All Downloads are FREE. Search and download functionalities are using the official Maven repository.

org.ojalgo.random.TDistribution Maven / Gradle / Ivy

Go to download

oj! Algorithms - ojAlgo - is Open Source Java code that has to do with mathematics, linear algebra and optimisation.

There is a newer version: 55.0.1
Show newest version
/*
 * Copyright 1997-2024 Optimatika
 *
 * 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.ojalgo.random;

import static org.ojalgo.function.constant.PrimitiveMath.*;

import org.ojalgo.function.special.BetaFunction;
import org.ojalgo.function.special.GammaFunction;

public class TDistribution extends AbstractContinuous {

    static final class Degree1 extends TDistribution {

        private final Cauchy myCauchy = new Cauchy();

        Degree1() {
            super(ONE);
        }

        @Override
        public double getDensity(final double value) {
            return myCauchy.getDensity(value);
        }

        @Override
        public double getDistribution(final double value) {
            return myCauchy.getDistribution(value);
        }

        @Override
        public double getQuantile(final double probability) {
            return myCauchy.getQuantile(probability);
        }

        @Override
        protected double generate() {
            return myCauchy.generate();
        }

    }

    static final class Degree2 extends TDistribution {

        Degree2() {
            super(TWO);
        }

        @Override
        public double getDensity(final double value) {
            return ONE / Math.pow(TWO + (value * value), THREE / TWO);
        }

        @Override
        public double getDistribution(final double value) {
            return HALF + (value / (TWO * Math.sqrt(TWO + (value * value))));
        }

        @Override
        public double getQuantile(final double probability) {
            double alpha = FOUR * probability * (ONE - probability);
            return TWO * (probability - HALF) * Math.sqrt(TWO / alpha);
        }

    }

    static final class Degree3 extends TDistribution {

        Degree3() {
            super(THREE);
        }

        @Override
        public double getDensity(final double value) {
            return (SIX * Math.sqrt(THREE)) / (PI * Math.pow(THREE + (value * value), TWO));
        }

        @Override
        public double getDistribution(final double value) {
            return HALF + ((ONE / PI) * (((ONE / SQRT.invoke(THREE)) * (value / (ONE + ((value * value) / THREE)))) + ATAN.invoke(value / SQRT.invoke(THREE))));
        }
    }

    static final class Degree4 extends TDistribution {

        Degree4() {
            super(FOUR);
        }

        @Override
        public double getDensity(final double value) {
            return THREE / (EIGHT * POW.invoke(ONE + ((value * value) / FOUR), FIVE / TWO));
        }

        @Override
        public double getDistribution(final double value) {
            double x = ONE + ((value * value) / FOUR);
            return HALF + (((THREE / EIGHT) * (value / SQRT.invoke(x))) * (ONE - (TWELFTH * ((value * value) / x))));
        }

        @Override
        public double getQuantile(final double probability) {
            double alpha = FOUR * probability * (ONE - probability);
            double sqrt = Math.sqrt(alpha);
            double q = Math.cos(THIRD * Math.acos(sqrt)) / sqrt;
            return Math.signum(probability - HALF) * TWO * Math.sqrt(q - ONE);
        }

    }

    static final class Degree5 extends TDistribution {

        private static final double SQRT5 = SQRT.invoke(FIVE);

        Degree5() {
            super(FIVE);
        }

        @Override
        public double getDensity(final double value) {
            return EIGHT / (THREE * PI * SQRT5 * POW.invoke(ONE + ((value * value) / FIVE), THREE));
        }

        @Override
        public double getDistribution(final double value) {
            double x = ONE + ((value * value) / FIVE);
            return HALF + ((((value / (SQRT5 * x)) * (ONE + (TWO / (THREE * x)))) + ATAN.invoke(value / SQRT5)) / PI);
        }

    }

    static final class DegreeInfinity extends TDistribution {

        private final Normal myNormal = new Normal();

