Class RandomDenseTensor

java.lang.Object

org.flag4j.rng.RandomDenseTensor

public class RandomDenseTensor
extends Object

A utility class for generating random dense tensors, matrices, and vectors with customizable distributions.

Features:

• Generate random tensors, matrices, and vectors with real or complex entries.
• Support for configurable uniform and Gaussian distributions.
• Create specialized matrices, such as symmetric, orthogonal, unitary, triangular, and positive definite.

This class relies on the RandomComplex and RandomArray classes for pseudorandom number generation and array population.

Example Usage:

     RandomDenseTensor rtg = new RandomDenseTensor(12345L);

     // Uniformly distributed tensor.
     Tensor tensor = rtg.randomTensor(new Shape(3, 3, 5), 0.0, 10.0);

     // Random unitary matrix.
     CMatrix unitaryMatrix = rtg.randomUnitaryMatrix(5);

     // Random symmetric positive definite matrix.
     Matrix symmPosDef = rtg.randomSymPosDef(12);
 

See Also:

• RandomSparseTensor
• RandomComplex
• RandomArray

Constructor Summary

Constructors

Constructor	Description
RandomDenseTensor()	Constructs a new pseudorandom tensor generator with a seed which is unlikely to be the same as other from any other invocation of this constructor.
RandomDenseTensor(long seed)	Constructs a pseudorandom tensor generator with a specified seed.

