Package org.flag4j.arrays.sparse

Class CooTensor

java.lang.Object

org.flag4j.core.TensorBase<CooTensor,Tensor,CooCTensor,CTensor,CooTensor,double[],Double>

org.flag4j.core.sparse_base.SparseTensorBase<CooTensor,Tensor,CooCTensor,CTensor,CooTensor,double[],Double>

org.flag4j.core.sparse_base.RealSparseTensorBase<CooTensor,Tensor,CooCTensor,CTensor>

org.flag4j.arrays.sparse.CooTensor

All Implemented Interfaces:

Serializable, RealTensorMixin<CooTensor,CooCTensor>, SparseTensorMixin, TensorComparisonsMixin, TensorExclusiveMixin<CooTensor,Tensor,CooTensor,CooCTensor>, TensorManipulationsMixin<CooTensor>, TensorOperationsMixin<CooTensor,Tensor,CooCTensor,CTensor,CooTensor,Double>, TensorPropertiesMixin

public class CooTensor
extends RealSparseTensorBase<CooTensor,Tensor,CooCTensor,CTensor>
implements TensorExclusiveMixin<CooTensor,Tensor,CooTensor,CooCTensor>

Real sparse tensor. Can be any rank. Stored in coordinate (COO) format.

See Also:

• Serialized Form

Field Summary

Fields inherited from class org.flag4j.core.sparse_base.SparseTensorBase

indices, nnz

Fields inherited from class org.flag4j.core.TensorBase

DEFAULT_ROUND_TO_ZERO_THRESHOLD, entries, shape

Constructor Summary

Constructors

Constructor	Description
CooTensor(CooTensor A)	Constructs a sparse tensor whose shape and non-zero values/indices are given by another sparse tensor.
CooTensor(Shape shape)	Creates a sparse tensor with specified shape filled with zeros.
CooTensor(Shape shape, double[] nonZeroEntries, int[][] indices)	Creates a sparse tensor with specified shape and non-zero values/indices.
CooTensor(Shape shape, int[] nonZeroEntries, int[][] indices)	Creates a sparse tensor with specified shape and non-zero values/indices.
CooTensor(Shape shape, List<Double> nonZeroEntries, List<int[]> indices)	Creates a sparse tensor with specified shape and non-zero values/indices.

Method Summary

Modifier and Type	Method	Description
Tensor	add(double a)	Adds specified value to all entries of this tensor.
CTensor	add(CTensor B)	Computes the element-wise addition between two tensors of the same rank.
Tensor	add(Tensor B)	Computes the element-wise addition between two tensors of the same rank.
CooCTensor	add(CooCTensor B)	Computes the element-wise addition between two tensors of the same rank.
CooTensor	add(CooTensor B)	Computes the element-wise addition between two tensors of the same rank.
CTensor	add(CNumber a)	Adds specified value to all entries of this tensor.
boolean	allClose(CooTensor tensor, double relTol, double absTol)	Checks if all entries of this tensor are close to the entries of the argument tensor.
CooTensor	copy()	Creates a copy of this tensor.
CooCTensor	elemDiv(CTensor B)	Computes the element-wise division between two tensors.
CooTensor	elemDiv(Tensor B)	Computes the element-wise division between two tensors.
CTensor	elemMult(CTensor B)	Computes the element-wise multiplication between two tensors.
CooTensor	elemMult(Tensor B)	Computes the element-wise multiplication between two tensors.
CooCTensor	elemMult(CooCTensor B)	Computes the element-wise multiplication between two tensors.
CooTensor	elemMult(CooTensor B)	Computes the element-wise multiplication between two tensors.
boolean	equals(Object object)	Checks if an object is equal to this sparse COO tensor.
CooTensor	flatten()	Flattens tensor to single dimension.
CooTensor	flatten(int axis)	Flattens a tensor along the specified axis.
static CooTensor	fromDense(Tensor src)	Converts a sparse CooTensor from a dense Tensor.
Double	get(int... indices)	Gets the element in this tensor at the specified indices.
protected CooTensor	getSelf()	Simply returns a reference of this tensor.
protected CooCTensor	makeComplexTensor(Shape shape, CNumber[] entries, int[][] indices)	A factory for creating a complex sparse tensor.
protected Tensor	makeDenseTensor(Shape shape, double[] entries)	A factory for creating a real dense tensor.
protected CooTensor	makeTensor(Shape shape, double[] entries, int[][] indices)	A factory for creating a real sparse tensor.
CooTensor	reshape(Shape newShape)	Copies and reshapes tensor if possible.
CooTensor	set(double value, int... indices)	Sets an index of this tensor to a specified value.
Tensor	sub(double a)	Adds specified value to all entries of this tensor.
CTensor	sub(CTensor B)	Computes the element-wise subtraction between two tensors of the same rank.
Tensor	sub(Tensor B)	Computes the element-wise addition between two tensors of the same rank.
CooCTensor	sub(CooCTensor B)	Computes the element-wise subtraction between two tensors of the same rank.
CooTensor	sub(CooTensor B)	Computes the element-wise subtraction between two tensors of the same rank.
CTensor	sub(CNumber a)	Subtracts a specified value from all entries of this tensor.
CooTensor	T()	Computes the transpose of a tensor.
CooTensor	T(int... axes)	Computes the transpose of this tensor.
CooTensor	T(int axis1, int axis2)	Computes the transpose of a tensor.
Tensor	tensorDot(CooTensor src2, int[] aAxes, int[] bAxes)	Computes the tensor contraction of this tensor with a specified tensor over the specified set of axes.
Tensor	tensorInv(int numIndices)	Computes the 'inverse' of this tensor.
CooCTensor	toComplex()	Converts this tensor to an equivalent complex tensor.
Tensor	toDense()	Converts this sparse tensor to an equivalent dense tensor.
CooMatrix	toMatrix(Shape matShape)	Converts this tensor to a matrix with the specified shape.
CooTensor	transpose()	Computes the transpose of a tensor.

