smile.stat.distribution.FDistribution Maven / Gradle / Ivy

Go to download
Show more of this group Show more artifacts with this name
Show all versions of smile-base Show documentation
smile-base
There is a newer version: 4.2.0
/*
 * Copyright (c) 2010-2021 Haifeng Li. All rights reserved.
 *
 * Smile is free software: you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation, either version 3 of the License, or
 * (at your option) any later version.
 *
 * Smile 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 General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with Smile.  If not, see .
 */

package smile.stat.distribution;

import smile.math.special.Beta;
import smile.math.special.Gamma;

import java.io.Serial;

/**
 * F-distribution arises in the testing of whether two observed samples have
 * the same variance. A random variate of the F-distribution arises as the
 * ratio of two chi-squared variates.
 *
 * @author Haifeng Li
 */
public class FDistribution implements Distribution {
    @Serial
    private static final long serialVersionUID = 2L;

    /**
     * The degrees of freedom of chi-square distribution in numerator.
     */
    public final int nu1;

    /**
     * The degrees of freedom chi-square distribution in denominator.
     */
    public final int nu2;

    /** The constant part in the pdf function. */
    private final double fac;

    /**
     * Constructor.
     * @param nu1 the degree of freedom of chi-square distribution in numerator.
     * @param nu2 the degree of freedom of chi-square distribution in denominator.
     */
    public FDistribution(int nu1, int nu2) {
        if (nu1 < 1) {
            throw new IllegalArgumentException("Invalid nu1 = " + nu1);
        }

        if (nu2 < 1) {
            throw new IllegalArgumentException("Invalid nu2 = " + nu2);
        }

        this.nu1 = nu1;
        this.nu2 = nu2;
        fac = 0.5 * (nu1 * Math.log(nu1) + nu2 * Math.log(nu2)) + Gamma.lgamma(0.5 * (nu1 + nu2))
                - Gamma.lgamma(0.5 * nu1) - Gamma.lgamma(0.5 * nu2);
    }

    @Override
    public int length() {
        return 2;
    }

    @Override
    public double mean() {
        return nu2 / (nu2 - 2.0);
    }

    @Override
    public double variance() {
        return 2.0 * nu2 * nu2 * (nu1 + nu2 - 2) / (nu1 * (nu2 - 2) * (nu2 - 2) * (nu2 - 4));
    }

    /**
     * Shannon's entropy. Not supported.
     */
    @Override
    public double entropy() {
        throw new UnsupportedOperationException("F-distribution does not support entropy()");
    }

    @Override
    public String toString() {
        return String.format("F-distribution(%d, %d)", nu1, nu2);
    }

    @Override
    public double rand() {
        return inverseTransformSampling();
    }

    @Override
    public double p(double x) {
        if (x <= 0.0) {
            throw new IllegalArgumentException("Invalid x: " + x);
        }

        return Math.exp((0.5 * nu1 - 1.0) * Math.log(x) - 0.5 * (nu1 + nu2) * Math.log(nu2 + nu1 * x) + fac);
    }

    @Override
    public double logp(double x) {
        if (x <= 0.0) {
            throw new IllegalArgumentException("Invalid x: " + x);
        }

        return (0.5 * nu1 - 1.0) * Math.log(x) - 0.5 * (nu1 + nu2) * Math.log(nu2 + nu1 * x) + fac;
    }

    @Override
    public double cdf(double x) {
        if (x < 0.0) {
            throw new IllegalArgumentException("Invalid x: " + x);
        }

        return Beta.regularizedIncompleteBetaFunction(0.5 * nu1, 0.5 * nu2, nu1 * x / (nu2 + nu1 * x));
    }

    @Override
    public double quantile(double p) {
        if (p < 0.0 || p > 1.0) {
            throw new IllegalArgumentException("Invalid p: " + p);
        }

        double x = Beta.inverseRegularizedIncompleteBetaFunction(0.5 * nu1, 0.5 * nu2, p);
        return nu2 * x / (nu1 * (1.0 - x));
    }
}