de.citec.ml.rng

Class CheckFunctions

java.lang.Object

de.citec.ml.rng.CheckFunctions

public final class CheckFunctions
extends Object

This is a class for consistency checks on input parameters.

Author:

Benjamin Paassen - bpaassen(at)techfak.uni-bielefeld.de

Method Summary

Modifier and Type	Method and Description
static IllegalArgumentException	checkAssignmentsVector(int N, int K, int[] assignments) Checks if the given assignment vector does indeed only map to entries between 0 and C, where C is the number of clusters/prototypes.
static IllegalArgumentException	checkConvexCoefficients(int m, double[] alpha) Checks if the given vector has the correct size (N), is non-negative and sums up to 1.
static IllegalArgumentException	checkConvexCoefficients(int m, int K, double[][] Alpha) Checks if the given matrix contains proper convex coefficient vectors.
static IllegalArgumentException	checkDissimilaritiesOfDatapointsToPrototypes(double[][] Dp) Checks if the given matrix is not empty and consistent.
static IllegalArgumentException	checkDissimilarityMatrix(double[][] D) Checks if the given matrix is quadratic, symmetric and reflexive.

Methods inherited from class java.lang.Object

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

Method Detail

checkDissimilarityMatrix

public static IllegalArgumentException checkDissimilarityMatrix(double[][] D)

Checks if the given matrix is quadratic, symmetric and reflexive. If it does not meet one of the criteria, this methods returns an exception accordingly. If all criteria are met, this method returns null.

Parameters:

D - supposedly a distance matrix.

Returns:

null if the matrix is a proper distance matrix, and an exception if it is not.

checkConvexCoefficients

public static IllegalArgumentException checkConvexCoefficients(int m,
                                                               int K,
                                                               double[][] Alpha)

Checks if the given matrix contains proper convex coefficient vectors.

Parameters:

m - the expected length of the convex combination.

K - the expected number of clusters/prototypes.

Alpha - supposedly convex combinations.

Returns:

null if Alpha contains proper convex combinations, and an exception if it does not.

checkConvexCoefficients

public static IllegalArgumentException checkConvexCoefficients(int m,
                                                               double[] alpha)

Checks if the given vector has the correct size (N), is non-negative and sums up to 1.

Parameters:

m - the expected length of the convex combination.

alpha - supposedly a convex combination.

Returns:

null if alpha is a proper convex combination, and an exception if it is not.

checkAssignmentsVector

public static IllegalArgumentException checkAssignmentsVector(int N,
                                                              int K,
                                                              int[] assignments)

Checks if the given assignment vector does indeed only map to entries between 0 and C, where C is the number of clusters/prototypes.

Parameters:

N - the expected length of the assignments vector.

K - the expected number of clusters/prototypes.

assignments - supposedly an assignments vector.

Returns:

null if the input is a proper assignments vector, and an exception if it is not.

checkDissimilaritiesOfDatapointsToPrototypes

public static IllegalArgumentException checkDissimilaritiesOfDatapointsToPrototypes(double[][] Dp)

Checks if the given matrix is not empty and consistent.

Parameters:

Dp - a matrix of distances from data points to prototypes.

Returns:

null if the input is a proper matrix and an exception if it is not.