Methods inherited from class org.flag4j.core.sparse_base.RealSparseTensorBase

abs, argMax, argMin, div, div, isNeg, isOnes, isPos, isZeros, max, maxAbs, min, minAbs, mult, mult, nonZeroEntries, recip, reshape, round, round, roundToZero, roundToZero, sortIndices, sqrt, sum

Methods inherited from class org.flag4j.core.sparse_base.SparseTensorBase

density, sparsity

Methods inherited from class org.flag4j.core.TensorBase

allClose, getEntries, getRank, getShape, sameLength, sameShape, tensorEquals, totalEntries

Methods inherited from class java.lang.Object

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

Methods inherited from interface org.flag4j.core.TensorExclusiveMixin

getRank, tensorDot, tensorDot, tensorDot, tensorInv, transpose, transpose

Constructor Details

CooTensor

public CooTensor(Shape shape)

Creates a sparse tensor with specified shape filled with zeros.

Parameters:

shape - Shape of the tensor.

CooTensor

public CooTensor(Shape shape,
 double[] nonZeroEntries,
 int[][] indices)

Creates a sparse tensor with specified shape and non-zero values/indices.

Parameters:

shape - Shape of the tensor.

nonZeroEntries - Non-zero entries of the tensor.

indices - Indices of the non-zero entries of the tensor.

CooTensor

public CooTensor(Shape shape,
 int[] nonZeroEntries,
 int[][] indices)

Creates a sparse tensor with specified shape and non-zero values/indices.

Parameters:

shape - Shape of the tensor.

nonZeroEntries - Non-zero entries of the tensor.

indices - Indices of the non-zero entries of the tensor.

CooTensor

public CooTensor(Shape shape,
 List<Double> nonZeroEntries,
 List<int[]> indices)

Creates a sparse tensor with specified shape and non-zero values/indices.

Parameters:

shape - Shape of the tensor.

nonZeroEntries - Non-zero entries of the tensor.

indices - Indices of the non-zero entries of the tensor.

CooTensor

public CooTensor(CooTensor A)

Constructs a sparse tensor whose shape and non-zero values/indices are given by another sparse tensor. This effectively copies the tensor.

Parameters:

A - Tensor to copy.

Method Details

equals

public boolean equals(Object object)

Checks if an object is equal to this sparse COO tensor.

Specified by:

equals in interface TensorComparisonsMixin

Overrides:

equals in class Object

Parameters:

object - Object to compare this sparse COO tensor to.

