Module flag4j

Package org.flag4j.linalg.ops.sparse.coo

Class CooDataSorter<T>

java.lang.Object

org.flag4j.linalg.ops.sparse.coo.CooDataSorter<T>

public final class CooDataSorter<T>
extends Object

A class which can be used to sort the indices, along with data values, in lexicographical order.

This class is not thread safe.

Specifically, if a sparse tensor has shape (10, 15, 5, 2) and the following indices and non-zero values,

      - Indices: {{4, 1, 2, 0}, {4, 0, 1, 2},
                  {1, 2, 3, 0}, {2, 3, 5, 1},
                  {9, 10, 4, 1}, {1, 2, 1, 1}}
      - Values: {1.1, 2.2, 3.3, 4.4, 5.5, 6.6}
 

then the sorted indices and non-zero values will be,

      - Indices: {{1, 2, 1, 1}, {1, 2, 3, 0},
                  {2, 3, 5, 1}, {4, 0, 1, 2},
                  {4, 1, 2, 0}, {9, 10, 4, 1}}
      - Values: {6.6, 3.3, 4.4, 2.2, 1.1, 5.5}
 

Field Summary

Fields

Modifier and Type	Field	Description
private final List<Integer>[]	keys	Stores wrapped indices.
private final List<T>	values	Non-zero values of the sparse tensor to sort.

Constructor Summary

Constructors

Modifier	Constructor	Description
	CooDataSorter(List<T> values, List<Integer> rowIndices, List<Integer> colIndices)	Creates a sorter which can sort the indices and values of a COO matrix lexicographically.
private	CooDataSorter(T[] values, int[][] indices, boolean transpose)	Wraps the data, and it's associated indices, of a sparse tensor.

Method Summary

Modifier and Type	Method	Description
CooDataSorter	sparseSort()	Sorts the wrapped indices in lexicographical order while maintaining the order of the non-zero values so that their order matches the order of the indices.
private void	sparseSortHelper(int keyIdx, int start, int stop)	Sorts the specified key list over a specified range.
void	unwrap(double[] values, int[] indices)	Unwraps sparse data values and indices.
void	unwrap(double[] values, int[][] indices)	Unwraps sparse data values and indices.
void	unwrap(double[] values, int[] rowIndices, int[] colIndices)	Unwraps sparse data values and indices.
void	unwrap(Object[] values, int[] indices)	Unwraps sparse data values and indices.
void	unwrap(Object[] values, int[][] indices)	Unwraps sparse data values and indices.
void	unwrap(Object[] values, int[] rowIndices, int[] colIndices)	Unwraps sparse data values and indices.
static CooDataSorter	wrap(double[] values, int[] indices)	Factory method which wraps data in an instance of CooDataSorter and returns that instance.
static CooDataSorter	wrap(double[] values, int[][] indices)	Factory method which wraps data in an instance of CooDataSorter and returns that instance.
static CooDataSorter	wrap(double[] values, int[] rowIndices, int[] colIndices)	Factory method which wraps data in an instance of CooDataSorter and returns that instance.
static CooDataSorter	wrap(Object[] values, int[] indices)	Factory method which wraps data in an instance of CooDataSorter and returns that instance.
static CooDataSorter	wrap(Object[] values, int[][] indices)	Factory method which wraps data in an instance of CooDataSorter and returns that instance.
static CooDataSorter	wrap(Object[] values, int[] rowIndices, int[] colIndices)	Factory method which wraps data in an instance of CooDataSorter and returns that instance.

Methods inherited from class java.lang.Object

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

Field Details

values

private final List<T> values

Non-zero values of the sparse tensor to sort.

keys

private final List<Integer>[] keys

Stores wrapped indices. Each list is a single index for the values data. If the tensor is of rank N, then each List in keys will have N entries. If there are M non-zero entries in the sparse tensor, then there will be M lists in the array.

Constructor Details

CooDataSorter

public CooDataSorter(List<T> values,
 List<Integer> rowIndices,
 List<Integer> colIndices)

Creates a sorter which can sort the indices and values of a COO matrix lexicographically.

When sparseSort() is called on an instance created with this constructor, the sorting will be done in place.

Parameters:

values - Non-zero values of the sparse COO matrix.

rowIndices - Row indices of the sparse COO matrix.

colIndices - Row indices of the sparse COO matrix.

CooDataSorter

private CooDataSorter(T[] values,
 int[][] indices,
 boolean transpose)

Wraps the data, and it's associated indices, of a sparse tensor.

Parameters:

values - Non-zero values of the sparse tensor.

indices - Indices of non-zero values in the sparse tensor.

transpose - Indicates if a transpose should be applied to the indices array.

Method Details

wrap

public static CooDataSorter wrap(Object[] values,
 int[][] indices)

Factory method which wraps data in an instance of CooDataSorter and returns that instance.