Method Summary

Modifier and Type	Method	Description
CMatrix	randnCMatrix(Shape shape)	Generates a matrix filled with pseudorandom values sampled from a bivariate standard Gaussian (normal) distribution with mean zero and standard deviation one along both the real and imaginary axes.
CMatrix	randnCMatrix(Shape shape, double mean, double std)	Generates a matrix filled with pseudorandom values sampled from a bivariate Gaussian (normal) distribution with specified mean and standard deviation along both the real and imaginary axes.
CMatrix	randnCMatrix(Shape shape, double meanRe, double stdRe, double meanIm, double stdIm, double corrCoeff)	Generates a matrix filled with pseudorandom values sampled from a bivariate Gaussian (normal) distribution.
CTensor	randnCTensor(Shape shape)	Generates a tensor filled with pseudorandom values sampled from a bivariate standard Gaussian (normal) distribution with mean zero and standard deviation one along both the real and imaginary axes.
CTensor	randnCTensor(Shape shape, double mean, double std)	Generates a tensor filled with pseudorandom values sampled from a bivariate Gaussian (normal) distribution with specified mean and standard deviation along both the real and imaginary axes.
CTensor	randnCTensor(Shape shape, double meanRe, double stdRe, double meanIm, double stdIm, double corrCoeff)	Generates a tensor filled with pseudorandom values sampled from a bivariate Gaussian (normal) distribution.
CVector	randnCVector(int size)	Generates a vector filled with pseudorandom values sampled from a bivariate standard Gaussian (normal) distribution with mean zero and standard deviation one along both the real and imaginary axes.
CVector	randnCVector(int size, double mean, double std)	Generates a vector filled with pseudorandom values sampled from a bivariate Gaussian (normal) distribution with specified mean and standard deviation along both the real and imaginary axes.
CVector	randnCVector(int size, double meanRe, double stdRe, double meanIm, double stdIm, double corrCoeff)	Generates a vector filled with pseudorandom values sampled from a bivariate Gaussian (normal) distribution.
Matrix	randnMatrix(Shape shape)	Generates a matrix filled with pseudorandom values sampled from a normal distribution with mean of 0.0 and standard deviation of 1.0.
Matrix	randnMatrix(Shape shape, double mean, double std)	Generates a matrix filled with pseudorandom values sampled from a normal distribution with specified mean and standard deviation.
Tensor	randnTensor(Shape shape)	Generates a tensor filled with pseudorandom values sampled from a normal distribution with mean of 0.0 and standard deviation of 1.0.
Tensor	randnTensor(Shape shape, double mean, double std)	Generates a tensor filled with pseudorandom values sampled from a normal distribution with specified mean and standard deviation.
Vector	randnVector(int size)	Generates a vector filled with pseudorandom values sampled from a normal distribution with mean of 0.0 and standard deviation of 1.0.
Vector	randnVector(int size, double mean, double std)	Generates a vector filled with pseudorandom values sampled from a normal distribution with specified mean and standard deviation.
CMatrix	randomCMatrix(Shape shape)	Generates a matrix filled with pseudorandom complex values uniformly distributed in the unit disk centered at the origin of the complex plane.
CMatrix	randomCMatrix(Shape shape, double min, double max)	Generates a matrix filled with pseudorandom complex values uniformly distributed in an annulus (i.e. washer) centered at the origin of the complex plane.
CTensor	randomCTensor(Shape shape)	Generates a tensor filled with pseudorandom complex values uniformly distributed in the unit disk centered at the origin of the complex plane.
CTensor	randomCTensor(Shape shape, double min, double max)	Generates a tensor filled with pseudorandom complex values uniformly distributed in an annulus (i.e. washer) centered at the origin of the complex plane.
CVector	randomCVector(int size)	Generates a vector filled with pseudorandom complex values uniformly distributed in the unit disk centered at the origin of the complex plane.
CVector	randomCVector(int size, double min, double max)	Generates a vector filled with pseudorandom complex values uniformly distributed in an annulus (i.e. washer) centered at the origin of the complex plane.
Matrix	randomDiagMatrix(int size, int min, int max)	Generates a random diagonal matrix whose diagonal entries are uniformly distributed in [min, max).
CMatrix	randomHermPosDefMatrix(int size)	Generates a pseudorandom symmetric positive definite matrix.
Matrix	randomMatrix(Shape shape)	Generates a matrix filled with pseudorandom values uniformly distributed in [0, 1).
Matrix	randomMatrix(Shape shape, double min, double max)	Generates a matrix filled with pseudorandom values uniformly distributed in [min, max).
Matrix	randomOrthogonalMatrix(int size)	Gets a pseudorandom orthogonal matrix.
Matrix	randomSymmetricMatrix(int size)	Generates a symmetric matrix filled with pseudorandom values uniformly distributed in [0, 1).
Matrix	randomSymmPosDefMatrix(int size)	Generates a pseudorandom symmetric positive definite matrix.
Tensor	randomTensor(Shape shape)	Generates a tensor filled with pseudorandom values uniformly distributed in [0, 1).
Tensor	randomTensor(Shape shape, double min, double max)	Generates a tensor filled with pseudorandom values uniformly distributed in [min, max).
Matrix	randomTrilMatrix(int size, int min, int max)	Gets a pseudorandom lower triangular matrix of the specified size.
Matrix	randomTriuMatrix(int size, int min, int max)	Gets a pseudorandom upper triangular matrix of the specified size.
CMatrix	randomUnitaryMatrix(int size)	Gets a pseudorandom unitary matrix.
Vector	randomVector(int size)	Generates a vector filled with pseudorandom values uniformly distributed in [0, 1).
Vector	randomVector(int size, double min, double max)	Generates a vector filled with pseudorandom values uniformly distributed in [min, max).

Methods inherited from class java.lang.Object

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

Constructor Details

RandomDenseTensor

public RandomDenseTensor()

Constructs a new pseudorandom tensor generator with a seed which is unlikely to be the same as other from any other invocation of this constructor.

RandomDenseTensor

public RandomDenseTensor(long seed)

Constructs a pseudorandom tensor generator with a specified seed. Use this constructor for reproducible results.

Parameters:

seed - Seed of the pseudorandom tensor generator.

Method Details

randomTensor

public Tensor randomTensor(Shape shape)

Generates a tensor filled with pseudorandom values uniformly distributed in [0, 1).

Parameters:

shape - Shape of the tensor.

Returns:

A tensor filled with pseudorandom values uniformly distributed in [0, 1).

randomTensor

public Tensor randomTensor(Shape shape,
 double min,
 double max)

Generates a tensor filled with pseudorandom values uniformly distributed in [min, max).

Parameters:

shape - Shape of the tensor.

min - Minimum value for the uniform distribution.

max - Maximum value for the uniform distribution.

