package dr.math.distributions;
import dr.math.UnivariateFunction;
/**
* @author Alexei Drummond
*/
public class LaplaceDistribution implements Distribution {
// the mean parameter
double mu;
// the scale parameter
double beta;
// the maximum density
double c;
/**
* This general constructor creates a new laplace distribution with a
* specified mean and scale
*
* @param mu mean
* @param beta scale
*/
public LaplaceDistribution(double mu, double beta) {
setParameters(mu, beta);
}
/**
* This default constructor creates a new laplace distribution with rate 1
*/
public LaplaceDistribution() {
this(0, 1);
}
/**
* Set parameters of this distribution
*
* @param k mean
* @param b scale
*/
public void setParameters(double k, double b) {
if (b <= 0) b = 1;
mu = k;
beta = b;
//Normalizing constant
c = 1 / (2 * beta);
}
/**
* @return mean parameter
*/
public double getMu() {
return mu;
}
/**
* @return scale parameter
*/
public double getBeta() {
return beta;
}
/**
* @return maximum value of pdf function
*/
public double getMaxDensity() {
return c;
}
/**
* Cumulative distribution function
*/
public double cdf(double x) {
if (x == mu) return 0.5;
else return (0.5) * (1 + ((x - mu) / Math.abs(x - mu))
* (1 - Math.exp(-Math.abs(x - mu) / beta)));
}
/**
* Density function
*/
public double pdf(double x) {
return c * Math.exp(-Math.abs(x - mu) / beta);
}
public double logPdf(double x) {
return Math.log(c) - (Math.abs(x - mu) / beta);
}
public double quantile(double y) {
double sign = 1;
if (y < 0.5) sign = -1;
return mu - beta * sign * Math.log(1 - 2 * Math.abs(y - 0.5));
}
public double mean() {
return mu;
}
public double variance() {
return 2 * beta * beta;
}
public UnivariateFunction getProbabilityDensityFunction() {
return pdfFunction;
}
private final UnivariateFunction pdfFunction = new UnivariateFunction() {
public final double evaluate(double x) {
return pdf(x);
}
public final double getLowerBound() {
return Double.NEGATIVE_INFINITY;
}
public final double getUpperBound() {
return Double.POSITIVE_INFINITY;
}
};
}