SSJ
V. labo.

umontreal.iro.lecuyer.probdistmulti
Class ContinuousDistributionMulti

java.lang.Object
  extended by umontreal.iro.lecuyer.probdistmulti.ContinuousDistributionMulti
Direct Known Subclasses:
ContinuousDistribution2Dim, DirichletDist, MultiNormalDist

public abstract class ContinuousDistributionMulti
extends Object

ContinuousDistributionMulti

Classes implementing continuous multi-dimensional distributions should inherit from this class. Such distributions are characterized by a density function f (x1, x2,…, xd); thus the signature of a density method is supplied here. All array indices start at 0.


Constructor Summary
ContinuousDistributionMulti()
           
 
Method Summary
abstract  double density(double[] x)
          .
abstract  double[][] getCorrelation()
          .
abstract  double[][] getCovariance()
          .
 int getDimension()
          .
abstract  double[] getMean()
          .
 
Methods inherited from class java.lang.Object
equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Constructor Detail

ContinuousDistributionMulti

public ContinuousDistributionMulti()
Method Detail

density

public abstract double density(double[] x)
. \begin{tabb}Returns $f(x_1, x_2, \ldots, x_d)$, the probability density of
 $X$\...
 ..._2, \ldots, x_d\}$. The convention is that
 $\texttt{x[i-1]} = x_i$.
 \end{tabb}
xvalue at which the density is evaluated density function evaluated at x


getDimension

public int getDimension()
. \begin{tabb}
 Returns the dimension $d$\ of the distribution.
 \end{tabb}


getMean

public abstract double[] getMean()
. \begin{tabb}
 Returns the mean vector of the distribution, defined as $\mu_{i} = E[X_i]$.
 \end{tabb}


getCovariance

public abstract double[][] getCovariance()
. \begin{tabb}
 Returns the variance-covariance matrix of the distribution, defined as\\
 $\sigma_{ij} = E[(X_i - \mu_i)(X_j - \mu_j)]$.
 \end{tabb}


getCorrelation

public abstract double[][] getCorrelation()
. \begin{tabb}
 Returns the correlation matrix of the distribution, defined as
 $\rho_{ij} = \sigma_{ij}/\sqrt{\sigma_{ii}\sigma_{jj}}$.
 \end{tabb}


SSJ
V. labo.

To submit a bug or ask questions, send an e-mail to Pierre L'Ecuyer.