Returns:

A tensor filled with pseudorandom values uniformly distributed in [min, max).

Throws:

IllegalArgumentException - If max is less than min.

randnTensor

public Tensor randnTensor(Shape shape)

Generates a tensor filled with pseudorandom values sampled from a normal distribution with mean of 0.0 and standard deviation of 1.0.

Parameters:

shape - Shape of the tensor.

Returns:

A tensor filled with pseudorandom values sampled from a normal distribution with mean of 0.0 and standard deviation of 1.0.

randnTensor

public Tensor randnTensor(Shape shape,
 double mean,
 double std)

Generates a tensor filled with pseudorandom values sampled from a normal distribution with specified mean and standard deviation.

Parameters:

shape - Shape of the tensor.

mean - Mean of the normal distribution to sample from.

std - Standard deviation of normal distribution to sample from.

Returns:

A tensor filled with pseudorandom values sampled from a normal distribution with specified mean and standard deviation.

Throws:

IllegalArgumentException - If standard deviation is negative.

randomCTensor

public CTensor randomCTensor(Shape shape)

Generates a tensor filled with pseudorandom complex values uniformly distributed in the unit disk centered at the origin of the complex plane.

Parameters:

shape - Shape of the tensor.

Returns:

A tensor filled with pseudorandom complex values uniformly distributed in the unit disk centered at the origin of the complex plane.

randomCTensor

public CTensor randomCTensor(Shape shape,
 double min,
 double max)

Generates a tensor filled with pseudorandom complex values uniformly distributed in an annulus (i.e. washer) centered at the origin of the complex plane.

Parameters:

shape - Shape of the tensor.

min - Inner radius of annulus.

max - Outer radius of annulus.

Returns:

A tensor filled with pseudorandom values with magnitudes uniformly distributed in an annulus (i.e. washer) centered at the origin of the complex plane.

Throws:

IllegalArgumentException - If min >= max or min < 0.

randnCTensor

public CTensor randnCTensor(Shape shape)

Generates a tensor filled with pseudorandom values sampled from a bivariate standard Gaussian (normal) distribution with mean zero and standard deviation one along both the real and imaginary axes. The correlation coefficient of the distribution will be zero.

Parameters:

shape - Shape of the tensor.

Returns:

A pseudorandom values sampled from a bivariate standard Gaussian (normal) distribution with a correlation coefficient of zero.

See Also:

• RandomComplex.randnComplex128()

randnCTensor

public CTensor randnCTensor(Shape shape,
 double mean,
 double std)

Generates a tensor filled with pseudorandom values sampled from a bivariate Gaussian (normal) distribution with specified mean and standard deviation along both the real and imaginary axes. The correlation coefficient of the distribution will be zero.

Parameters:

shape - Shape of the tensor.

mean - Mean of the Gaussian distribution.

std - Standard deviation of the Gaussian distribution.

Returns:

A tensor filled with pseudorandom values sampled from a bivariate Gaussian distribution with specified mean and standard deviation.

See Also:

• RandomComplex.randnComplex128(double, double)

randnCTensor

public CTensor randnCTensor(Shape shape,
 double meanRe,
 double stdRe,
 double meanIm,
 double stdIm,
 double corrCoeff)

Generates a tensor filled with pseudorandom values sampled from a bivariate Gaussian (normal) distribution.

Parameters:

shape - Shape of the tensor.

meanRe - Mean along real axis of complex plane for the distribution.

stdRe - Standard deviation along real axis of complex plane for the distribution.

meanIm - Mean along imaginary axis of complex plane for the distribution.

stdIm - Standard deviation along imaginary axis of complex plane for the distribution.

corrCoeff - Correlation coefficient of the bivariate distribution.

rng - Pseudorandom number generator to use when randomly sampling from this distribution.

Returns:

A tensor filled with pseudorandom values sampled from a bivariate Gaussian distribution with specified means, standard deviations and correlation coefficient.

See Also:

• RandomComplex.randnComplex128(double, double, double, double, double)

randomVector

public Vector randomVector(int size)

Generates a vector filled with pseudorandom values uniformly distributed in [0, 1).

Parameters:

int - size of the vector.

