Example usage for org.apache.commons.math.random RandomData nextUniform

List of usage examples for org.apache.commons.math.random RandomData nextUniform

  1. HOME
  2. Java
  3. org.apache.commons
  4. org.apache.commons.math.random.*
  5. RandomData
  6. nextUniform

Introduction

In this page you can find the example usage for org.apache.commons.math.random RandomData nextUniform.

Prototype

double nextUniform(double lower, double upper);

Source Link

Document

Generates a uniformly distributed random value from the open interval (lower,upper) (i.e., endpoints excluded).

Usage

From source file:datafu.pig.hash.lsh.RepeatingLSH.java

public RepeatingLSH(List<LSH> lshList) throws MathException {
    super(lshList.get(0).getDim(), lshList.get(0).getRandomGenerator());
    this.lshList = lshList;
    RandomGenerator rg = lshList.get(0).getRandomGenerator();
    RandomData rd = new RandomDataImpl(rg);
    /*/*  w  ww . j a  va 2  s.  com*/
     * Compute a random vector of lshList.size() with each component taken from U(0,10)
     */
    randomVec = new ArrayRealVector(lshList.size());
    for (int i = 0; i < randomVec.getDimension(); ++i) {
        randomVec.setEntry(i, rd.nextUniform(0, 10.0));
    }
}

From source file:br.upe.ecomp.doss.algorithm.apso.APSO.java

private void adaptParameters(double evolFactor) {
    RandomData random = new RandomDataImpl();
    double accelerationRate = random.nextUniform(0.05, 0.1);
    double c1 = getC1();
    double c2 = getC2();
    if (currentEvolutionaryState.equals(EnumEvolutionaryState.CONVERGENCE)) {
        // C1 - increase slightly | C2 increase slightly
        c1 = c1 + (c1 * accelerationRate * 0.5);
        c2 = c2 + (c2 * accelerationRate * 0.5);
    } else if (currentEvolutionaryState.equals(EnumEvolutionaryState.EXPLOITATION)) {
        // C1 - increase | C2 decrease
        c1 = c1 + (c1 * accelerationRate);
        c2 = c2 - (c2 * accelerationRate);
    } else if (currentEvolutionaryState.equals(EnumEvolutionaryState.EXPLORATION)) {
        // C1 - increase slightly | C2 decrease slightly
        c1 = c1 + (c1 * accelerationRate * 0.5);
        c2 = c2 - (c2 * accelerationRate * 0.5);
    } else if (currentEvolutionaryState.equals(EnumEvolutionaryState.JUMPING_OUT)) {
        // C1 - decrease | C2 increase
        c1 = c1 - (c1 * accelerationRate);
        c2 = c2 + (c2 * accelerationRate);
    }/* w  w w  .ja  v  a  2  s. c  om*/
    // Normalization
    if ((c1 + c2) > (cMin + cMax)) {
        double c1Temp = c1;
        double c2Temp = c2;
        c1Temp = c1 * (cMin + cMax) / (c1 + c2);
        c2Temp = c2 * (cMin + cMax) / (c1 + c2);
        c1 = c1Temp;
        c2 = c2Temp;
    }

    // Verify and adjust C1 and C2 by bounds
    if (c1 < cMin) {
        c1 = cMin;
    }
    if (c2 < cMin) {
        c2 = cMin;
    }
    if (c1 > cMax) {
        c1 = cMax;
    }
    if (c2 > cMax) {
        c2 = cMax;
    }

    setC1(c1);
    setC2(c2);

    // Adapt inertia weight
    double inertiaWeight = 1 / (1 + 1.5 * Math.exp(-2.6 * evolFactor));
    // Verify and adjust inertia weight by bounds
    if (inertiaWeight < inertiaWeightMin) {
        inertiaWeight = inertiaWeightMin;
    }
    if (inertiaWeight > inertiaWeightMax) {
        inertiaWeight = inertiaWeightMax;
    }
    setInertiaWeight(inertiaWeight);
}

From source file:br.upe.ecomp.doss.algorithm.chargedpso.ChargedPSOParticle.java

private double getRandomInRange(double value1, double value2) {
    RandomData random = new RandomDataImpl();
    double lowerBound;
    double upperBound;
    if (value1 < value2) {
        lowerBound = value1;/*from   ww w .j  a  va2  s  . co  m*/
        upperBound = value2;
    } else {
        lowerBound = value2;
        upperBound = value1;
    }
    return random.nextUniform(lowerBound, upperBound);
}