Returns:

True if the object is a CooTensor, has the same shape as this tensor, and is element-wise equal to this tensor.

toComplex

public CooCTensor toComplex()

Converts this tensor to an equivalent complex tensor. That is, the entries of the resultant matrix will be exactly the same value but will have type CNumber rather than Double.

Specified by:

toComplex in interface RealTensorMixin<CooTensor,CooCTensor>

Returns:

A complex matrix which is equivalent to this matrix.

fromDense

public static CooTensor fromDense(Tensor src)

Converts a sparse CooTensor from a dense Tensor. This is likely only worthwhile for very sparse tensors.

Parameters:

src - Dense tensor to convert to sparse COO tensor.

Returns:

A COO tensor which is equivalent to the src dense tensor.

makeTensor

protected CooTensor makeTensor(Shape shape,
 double[] entries,
 int[][] indices)

A factory for creating a real sparse tensor.

Specified by:

makeTensor in class RealSparseTensorBase<CooTensor,Tensor,CooCTensor,CTensor>

Parameters:

shape - Shape of the sparse tensor to make.

entries - Non-zero entries of the sparse tensor to make.

indices - Non-zero indices of the sparse tensor to make.

Returns:

A tensor created from the specified parameters.

makeDenseTensor

protected Tensor makeDenseTensor(Shape shape,
 double[] entries)

A factory for creating a real dense tensor.

Specified by:

makeDenseTensor in class RealSparseTensorBase<CooTensor,Tensor,CooCTensor,CTensor>

Parameters:

shape - Shape of the tensor to make.

entries - Entries of the dense tensor to make.

Returns:

A tensor created from the specified parameters.

makeComplexTensor

protected CooCTensor makeComplexTensor(Shape shape,
 CNumber[] entries,
 int[][] indices)

A factory for creating a complex sparse tensor.

Specified by:

makeComplexTensor in class RealSparseTensorBase<CooTensor,Tensor,CooCTensor,CTensor>

Parameters:

shape - Shape of the tensor to make.

entries - Non-zero entries of the sparse tensor to make.

indices - Non-zero indices of the sparse tensor to make.

Returns:

A tensor created from the specified parameters.

set

public CooTensor set(double value,
 int... indices)

Sets an index of this tensor to a specified value.

Specified by:

set in interface TensorManipulationsMixin<CooTensor>

Parameters:

value - Value to set.

indices - The indices of this tensor for which to set the value.

Returns:

A reference to this tensor.

reshape

public CooTensor reshape(Shape newShape)

Copies and reshapes tensor if possible. The total number of entries in this tensor must match the total number of entries in the reshaped tensor.

Specified by:

reshape in interface TensorManipulationsMixin<CooTensor>

Parameters:

newShape - Shape of the new tensor.

Returns:

A tensor which is equivalent to this tensor but with the specified shape.

Throws:

IllegalArgumentException - If this tensor cannot be reshaped to the specified dimensions.

flatten

public CooTensor flatten()

Flattens tensor to single dimension. To flatten tensor along a single axis see flatten(int).

Specified by:

flatten in interface TensorManipulationsMixin<CooTensor>

Returns:

The flattened tensor.

See Also:

• flatten(int)

tensorDot

public Tensor tensorDot(CooTensor src2,
 int[] aAxes,
 int[] bAxes)

Computes the tensor contraction of this tensor with a specified tensor over the specified set of axes. That is, computes the sum of products between the two tensors along the specified set of axes.

Specified by:

tensorDot in interface TensorExclusiveMixin<CooTensor,Tensor,CooTensor,CooCTensor>

Parameters:

src2 - Tensor to contract with this tensor.

aAxes - Axes along which to compute products for this tensor.

bAxes - Axes along which to compute products for src2 tensor.

Returns:

The tensor dot product over the specified axes.

Throws:

IllegalArgumentException - If the two tensors shapes do not match along the specified axes pairwise in aAxes and bAxes.

IllegalArgumentException - If aAxes and bAxes do not match in length, or if any of the axes are out of bounds for the corresponding tensor.

T

public CooTensor T(int axis1,
 int axis2)

Computes the transpose of a tensor. Same as TensorExclusiveMixin.transpose(int, int). In the context of a tensor, this exchanges the specified axes. Also see transpose() and T() to exchange first and last axes.