Returns:

A vector filled with pseudorandom values uniformly distributed in [0, 1).

randomVector

public Vector randomVector(int size,
 double min,
 double max)

Generates a vector filled with pseudorandom values uniformly distributed in [min, max).

Parameters:

min - Minimum value for the uniform distribution.

max - Maximum value for the uniform distribution.

int - size of the vector.

Returns:

A vector filled with pseudorandom values uniformly distributed in [min, max).

Throws:

IllegalArgumentException - If max is less than min.

randnVector

public Vector randnVector(int size)

Generates a vector filled with pseudorandom values sampled from a normal distribution with mean of 0.0 and standard deviation of 1.0.

Parameters:

int - size of the vector.

Returns:

A vector filled with pseudorandom values sampled from a normal distribution with mean of 0.0 and standard deviation of 1.0.

randnVector

public Vector randnVector(int size,
 double mean,
 double std)

Generates a vector filled with pseudorandom values sampled from a normal distribution with specified mean and standard deviation.

Parameters:

mean - Mean of the normal distribution to sample from.

std - Standard deviation of normal distribution to sample from.

int - size of the vector.

Returns:

A vector filled with pseudorandom values sampled from a normal distribution with specified mean and standard deviation.

Throws:

IllegalArgumentException - If standard deviation is negative.

randomCVector

public CVector randomCVector(int size)

Generates a vector filled with pseudorandom complex values uniformly distributed in the unit disk centered at the origin of the complex plane.

Parameters:

int - size of the vector.

Returns:

A vector filled with pseudorandom complex values uniformly distributed in the unit disk centered at the origin of the complex plane.

randomCVector

public CVector randomCVector(int size,
 double min,
 double max)

Generates a vector filled with pseudorandom complex values uniformly distributed in an annulus (i.e. washer) centered at the origin of the complex plane.

Parameters:

min - Inner radius of annulus.

max - Outer radius of annulus.

int - size of the vector.

Returns:

A vector filled with pseudorandom values with magnitudes uniformly distributed in an annulus (i.e. washer) centered at the origin of the complex plane.

Throws:

IllegalArgumentException - If min >= max or min < 0.

randnCVector

public CVector randnCVector(int size)

Generates a vector filled with pseudorandom values sampled from a bivariate standard Gaussian (normal) distribution with mean zero and standard deviation one along both the real and imaginary axes. The correlation coefficient of the distribution will be zero.

Parameters:

int - size of the vector.

Returns:

A pseudorandom values sampled from a bivariate standard Gaussian (normal) distribution with a correlation coefficient of zero.

See Also:

• RandomComplex.randnComplex128()

randnCVector

public CVector randnCVector(int size,
 double mean,
 double std)

Generates a vector filled with pseudorandom values sampled from a bivariate Gaussian (normal) distribution with specified mean and standard deviation along both the real and imaginary axes. The correlation coefficient of the distribution will be zero.

Parameters:

mean - Mean of the Gaussian distribution.

std - Standard deviation of the Gaussian distribution.

int - size of the vector.

Returns:

A vector filled with pseudorandom values sampled from a bivariate Gaussian distribution with specified mean and standard deviation.

See Also:

• RandomComplex.randnComplex128(double, double)

randnCVector

public CVector randnCVector(int size,
 double meanRe,
 double stdRe,
 double meanIm,
 double stdIm,
 double corrCoeff)

Generates a vector filled with pseudorandom values sampled from a bivariate Gaussian (normal) distribution.

Parameters:

meanRe - Mean along real axis of complex plane for the distribution.

stdRe - Standard deviation along real axis of complex plane for the distribution.

meanIm - Mean along imaginary axis of complex plane for the distribution.

stdIm - Standard deviation along imaginary axis of complex plane for the distribution.

corrCoeff - Correlation coefficient of the bivariate distribution.

int - size of the vector.

rng - Pseudorandom number generator to use when randomly sampling from this distribution.

Returns:

A vector filled with pseudorandom values sampled from a bivariate Gaussian distribution with specified means, standard deviations and correlation coefficient.

