org.briljantframework.data.dataframe.transform.ZNormalizer.java Source code

Introduction

Here is the source code for org.briljantframework.data.dataframe.transform.ZNormalizer.java

Source

/**
 * The MIT License (MIT)
 *
 * Copyright (c) 2016 Isak Karlsson
 *
 * Permission is hereby granted, free of charge, to any person obtaining a copy of this software and
 * associated documentation files (the "Software"), to deal in the Software without restriction,
 * including without limitation the rights to use, copy, modify, merge, publish, distribute,
 * sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is
 * furnished to do so, subject to the following conditions:
 *
 * The above copyright notice and this permission notice shall be included in all copies or
 * substantial portions of the Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT
 * NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
 * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM,
 * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 */
package org.briljantframework.data.dataframe.transform;

import org.apache.commons.math3.stat.descriptive.StatisticalSummary;
import org.briljantframework.Check;
import org.briljantframework.data.Is;
import org.briljantframework.data.dataframe.DataFrame;
import org.briljantframework.data.vector.Vector;
import org.briljantframework.data.vector.Vectors;

/**
 * Z normalization is also known as "Normalization to Zero Mean and Unit of Energy" first mentioned
 * in Goldin & Kanellakis. It ensures that all elements of the input vector are transformed into the
 * output vector whose mean is approximately 0 while the standard deviation are in a range close to
 * 1.
 *
 * @author Isak Karlsson
 */
public class ZNormalizer implements Transformation {

    @Override
    public Transformer fit(DataFrame df) {
        Vector.Builder meanBuilder = Vector.Builder.of(Double.class);
        Vector.Builder stdBuilder = Vector.Builder.of(Double.class);
        for (Object columnKey : df) {
            StatisticalSummary stats = Vectors.statisticalSummary(df.get(columnKey));
            if (stats.getN() <= 0 || Is.NA(stats.getMean()) || Is.NA(stats.getStandardDeviation())) {
                throw new IllegalArgumentException("Illegal value for column " + columnKey);
            }
            meanBuilder.set(columnKey, stats.getMean());
            stdBuilder.set(columnKey, stats.getStandardDeviation());
        }
        Vector mean = meanBuilder.build();
        Vector sigma = stdBuilder.build();

        return new ZNormalizerTransformer(mean, sigma);
    }

    @Override
    public String toString() {
        return "ZNormalizer";
    }

    private static class ZNormalizerTransformer implements Transformer {

        private final Vector mean;
        private final Vector sigma;

        public ZNormalizerTransformer(Vector mean, Vector sigma) {
            this.mean = mean;
            this.sigma = sigma;
        }

        @Override
        public DataFrame transform(DataFrame x) {
            Check.argument(mean.getIndex().equals(x.getColumnIndex()), "Columns must match.");
            DataFrame.Builder builder = x.newBuilder();
            for (Object columnKey : x) {
                Vector column = x.get(columnKey);
                double m = mean.getAsDouble(columnKey);
                double std = sigma.getAsDouble(columnKey);
                Vector.Builder normalized = column.newBuilder(column.size());
                for (int i = 0, size = column.size(); i < size; i++) {
                    double v = column.loc().getAsDouble(i);
                    if (Is.NA(v)) {
                        normalized.addNA();
                    } else if (std == 0) {
                        normalized.add(0);
                    } else {
                        normalized.add((v - m) / std);
                    }
                }
                builder.set(columnKey, normalized);
            }
            return builder.setIndex(x.getIndex()).build();
        }

        @Override
        public String toString() {
            return "ZNormalizerTransformer{" + "mean=" + mean + ", sigma=" + sigma + '}';
        }
    }
}