Parameters:

values - Non-zero values of the sparse tensor.

indices - Indices of non-zero values in the sparse tensor.

Returns:

A new instance of CooDataSorter which wraps the specified values and indices.

wrap

public static CooDataSorter wrap(Object[] values,
 int[] indices)

Factory method which wraps data in an instance of CooDataSorter and returns that instance.

Parameters:

values - Non-zero values of the sparse tensor.

indices - Indices of non-zero values in the sparse tensor.

Returns:

A new instance of CooDataSorter which wraps the specified values and indices.

wrap

public static CooDataSorter wrap(double[] values,
 int[][] indices)

Factory method which wraps data in an instance of CooDataSorter and returns that instance.

Parameters:

values - Non-zero values of the sparse tensor.

indices - Indices of non-zero values in the sparse tensor.

Returns:

A new instance of CooDataSorter which wraps the specified values and indices.

wrap

public static CooDataSorter wrap(double[] values,
 int[] indices)

Factory method which wraps data in an instance of CooDataSorter and returns that instance.

Parameters:

values - Non-zero values of the sparse tensor.

indices - Indices of non-zero values in the sparse tensor.

Returns:

A new instance of CooDataSorter which wraps the specified values and indices.

wrap

public static CooDataSorter wrap(double[] values,
 int[] rowIndices,
 int[] colIndices)

Factory method which wraps data in an instance of CooDataSorter and returns that instance.

Parameters:

values - Non-zero values of the sparse tensor.

rowIndices - Row indices of non-zero entries of the sparse tensor.

colIndices - Column indices of non-zero entries of the sparse tensor.

Returns:

A new instance of CooDataSorter which wraps the specified values and indices.

wrap

public static CooDataSorter wrap(Object[] values,
 int[] rowIndices,
 int[] colIndices)

Factory method which wraps data in an instance of CooDataSorter and returns that instance.

Parameters:

values - Non-zero values of the sparse tensor.

rowIndices - Row indices of non-zero entries of the sparse tensor.

colIndices - Column indices of non-zero entries of the sparse tensor.

Returns:

A new instance of CooDataSorter which wraps the specified values and indices.

sparseSort

public CooDataSorter sparseSort()

Sorts the wrapped indices in lexicographical order while maintaining the order of the non-zero values so that their order matches the order of the indices.

Returns:

A reference to this sparse data wrapper.

sparseSortHelper

private void sparseSortHelper(int keyIdx,
 int start,
 int stop)

Sorts the specified key list over a specified range. During sorting, when swaps are made in the key list, the same indices are swapped in all other lists within the wrapper. This is done recursively top down.

Parameters:

keyIdx - Index of key list within keys.

start - Staring index of range to sort the key list over (inclusive).

stop - Stopping index of range to sort the key list over (exclusive).

unwrap

public void unwrap(Object[] values,
 int[][] indices)

Unwraps sparse data values and indices.

Parameters:

values - Storage for unwrapped values. Must have the same length as that passed to the constructor. Modified.

indices - Storage for unwrapped indices. Must have the same shape as that passed to the constructor. Modified.

unwrap

public void unwrap(Object[] values,
 int[] indices)

Unwraps sparse data values and indices.

Parameters:

values - Storage for unwrapped values. Must have the same length as that passed to the constructor. Modified.

indices - Storage for unwrapped indices. Must have the same shape as that passed to the constructor. Modified.

unwrap

public void unwrap(double[] values,
 int[][] indices)

Unwraps sparse data values and indices.

Parameters:

values - Storage for unwrapped values. Must have the same length as that passed to the constructor. Modified.

indices - Storage for unwrapped indices. Must have the same shape as that passed to the constructor. Modified.

unwrap

public void unwrap(double[] values,
 int[] rowIndices,
 int[] colIndices)

Unwraps sparse data values and indices.

Parameters:

values - Storage for unwrapped values. Must have the same length as that passed to the constructor. Modified.

rowIndices - Storage for unwrapped row indices. Must have the same shape as that passed to the constructor. Modified.

colIndices - Storage for unwrapped column indices. Must have the same shape as that passed to the constructor. Modified.

unwrap

public void unwrap(Object[] values,
 int[] rowIndices,
 int[] colIndices)

Unwraps sparse data values and indices.

Parameters:

values - Storage for unwrapped values. Must have the same length as that passed to the constructor. Modified.

rowIndices - Storage for unwrapped row indices. Must have the same shape as that passed to the constructor. Modified.

colIndices - Storage for unwrapped column indices. Must have the same shape as that passed to the constructor. Modified.

unwrap

public void unwrap(double[] values,
 int[] indices)

Unwraps sparse data values and indices.

Parameters:

values - Storage for unwrapped values. Must have the same length as that passed to the constructor. Modified.

indices - Storage for unwrapped indices. Must have the same shape as that passed to the constructor. Modified.