Specified by:

T in interface TensorExclusiveMixin<CooTensor,Tensor,CooTensor,CooCTensor>

Parameters:

axis1 - First axis to exchange.

axis2 - Second axis to exchange.

Returns:

The transpose of this tensor.

T

public CooTensor T(int... axes)

Computes the transpose of this tensor. That is, interchanges the axes of this tensor so that it matches the specified axes permutation. Same as TensorExclusiveMixin.transpose(int[]).

Specified by:

T in interface TensorExclusiveMixin<CooTensor,Tensor,CooTensor,CooCTensor>

Parameters:

axes - Permutation of tensor axis. If the tensor has rank N, then this must be an array of length N which is a permutation of {0, 1, 2, ..., N-1}.

Returns:

The transpose of this tensor with its axes permuted by the axes array.

Throws:

IllegalArgumentException - If axes is not a permutation of {1, 2, 3, ... N-1}.

add

public CooTensor add(CooTensor B)

Computes the element-wise addition between two tensors of the same rank.

Specified by:

add in interface TensorExclusiveMixin<CooTensor,Tensor,CooTensor,CooCTensor>

Specified by:

add in interface TensorOperationsMixin<CooTensor,Tensor,CooCTensor,CTensor,CooTensor,Double>

Parameters:

B - Second tensor in the addition.

Returns:

The result of adding the tensor B to this tensor element-wise.

Throws:

IllegalArgumentException - If this tensor and B have different shapes.

add

public Tensor add(Tensor B)

Computes the element-wise addition between two tensors of the same rank.

Specified by:

add in interface TensorExclusiveMixin<CooTensor,Tensor,CooTensor,CooCTensor>

Parameters:

B - Second tensor in the addition.

Returns:

The result of adding the tensor B to this tensor element-wise.

Throws:

IllegalArgumentException - If this tensor and B have different shapes.

sub

public Tensor sub(Tensor B)

Computes the element-wise addition between two tensors of the same rank.

Specified by:

sub in interface TensorExclusiveMixin<CooTensor,Tensor,CooTensor,CooCTensor>

Parameters:

B - Second tensor in the addition.

Returns:

The result of adding the tensor B to this tensor element-wise.

Throws:

IllegalArgumentException - If this tensor and B have different shapes.

add

public CTensor add(CTensor B)

Computes the element-wise addition between two tensors of the same rank.

Specified by:

add in interface TensorExclusiveMixin<CooTensor,Tensor,CooTensor,CooCTensor>

Parameters:

B - Second tensor in the addition.

Returns:

The result of adding the tensor B to this tensor element-wise.

Throws:

IllegalArgumentException - If this tensor and B have different shapes.

add

public CooCTensor add(CooCTensor B)

Computes the element-wise addition between two tensors of the same rank.

Specified by:

add in interface TensorExclusiveMixin<CooTensor,Tensor,CooTensor,CooCTensor>

Parameters:

B - Second tensor in the addition.

Returns:

The result of adding the tensor B to this tensor element-wise.

Throws:

IllegalArgumentException - If this tensor and B have different shapes.

add

public Tensor add(double a)

Adds specified value to all entries of this tensor.

Specified by:

add in interface TensorOperationsMixin<CooTensor,Tensor,CooCTensor,CTensor,CooTensor,Double>

Parameters:

a - Value to add to all entries of this tensor.

Returns:

The result of adding the specified value to each entry of this tensor.

add

public CTensor add(CNumber a)

Adds specified value to all entries of this tensor.

Specified by:

add in interface TensorOperationsMixin<CooTensor,Tensor,CooCTensor,CTensor,CooTensor,Double>

Parameters:

a - Value to add to all entries of this tensor.

Returns:

The result of adding the specified value to each entry of this tensor.

sub

public CooTensor sub(CooTensor B)

Computes the element-wise subtraction between two tensors of the same rank.

Specified by:

sub in interface TensorExclusiveMixin<CooTensor,Tensor,CooTensor,CooCTensor>

Specified by:

sub in interface TensorOperationsMixin<CooTensor,Tensor,CooCTensor,CTensor,CooTensor,Double>

Parameters:

B - Second tensor in element-wise subtraction.

Returns:

The result of subtracting the tensor B from this tensor element-wise.

Throws:

IllegalArgumentException - If this tensor and B have different shapes.

sub

public CTensor sub(CTensor B)

Computes the element-wise subtraction between two tensors of the same rank.