See Also:

• RandomComplex.randnComplex128(double, double, double, double, double)

randomMatrix

public Matrix randomMatrix(Shape shape)

Generates a matrix filled with pseudorandom values uniformly distributed in [0, 1).

Parameters:

shape - Shape of the matrix.

Returns:

A matrix filled with pseudorandom values uniformly distributed in [0, 1).

randomMatrix

public Matrix randomMatrix(Shape shape,
 double min,
 double max)

Generates a matrix filled with pseudorandom values uniformly distributed in [min, max).

Parameters:

shape - Shape of the matrix.

min - Minimum value for the uniform distribution.

max - Maximum value for the uniform distribution.

Returns:

A matrix filled with pseudorandom values uniformly distributed in [min, max).

Throws:

IllegalArgumentException - If max is less than min.

randnMatrix

public Matrix randnMatrix(Shape shape)

Generates a matrix filled with pseudorandom values sampled from a normal distribution with mean of 0.0 and standard deviation of 1.0.

Parameters:

shape - Shape of the matrix.

Returns:

A matrix filled with pseudorandom values sampled from a normal distribution with mean of 0.0 and standard deviation of 1.0.

randnMatrix

public Matrix randnMatrix(Shape shape,
 double mean,
 double std)

Generates a matrix filled with pseudorandom values sampled from a normal distribution with specified mean and standard deviation.

Parameters:

shape - Shape of the matrix.

mean - Mean of the normal distribution to sample from.

std - Standard deviation of normal distribution to sample from.

Returns:

A matrix filled with pseudorandom values sampled from a normal distribution with specified mean and standard deviation.

Throws:

IllegalArgumentException - If standard deviation is negative.

randomCMatrix

public CMatrix randomCMatrix(Shape shape)

Generates a matrix filled with pseudorandom complex values uniformly distributed in the unit disk centered at the origin of the complex plane.

Parameters:

shape - Shape of the matrix.

Returns:

A matrix filled with pseudorandom complex values uniformly distributed in the unit disk centered at the origin of the complex plane.

randomCMatrix

public CMatrix randomCMatrix(Shape shape,
 double min,
 double max)

Generates a matrix filled with pseudorandom complex values uniformly distributed in an annulus (i.e. washer) centered at the origin of the complex plane.

Parameters:

shape - Shape of the matrix.

min - Inner radius of annulus.

max - Outer radius of annulus.

Returns:

A matrix filled with pseudorandom values with magnitudes uniformly distributed in an annulus (i.e. washer) centered at the origin of the complex plane.

Throws:

IllegalArgumentException - If min >= max or min < 0.

randnCMatrix

public CMatrix randnCMatrix(Shape shape)

Generates a matrix filled with pseudorandom values sampled from a bivariate standard Gaussian (normal) distribution with mean zero and standard deviation one along both the real and imaginary axes. The correlation coefficient of the distribution will be zero.

Parameters:

shape - Shape of the matrix.

Returns:

A pseudorandom values sampled from a bivariate standard Gaussian (normal) distribution with a correlation coefficient of zero.

See Also:

• RandomComplex.randnComplex128()

randnCMatrix

public CMatrix randnCMatrix(Shape shape,
 double mean,
 double std)

Generates a matrix filled with pseudorandom values sampled from a bivariate Gaussian (normal) distribution with specified mean and standard deviation along both the real and imaginary axes. The correlation coefficient of the distribution will be zero.

Parameters:

shape - Shape of the matrix.

mean - Mean of the Gaussian distribution.

std - Standard deviation of the Gaussian distribution.

Returns:

A matrix filled with pseudorandom values sampled from a bivariate Gaussian distribution with specified mean and standard deviation.

See Also:

• RandomComplex.randnComplex128(double, double)

randnCMatrix

public CMatrix randnCMatrix(Shape shape,
 double meanRe,
 double stdRe,
 double meanIm,
 double stdIm,
 double corrCoeff)

Generates a matrix filled with pseudorandom values sampled from a bivariate Gaussian (normal) distribution.