From source file:br.upe.ecomp.doss.algorithm.fss.Fish.java

public void performIndividualMovement(double[] stepInd, Problem problem) {
    RandomData random = new RandomDataImpl();
    int dimensions = getDimensions();
    double[] currentPosition = getCurrentPosition().clone();

    double[] nextPosition = new double[getDimensions()];
    for (int i = 0; i < dimensions; i++) {
        nextPosition[i] = currentPosition[i] + random.nextUniform(-1.0, 1.0) * stepInd[i];
        if (nextPosition[i] > problem.getUpperBound(i) || nextPosition[i] < problem.getLowerBound(i)) {
            nextPosition[i] = currentPosition[i];
        }//  ww w . j  a  va  2 s.  c  o m
    }

    deltaFitness = problem.getFitness(nextPosition) - getCurrentFitness();
    if (problem.isFitnessBetterThan(getCurrentFitness(), problem.getFitness(nextPosition))
            && deltaFitness < 0) {
        deltaFitness *= -1;
    }

    if (problem.isFitnessBetterThan(getCurrentFitness(), problem.getFitness(nextPosition))) {
        previousPosition = getCurrentPosition().clone();
        updateCurrentPosition(nextPosition, problem.getFitness(nextPosition));

        if (problem.isFitnessBetterThan(getBestFitness(), getCurrentFitness())) {
            updateBestPosition(getCurrentPosition(), getCurrentFitness());
        }
    } else {
        deltaFitness = 0;
    }
}

From source file:br.upe.ecomp.doss.algorithm.chargedpso.ChargedPSOParticle.java

/**
 * Updates the current position of the particle.
 * /* ww  w .  ja  v a2 s . com*/
 * @param problem The problem that we are trying to solve.
 */
@Override
public void updateCurrentPosition(Problem problem) {
    RandomData random = new RandomDataImpl();
    double[] position = getCurrentPosition();
    double[] velocity = getVelocity();
    double newPosition;
    for (int i = 0; i < getDimensions(); i++) {
        double p = position[i];
        newPosition = position[i] + velocity[i];
        double p1 = newPosition;
        if (charge != 0) {
            double step;

            if (newPosition > problem.getUpperBound(i)) {
                step = Math.abs(random.nextUniform(0, problem.getUpperBound(i) - problem.getLowerBound(i)));
                newPosition = problem.getUpperBound(i) - step;
            } else if (newPosition < problem.getLowerBound(i)) {
                step = Math.abs(random.nextUniform(0, problem.getUpperBound(i) - problem.getLowerBound(i)));
                newPosition = problem.getLowerBound(i) + step;
            }

            if (newPosition > problem.getUpperBound(i)) {
                System.out.println("parou");
            } else if (newPosition < problem.getLowerBound(i)) {
                System.out.println("parou 1");
            }

        } else if (newPosition > problem.getUpperBound(i)) {
            newPosition = problem.getUpperBound(i);
        } else if (newPosition < problem.getLowerBound(i)) {
            newPosition = problem.getLowerBound(i);
        }
        if (newPosition > problem.getUpperBound(i)) {
            System.out.println("ops");
        }
        position[i] = newPosition;
    }
    updateCurrentPosition(position, problem.getFitness(position));

}

From source file:br.upe.ecomp.doss.algorithm.fss.FSS.java

private void performCollectiveVolitiveMovement() {
    RandomData random = new RandomDataImpl();
    double[] barycenter = calculateBarycenter();

    double actualFishSchoolWeight = getFishSchoolWeigth();

    double euclidianDistance;
    double[] position;
    double currentPosition;
    for (Fish fish : fishSchool) {

        euclidianDistance = MathUtils.distance(fish.getCurrentPosition(), barycenter);
        position = new double[dimensions];

        for (int i = 0; i < dimensions; i++) {
            currentPosition = fish.getCurrentPosition()[i];

            if (actualFishSchoolWeight > previousFishSchoolWeight) {

                position[i] = currentPosition - stepVol[i] * random.nextUniform(0.0, 1.0)
                        * ((currentPosition - barycenter[i]) / euclidianDistance);
                // * ((currentPosition - barycenter[i]) / euclidianDistance);
            } else {
                position[i] = currentPosition + stepVol[i] * random.nextUniform(0.0, 1.0)
                        * ((currentPosition - barycenter[i]) / euclidianDistance);
                // * ((currentPosition - barycenter[i]) / euclidianDistance);
            }//from w  w  w. j av a 2  s  .  c  om
            // * ((currentPosition - barycenter[i]) / euclidianDistance);
            if (position[i] > getProblem().getUpperBound(i) || position[i] < getProblem().getLowerBound(i)) {
                position[i] = currentPosition;
            }
        }

