++) {
066            for (int c = 0; c < nC; c++) {
067                final int count = hit[r][c];
068                if (count != 0) {
069                    error[r][c] /= count;
070                }
071            }
072        }
073
074        return error;
075    }
076}