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 */
017package org.apache.commons.rng.sampling.distribution;
018
019import org.apache.commons.rng.UniformRandomProvider;
020
021/**
022 * <a href="https://en.wikipedia.org/wiki/Box%E2%80%93Muller_transform">
023 * Box-Muller algorithm</a> for sampling from Gaussian distribution with
024 * mean 0 and standard deviation 1.
025 *
026 * <p>Sampling uses {@link UniformRandomProvider#nextDouble()}.</p>
027 *
028 * @since 1.1
029 */
030public class BoxMullerNormalizedGaussianSampler
031    implements NormalizedGaussianSampler, SharedStateContinuousSampler {
032    /** Next gaussian. */
033    private double nextGaussian = Double.NaN;
034    /** Underlying source of randomness. */
035    private final UniformRandomProvider rng;
036
037    /**
038     * @param rng Generator of uniformly distributed random numbers.
039     */
040    public BoxMullerNormalizedGaussianSampler(UniformRandomProvider rng) {
041        this.rng = rng;
042    }
043
044    /** {@inheritDoc} */
045    @Override
046    public double sample() {
047        double random;
048        if (Double.isNaN(nextGaussian)) {
049            // Generate a pair of Gaussian numbers.
050
051            // Avoid zero for the uniform deviate y.
052            // The extreme tail of the sample is:
053            // y = 2^-53
054            // r = 8.57167
055            final double x = rng.nextDouble();
056            final double y = InternalUtils.makeNonZeroDouble(rng.nextLong());
057            final double alpha = 2 * Math.PI * x;
058            final double r = Math.sqrt(-2 * Math.log(y));
059
060            // Return the first element of the generated pair.
061            random = r * Math.cos(alpha);
062
063            // Keep second element of the pair for next invocation.
064            nextGaussian = r * Math.sin(alpha);
065        } else {
066            // Use the second element of the pair (generated at the
067            // previous invocation).
068            random = nextGaussian;
069
070            // Both elements of the pair have been used.
071            nextGaussian = Double.NaN;
072        }
073
074        return random;
075    }
076
077    /** {@inheritDoc} */
078    @Override
079    public String toString() {
080        return "Box-Muller normalized Gaussian deviate [" + rng.toString() + "]";
081    }
082
083    /**
084     * {@inheritDoc}
085     *
086     * @since 1.3
087     */
088    @Override
089    public SharedStateContinuousSampler withUniformRandomProvider(UniformRandomProvider rng) {
090        return new BoxMullerNormalizedGaussianSampler(rng);
091    }
092
093    /**
094     * Create a new normalised Gaussian sampler.
095     *
096     * @param <S> Sampler type.
097     * @param rng Generator of uniformly distributed random numbers.
098     * @return the sampler
099     * @since 1.3
100     */
101    @SuppressWarnings("unchecked")
102    public static <S extends NormalizedGaussianSampler & SharedStateContinuousSampler> S
103            of(UniformRandomProvider rng) {
104        return (S) new BoxMullerNormalizedGaussianSampler(rng);
105    }
106}