Specified by:

sub in interface TensorExclusiveMixin<CooTensor,Tensor,CooTensor,CooCTensor>

Parameters:

B - Second tensor in element-wise subtraction.

Returns:

The result of subtracting the tensor B from this tensor element-wise.

Throws:

IllegalArgumentException - If this tensor and B have different shapes.

sub

public CooCTensor sub(CooCTensor B)

Computes the element-wise subtraction between two tensors of the same rank.

Specified by:

sub in interface TensorExclusiveMixin<CooTensor,Tensor,CooTensor,CooCTensor>

Parameters:

B - Second tensor in element-wise subtraction.

Returns:

The result of subtracting the tensor B from this tensor element-wise.

Throws:

IllegalArgumentException - If this tensor and B have different shapes.

elemMult

public CooTensor elemMult(Tensor B)

Computes the element-wise multiplication between two tensors.

Specified by:

elemMult in interface TensorExclusiveMixin<CooTensor,Tensor,CooTensor,CooCTensor>

Parameters:

B - Tensor to element-wise multiply to this tensor.

Returns:

The result of the element-wise tensor multiplication.

Throws:

IllegalArgumentException - If the tensors do not have the same shape.

sub

public Tensor sub(double a)

Adds specified value to all entries of this tensor.

Specified by:

sub in interface TensorOperationsMixin<CooTensor,Tensor,CooCTensor,CTensor,CooTensor,Double>

Parameters:

a - Value to add to all entries of this tensor.

Returns:

The result of adding the specified value to each entry of this tensor.

sub

public CTensor sub(CNumber a)

Subtracts a specified value from all entries of this tensor.

Specified by:

sub in interface TensorOperationsMixin<CooTensor,Tensor,CooCTensor,CTensor,CooTensor,Double>

Parameters:

a - Value to subtract from all entries of this tensor.

Returns:

The result of subtracting the specified value from each entry of this tensor.

transpose

public CooTensor transpose()

Computes the transpose of a tensor. Same as T(). In the context of a tensor, this exchanges the first and last axis of the tensor. Also see TensorExclusiveMixin.transpose(int, int) and T(int, int).

Specified by:

transpose in interface TensorOperationsMixin<CooTensor,Tensor,CooCTensor,CTensor,CooTensor,Double>

Returns:

The transpose of this tensor.

See Also:

• T()
• T(int, int)
• T(int...)

T

public CooTensor T()

Computes the transpose of a tensor. Same as transpose(). In the context of a tensor, this exchanges the first and last axis of the tensor.

Specified by:

T in interface TensorOperationsMixin<CooTensor,Tensor,CooCTensor,CTensor,CooTensor,Double>

Returns:

The transpose of this tensor.

See Also:

• transpose()
• T(int, int)
• T(int...)

get

public Double get(int... indices)

Gets the element in this tensor at the specified indices.

Specified by:

get in interface TensorOperationsMixin<CooTensor,Tensor,CooCTensor,CTensor,CooTensor,Double>

Parameters:

indices - Indices of element.

Returns:

The element at the specified indices.

Throws:

IllegalArgumentException - If the number of indices does not match the rank of this tensor.

copy

public CooTensor copy()

Creates a copy of this tensor.

Specified by:

copy in interface TensorOperationsMixin<CooTensor,Tensor,CooCTensor,CTensor,CooTensor,Double>

Returns:

A copy of this tensor.

elemMult

public CooTensor elemMult(CooTensor B)

Computes the element-wise multiplication between two tensors.

Specified by:

elemMult in interface TensorExclusiveMixin<CooTensor,Tensor,CooTensor,CooCTensor>

Specified by:

elemMult in interface TensorOperationsMixin<CooTensor,Tensor,CooCTensor,CTensor,CooTensor,Double>

Parameters:

B - Tensor to element-wise multiply to this tensor.

Returns:

The result of the element-wise tensor multiplication.

Throws:

IllegalArgumentException - If this tensor and B do not have the same shape.

elemMult

public CTensor elemMult(CTensor B)

Computes the element-wise multiplication between two tensors.

Specified by:

elemMult in interface TensorExclusiveMixin<CooTensor,Tensor,CooTensor,CooCTensor>

Parameters:

B - Tensor to element-wise multiply to this tensor.