        // if (getProblem().compareFitness(fish.getCurrentFitness(),
        // getProblem().getFitness(position))) {
        fish.updateCurrentPosition(position, getProblem().getFitness(position));

        if (getProblem().isFitnessBetterThan(fish.getBestFitness(), fish.getCurrentFitness())) {
            fish.updateBestPosition(fish.getCurrentPosition(), fish.getCurrentFitness());
        }
        // }
    }

    previousFishSchoolWeight = actualFishSchoolWeight;
}

From source file:br.upe.ecomp.doss.algorithm.volitiveapso.VolitiveAPSO.java

private void performCollectiveVolitiveMovement() {
    RandomData random = new RandomDataImpl();
    double[] barycenter = calculateBarycenter();

    double actualSwarmWeight = getSwarmWeigth();

    double euclidianDistance;
    double[] position;
    double currentPosition;

    WeightedAPSOParticle[] particles = (WeightedAPSOParticle[]) getParticles();
    for (WeightedAPSOParticle particle : particles) {

        euclidianDistance = MathUtils.distance(particle.getCurrentPosition(), barycenter);
        position = new double[getDimensions()];

        for (int i = 0; i < getDimensions(); i++) {
            currentPosition = particle.getCurrentPosition()[i];

            if (actualSwarmWeight > previousSwarmWeight) {

                position[i] = currentPosition - stepVol[i] * random.nextUniform(0.0, 1.0)
                        * ((currentPosition - barycenter[i]) / euclidianDistance);
            } else {
                position[i] = currentPosition + stepVol[i] * random.nextUniform(0.0, 1.0)
                        * ((currentPosition - barycenter[i]) / euclidianDistance);
            }//from www.  ja  va 2s  .c  o m
            if (position[i] > getProblem().getUpperBound(i) || position[i] < getProblem().getLowerBound(i)) {
                position[i] = currentPosition;
            }
        }

        // if (getProblem().compareFitness(fish.getCurrentFitness(),
        // getProblem().getFitness(position))) {
        particle.updateCurrentPosition(position, getProblem().getFitness(position));

        if (getProblem().isFitnessBetterThan(particle.getBestFitness(), particle.getCurrentFitness())) {
            particle.updateBestPosition(particle.getCurrentPosition(), particle.getCurrentFitness());
        }
        // }
    }

    previousSwarmWeight = actualSwarmWeight;
}

From source file:br.upe.ecomp.doss.algorithm.volitivepso.VolitivePSO.java

private void performCollectiveVolitiveMovement() {

    RandomData random = new RandomDataImpl();
    double[] barycenter = calculateBarycenter();

    double actualSwarmWeight = getSwarmWeigth();

    double euclidianDistance;
    double[] position;
    double currentPosition;

    WeightedPSOParticle[] particles = (WeightedPSOParticle[]) getParticles();
    for (WeightedPSOParticle particle : particles) {

        euclidianDistance = MathUtils.distance(particle.getCurrentPosition(), barycenter);
        position = new double[getDimensions()];

        for (int i = 0; i < getDimensions(); i++) {
            currentPosition = particle.getCurrentPosition()[i];

            if (actualSwarmWeight > previousSwarmWeight) {

                position[i] = currentPosition - stepVol[i] * random.nextUniform(0.0, 1.0)
                        * ((currentPosition - barycenter[i]) / euclidianDistance);
            } else {
                position[i] = currentPosition + stepVol[i] * random.nextUniform(0.0, 1.0)
                        * ((currentPosition - barycenter[i]) / euclidianDistance);
            }/*  w w w  . ja  v a  2  s.c  om*/
            if (position[i] > getProblem().getUpperBound(i) || position[i] < getProblem().getLowerBound(i)) {
                position[i] = currentPosition;
            }
        }

        // if (getProblem().compareFitness(fish.getCurrentFitness(),
        // getProblem().getFitness(position))) {
        particle.updateCurrentPosition(position, getProblem().getFitness(position));

        if (getProblem().isFitnessBetterThan(particle.getBestFitness(), particle.getCurrentFitness())) {
            particle.updateBestPosition(particle.getCurrentPosition(), particle.getCurrentFitness());
        }
        // }
    }

    previousSwarmWeight = actualSwarmWeight;
}

From source file:org.tolven.gen.model.ChooseScenario.java

/**
 * //from w  w  w  . j a  v a  2  s . c o m
 * @param patient
 * @param startTime
 * @param endTime
 * @param rng
 * @return Return true if one of the child scenarios did something (returned true), otherwise false. 
 */
public boolean apply(GenMedical patient, Date startTime, Date endTime, RandomData rng) {
    double choice = rng.nextUniform(0.0, 1.0);
    double base = 0.0;
    for (ProbabilityScenario s : getScenarios()) {
        base += s.getProbability();
        if (base > choice) {
            //            TolvenLogger.info( "Scenario Choice " + s.getTitle(), ChooseScenario.class);
            return s.apply(patient, startTime, endTime, rng);
        }
    }
    return false;
}

