repicea.stats

Class StatisticalUtility

java.lang.Object

repicea.stats.StatisticalUtility

public class StatisticalUtility
extends java.lang.Object

This class contains static methods that are useful for statistical regressions.

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	StatisticalUtility.TypeMatrixR

Constructor Summary

Constructors

Constructor and Description
StatisticalUtility()

Method Summary

Modifier and Type	Method and Description
static Matrix	combineMatrices(Matrix mat1, Matrix mat2) This method combines two row vectors of dummy variables.
static Matrix	constructRMatrix(Matrix coordinates, double rho, double gamma, double residual, StatisticalUtility.TypeMatrixR type) Construct a within-subject correlation matrix using a rho parameter, a gamma parameter, a residual parameter and a column vector of coordinates.
static Matrix	constructRMatrix(Matrix coordinates, double varianceParameter, double covarianceParameter, StatisticalUtility.TypeMatrixR type) Construct a within-subject correlation matrix using a variance parameter, a correlation parameter and a column vector of coordinates.
static Matrix	drawRandomVector(int nrow, Distribution.Type type) This method generates a random vector
static Matrix	drawRandomVector(int nrow, Distribution.Type type, java.util.Random random) This method generates a random vector
static long	getCombinations(int n, int d) This method returns the number of combinations.
static REpiceaRandom	getRandom() This method returns a Random generator.
static Matrix	performSpecialAdd(Matrix originalMatrix, Matrix matrixToAdd) This method performs a special addition in which only the elements different from 0 and 1 are involved.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

StatisticalUtility

public StatisticalUtility()

Method Detail

constructRMatrix

public static Matrix constructRMatrix(Matrix coordinates,
                                      double varianceParameter,
                                      double covarianceParameter,
                                      StatisticalUtility.TypeMatrixR type)

Construct a within-subject correlation matrix using a variance parameter, a correlation parameter and a column vector of coordinates.

Parameters:

coordinates - a column vector of coordinates from which the distances are calculated

varianceParameter - the variance parameter

covarianceParameter - the covariance parameter

type - the type of correlation

Returns:

the resulting matrix

constructRMatrix

public static Matrix constructRMatrix(Matrix coordinates,
                                      double rho,
                                      double gamma,
                                      double residual,
                                      StatisticalUtility.TypeMatrixR type)

Construct a within-subject correlation matrix using a rho parameter, a gamma parameter, a residual parameter and a column vector of coordinates.

Parameters:

coordinates - a column vector of coordinates from which the distances are calculated

rho - the rho parameter

gamma - the gamma parameter

residual - the residual parameter

type - the type of correlation

Returns:

the resulting matrix

drawRandomVector

public static Matrix drawRandomVector(int nrow,
                                      Distribution.Type type)

This method generates a random vector

Parameters:

nrow - the number of elements to be generated

type - the distribution type (a Distribution.Type enum variable)

Returns:

a Matrix instance

getRandom

public static REpiceaRandom getRandom()

This method returns a Random generator.

Returns:

a Random instance

drawRandomVector

public static Matrix drawRandomVector(int nrow,
                                      Distribution.Type type,
                                      java.util.Random random)

This method generates a random vector

Parameters:

nrow - the number of elements to be generated

type - the distribution type (a Distribution enum variable)

random - a Random instance

Returns:

a Matrix instance

performSpecialAdd

public static Matrix performSpecialAdd(Matrix originalMatrix,
                                       Matrix matrixToAdd)

This method performs a special addition in which only the elements different from 0 and 1 are involved. NOTE: this method is used with SAS output.

Parameters:

originalMatrix - the matrix of parameters

matrixToAdd - the matrix of parameter deviates

Returns:

the new parameters in a new Matrix instance

combineMatrices

public static Matrix combineMatrices(Matrix mat1,
                                     Matrix mat2)

This method combines two row vectors of dummy variables. Useful for regressions.

Parameters:

mat1 - the first row vector

mat2 - the second row vector

Returns:

the resulting matrix

getCombinations

public static long getCombinations(int n,
                                   int d)

This method returns the number of combinations.

Parameters:

n - the number of units

d - the number of units drawn in each combination

Returns:

a long