Parameters:

shape - Shape of the matrix.

meanRe - Mean along real axis of complex plane for the distribution.

stdRe - Standard deviation along real axis of complex plane for the distribution.

meanIm - Mean along imaginary axis of complex plane for the distribution.

stdIm - Standard deviation along imaginary axis of complex plane for the distribution.

corrCoeff - Correlation coefficient of the bivariate distribution.

rng - Pseudorandom number generator to use when randomly sampling from this distribution.

Returns:

A matrix filled with pseudorandom values sampled from a bivariate Gaussian distribution with specified means, standard deviations and correlation coefficient.

See Also:

• RandomComplex.randnComplex128(double, double, double, double, double)

randomSymmetricMatrix

public Matrix randomSymmetricMatrix(int size)

Generates a symmetric matrix filled with pseudorandom values uniformly distributed in [0, 1).

Parameters:

size - Number of rows and columns in the resulting matrix (the result will be a square matrix).

Returns:

A symmetric matrix filled with pseudorandom values uniformly distributed in [0, 1).

randomOrthogonalMatrix

public Matrix randomOrthogonalMatrix(int size)

Gets a pseudorandom orthogonal matrix.

The matrix is generated as if by randomMatrix(new Shape(size, size)). Then, a QR decomposition is computed on this pseudorandom matrix and the Q matrix from this decomposition is returned.

Parameters:

size - Size of the orthogonal matrix.

Returns:

A pseudorandom orthogonal matrix.

randomUnitaryMatrix

public CMatrix randomUnitaryMatrix(int size)

Gets a pseudorandom unitary matrix.

The matrix is generated as if by randomCMatrix(new Shape(size, size)). Then, a QR decomposition is computed on this pseudorandom matrix and the Q matrix from this decomposition is returned.

Parameters:

size - Size of the unitary matrix.

Returns:

A pseudorandom unitary matrix.

randomTriuMatrix

public Matrix randomTriuMatrix(int size,
 int min,
 int max)

Gets a pseudorandom upper triangular matrix of the specified size. The non-zero entries of the matrix are distributed according to a uniform distribution un [min, max).

Parameters:

size - Size if the upper triangular matrix.

min - Lower bound of the uniform distribution (inclusive).

max - Upper bound of the uniform distribution (exclusive).

Returns:

A pseudorandom upper triangular matrix of the specified size.

randomTrilMatrix

public Matrix randomTrilMatrix(int size,
 int min,
 int max)

Gets a pseudorandom lower triangular matrix of the specified size. The non-zero entries of the matrix are distributed according to a uniform distribution un [min, max).

Parameters:

size - Size if the lower triangular matrix.

min - Lower bound of the uniform distribution (inclusive).

max - Upper bound of the uniform distribution (exclusive).

Returns:

A pseudorandom lower triangular matrix of the specified size.

randomDiagMatrix

public Matrix randomDiagMatrix(int size,
 int min,
 int max)

Generates a random diagonal matrix whose diagonal entries are uniformly distributed in [min, max).

Parameters:

size - Size of the diagonal matrix to construct.

min - Lower bound of the uniform distribution (inclusive).

max - Upper bound of the uniform distribution (exclusive).

Returns:

A diagonal matrix with diagonal entries are uniformly distributed in [min, max).

randomSymmPosDefMatrix

public Matrix randomSymmPosDefMatrix(int size)

Generates a pseudorandom symmetric positive definite matrix. This is done as if by

     Matrix D = randomDiagMatrix(size, 0, 1);
     Matrix Q = randomOrthogonalMatrix(size);
     return Q.T().mult(D).mult(Q);

Parameters:

size - Size of the symmetric positive definite matrix to generate.

Returns:

A pseudorandom symmetric positive definite matrix.

See Also:

• randomHermPosDefMatrix(int)

randomHermPosDefMatrix

public CMatrix randomHermPosDefMatrix(int size)

Generates a pseudorandom symmetric positive definite matrix. This is done as if by

     Matrix D = randomDiagMatrix(size, 0, 1);
     CMatrix U = randomUnitaryMatrix(size);
     return U.H().mult(D).mult(U);

Parameters:

size - Size of the symmetric positive definite matrix to generate.

Returns:

A pseudorandom symmetric positive definite matrix.

See Also:

• randomSymmPosDefMatrix(int)