        DegreeInfinity() {
            super(POSITIVE_INFINITY);
        }

        @Override
        public double getDensity(final double value) {
            return myNormal.getDensity(value);
        }

        @Override
        public double getDistribution(final double value) {
            return myNormal.getDistribution(value);
        }

        @Override
        public double getQuantile(final double probability) {
            return myNormal.getQuantile(probability);
        }

        @Override
        protected double generate() {
            return myNormal.generate();
        }

    }

    private static final double _0_0001 = 0.0001;
    private static final Normal NORMAL = new Normal();

    public static TDistribution of(final int degreesOfFreedom) {
        switch (degreesOfFreedom) {
        case 1:
            return new Degree1();
        case 2:
            return new Degree2();
        case 3:
            return new Degree3();
        case 4:
            return new Degree4();
        case 5:
            return new Degree5();
        case Integer.MAX_VALUE:
            return new DegreeInfinity();
        default:
            return new TDistribution(degreesOfFreedom);
        }
    }

    public static TDistribution ofInfinity() {
        return TDistribution.of(Integer.MAX_VALUE);
    }

    /**
     * The density and distribution functions share a common constant factor
     */
    private final double myConstant;

    private final double myDegreesOfFreedom;

    TDistribution(final double degreesOfFreedom) {
        super();
        myDegreesOfFreedom = Math.min(100, degreesOfFreedom);
        myConstant = GammaFunction.gamma((myDegreesOfFreedom + ONE) / TWO)
                / (Math.sqrt(myDegreesOfFreedom * PI) * GammaFunction.gamma(myDegreesOfFreedom / TWO));
    }

    public double getDensity(final double value) {
        return myConstant * Math.pow(ONE + ((value * value) / myDegreesOfFreedom), (NEG - myDegreesOfFreedom) / TWO);
    }

    public double getDistribution(final double value) {

        if (value >= ZERO) {

            double x = myDegreesOfFreedom / ((value * value) + myDegreesOfFreedom);

            return ONE - (HALF * BetaFunction.Regularized.beta(x, myDegreesOfFreedom / TWO, HALF));

        } else {

            return ONE - this.getDistribution(-value);
        }

        //        return HALF + (value * myConstant
        //                * HypergeometricFunction.hypergeometric(HALF, (myDegreesOfFreedom - ONE) / TWO, THREE / TWO, -(value * value) / myDegreesOfFreedom));
    }

    public double getExpected() {
        if (myDegreesOfFreedom > ONE) {
            return ZERO;
        } else {
            return NaN;
        }
    }

    public double getQuantile(final double probability) {

        double retVal = this.approximateQuantile(probability);

        if (Double.isInfinite(retVal)) {
            return retVal;
        }

        double reverse = this.getDistribution(retVal);

        double lower = retVal, higher = retVal;

        if ((probability - reverse) > _0_0001) {
            do {
                higher *= TWO;
            } while (this.getDistribution(higher) <= probability);
        } else if ((reverse - probability) > _0_0001) {
            do {
                lower /= TWO;
            } while (this.getDistribution(lower) >= probability);
        } else {
            return retVal;
        }

        do {
            retVal = (lower + higher) / TWO;
            reverse = this.getDistribution(retVal);
            if (reverse < probability) {
                lower = retVal;
            } else if (reverse > probability) {
                higher = retVal;
            }
        } while (Math.abs(reverse - probability) >= _0_0001);

        return retVal;
    }

    @Override
    public double getVariance() {
        if (myDegreesOfFreedom > TWO) {
            return myDegreesOfFreedom / (myDegreesOfFreedom - TWO);
        } else if (myDegreesOfFreedom > ONE) {
            return POSITIVE_INFINITY;
        } else {
            return NaN;
        }
    }

    private double approximateQuantile(final double probability) {
        return NORMAL.getQuantile(probability);
    }

}




© 2015 - 2024 Weber Informatics LLC | Privacy Policy