Returns:

The result of the element-wise tensor multiplication.

Throws:

IllegalArgumentException - If the tensors do not have the same shape.

elemMult

public CooCTensor elemMult(CooCTensor B)

Computes the element-wise multiplication between two tensors.

Specified by:

elemMult in interface TensorExclusiveMixin<CooTensor,Tensor,CooTensor,CooCTensor>

Parameters:

B - Tensor to element-wise multiply to this tensor.

Returns:

The result of the element-wise tensor multiplication.

Throws:

IllegalArgumentException - If the tensors do not have the same shape.

elemDiv

public CooCTensor elemDiv(CTensor B)

Computes the element-wise division between two tensors.

Specified by:

elemDiv in interface TensorExclusiveMixin<CooTensor,Tensor,CooTensor,CooCTensor>

Parameters:

B - Tensor to element-wise divide with this tensor.

Returns:

The result of the element-wise tensor division.

Throws:

IllegalArgumentException - If the tensors do not have the same shape.

elemDiv

public CooTensor elemDiv(Tensor B)

Computes the element-wise division between two tensors.

Specified by:

elemDiv in interface TensorExclusiveMixin<CooTensor,Tensor,CooTensor,CooCTensor>

Specified by:

elemDiv in interface TensorOperationsMixin<CooTensor,Tensor,CooCTensor,CTensor,CooTensor,Double>

Parameters:

B - Tensor to element-wise divide with this tensor.

Returns:

The result of the element-wise tensor multiplication.

Throws:

IllegalArgumentException - If this tensor and B do not have the same shape.

tensorInv

public Tensor tensorInv(int numIndices)

Computes the 'inverse' of this tensor. That is, computes the tensor X=this.tensorInv() such that this.tensorDot(X, numIndices) is the 'identity' tensor for the tensor dot product operation. A tensor I is the identity for a tensor dot product if this.tensorDot(I, numIndices).equals(this).

WARNING: This method will convert this tensor to a dense tensor.

Specified by:

tensorInv in interface TensorExclusiveMixin<CooTensor,Tensor,CooTensor,CooCTensor>

Parameters:

numIndices - The number of first numIndices which are involved in the inverse sum.

Returns:

The 'inverse' of this tensor as defined in the above sense.

See Also:

• TensorExclusiveMixin.tensorInv()

flatten

public CooTensor flatten(int axis)

Flattens a tensor along the specified axis.

Specified by:

flatten in interface TensorManipulationsMixin<CooTensor>

Parameters:

axis - Axis along which to flatten tensor.

Throws:

IllegalArgumentException - If the axis is not positive or larger than this.TensorBase.getRank()-1.

getSelf

protected CooTensor getSelf()

Simply returns a reference of this tensor.

Specified by:

getSelf in class TensorBase<CooTensor,Tensor,CooCTensor,CTensor,CooTensor,double[],Double>

Returns:

A reference to this tensor.

allClose

public boolean allClose(CooTensor tensor,
 double relTol,
 double absTol)

Description copied from class: TensorBase

Checks if all entries of this tensor are close to the entries of the argument tensor.

Specified by:

allClose in class TensorBase<CooTensor,Tensor,CooCTensor,CTensor,CooTensor,double[],Double>

Parameters:

tensor - Tensor to compare this tensor to.

relTol - Relative tolerance.

absTol - Absolute tolerance.

Returns:

True if the argument tensor is the same shape as this tensor and all entries are 'close', i.e. elements a and b at the same positions in the two tensors respectively satisfy |a-b| <= (atol + rtol*|b|). Otherwise, returns false.

See Also:

• TensorBase.allClose(Object)

toMatrix

public CooMatrix toMatrix(Shape matShape)

Converts this tensor to a matrix with the specified shape.

Parameters:

matShape - Shape of the resulting matrix. Must be broadcastable with the shape of this tensor.

Returns:

A matrix of shape matShape with the values of this tensor.

toDense

public Tensor toDense()

Converts this sparse tensor to an equivalent dense tensor.

Specified by:

toDense in class SparseTensorBase<CooTensor,Tensor,CooCTensor,CTensor,CooTensor,double[],Double>

Returns:

A dense tensor which is equivalent to this sparse tensor.