From source file:org.tolven.gen.model.DiabetesGen.java

public DiabetesGen(RandomData rng) {
    this.rng = rng;
    setDiseaseName("Diabetes Mellitus");
    //      addMatcher( new DemogMatcher( 0, 20, "M", 0.04*demoSkew));
    //      addMatcher( new DemogMatcher( 0, 20, "F", 0.06*demoSkew));
    //      addMatcher( new DemogMatcher( 20, 40, "M", 0.012*demoSkew));
    //      addMatcher( new DemogMatcher( 20, 40, "F", 0.018*demoSkew));
    addMatcher(new DemogMatcher(40, 60, "M", 0.052 * demoSkew));
    addMatcher(new DemogMatcher(40, 60, "F", 0.078 * demoSkew));
    addMatcher(new DemogMatcher(60, 200, "M", 0.08 * demoSkew));
    addMatcher(new DemogMatcher(60, 200, "F", 0.12 * demoSkew));
    // Some scenario components we'll use below
    Scenario podiatry = new RadiologyScenario("Foot Exam", "No abnormality noted on either foot.", "Normal");
    Scenario eagerLab = new GroupScenario("0.50 scenario in Diabetes",
            new Scenario[] {
                    new LabTestScenario("hemoglobin A1C test", "%", "7 (or less)", 1,
                            new GaussianValueGen(6.0, rng.nextUniform(0.5, 1.3))),
                    new LabTestScenario("Creatinine", "mg/dL", "0.9 to 1.4 mg/dL", 1,
                            new GaussianValueGen(1.2, rng.nextUniform(0.1, 0.3))),
                    new BatteryScenario("Lipid Panel",
                            new Scenario[] {
                                    new LabTestScenario("low-density lipoprotein (LDL)", "mg/dL", ">100 mg/dL",
                                            0, new GaussianValueGen(85.0, rng.nextUniform(10.0, 20.0))),
                                    new LabTestScenario("triglycerides", "mg/dL", ">150 mg/dL", 0,
                                            new GaussianValueGen(140.0, rng.nextUniform(5.0, 15.0))) }),
                    new LabTestScenario("urine albumin", "mcg", ">30 mcg", 0,
                            new GaussianValueGen(29.0, rng.nextUniform(4.0, 7.0))) });
    Scenario lazyLab = new GroupScenario("0.25 scenario in Diabetes",
            new Scenario[] {
                    new LabTestScenario("hemoglobin A1C test", "%", "7 (or less)", 1,
                            new GaussianValueGen(9.0, rng.nextUniform(1.0, 3.5))),
                    new LabTestScenario("Creatinine", "mg/dL", "0.9 to 1.4 mg/dL", 1,
                            new GaussianValueGen(1.1, rng.nextUniform(0.3, 0.7))),
                    new BatteryScenario("Lipid Panel",
                            new Scenario[] {
                                    new LabTestScenario("low-density lipoprotein (LDL)", "mg/dL", ">100 mg/dL",
                                            0, new GaussianValueGen(165.0, rng.nextUniform(10.0, 40.0))),
                                    new LabTestScenario("triglycerides", "mg/dL", ">150 mg/dL", 0,
                                            new GaussianValueGen(180.0, rng.nextUniform(15.0, 28.0))) }),
                    new LabTestScenario("urine albumin", "mcg", ">30 mcg", 0,
                            new GaussianValueGen(35.0, rng.nextUniform(5.1, 11.2))) });
    Scenario annualAppt = new AppointmentScenario("12 Month diabetes followup", null, null, null);
    //      Scenario semiAnnualAppt = new AppointmentScenario( "6 Month diabetes lab work", null, null, null);

    // Specify the scenario here.
    this.setScenario(new ChooseScenario(new ProbabilityScenario[] {
            // This says on 12 month cycle, do an appointment from which lab and podiatry exam events occur and then, repeat once at a 6 month interval a second lab scenario.
            // In addition, we'll schedule an annual appointment for one time in the future. (Ignore six-month appoitment) 
            new ProbabilityScenario(0.50,
                    new RepeatScenario(GregorianCalendar.MONTH, 12, 1.5, false, 1, 0,
                            new GroupScenario("12 month repeat",
                                    new Scenario[] { annualAppt, eagerLab, podiatry,
                                            new RepeatScenario(GregorianCalendar.MONTH, 6, 1.5, false, 1, 1,
                                                    eagerLab) }),
                            annualAppt, 1)),
            new ProbabilityScenario(0.25,
                    new RepeatScenario(GregorianCalendar.MONTH, 8, 2.1, true, 0, 0,
                            new GroupScenario("12 month repeat",
                                    new Scenario[] { annualAppt, lazyLab, podiatry, new RepeatScenario(
                                            GregorianCalendar.MONTH, 6, 1.5, false, 1, 1, lazyLab) }),
                            annualAppt, 1)) }));
}