Class CooGetSet

java.lang.Object
org.flag4j.linalg.ops.sparse.coo.CooGetSet
public final class CooGetSet extends Object

A utility class that aids in getting or setting specified elements of a sparse COO tensor, matrix, or vector.

All methods in this class guarantee all results will be properly lexicographically sorted by indices.

  • Method Summary

    Modifier and Type
    Method
    Description
    static <T> void
    cooInsertNewValue(T value, int[] index, T[] srcEntries, int[][] srcIndices, int insertionPoint, T[] destEntries, int[][] destIndices)
    Inserts a new value into a sparse COO tensor.
    static <T> void
    cooInsertNewValue(T value, int rowIdx, int colIdx, T[] srcEntries, int[] srcRowIndices, int[] srcColIndices, int insertionPoint, T[] destEntries, int[] destRowIndices, int[] destColIndices)
    Inserts a new value into a sparse COO matrix.
    static <T> void
    cooInsertNewValue(T value, int index, T[] srcEntries, int[] srcIndices, int insertionPoint, T[] destEntries, int[] destIndices)
    Inserts a new value into a sparse COO vector.
    static <T> SparseVectorData<T>
    getCol(Shape shape, T[] entries, int[] rowIndices, int[] colIndices, int colIdx, int start, int end)
    Gets a specified column of a COO matrix between start (inclusive) and end (exclusive).
    static double
    getCoo(double[] entries, int[][] indices, int[] target)
    Gets an element of a sparse COO tensor at the specified index.
    static float
    getCoo(float[] entries, int[][] indices, int[] target)
    Gets an element of a sparse COO tensor at the specified index.
    static int
    getCoo(int[] entries, int[][] indices, int[] target)
    Gets an element of a sparse COO tensor at the specified index.
    static <T> T
    getCoo(T[] entries, int[][] indices, int[] target)
    Gets an element of a sparse COO tensor at the specified index.
    static <V> V
    getCoo(V[] entries, int[] indices, int index)
    Gets the specified element from a sparse COO vector.
    static <V> V
    getCoo(V[] entries, int[] rowIndices, int[] colIndices, int row, int col)
    Gets the specified element from a sparse COO matrix.
    static <T> SparseVectorData<T>
    getDiag(Shape shape, T[] entries, int[] rowIndices, int[] colIndices, int diagOffset)
    Gets the elements of a COO matrix along the specified diagonal.
    static <T> SparseVectorData<T>
    getRow(Shape shape, T[] entries, int[] rowIndices, int[] colIndices, int rowIdx, int start, int end)
    Gets a specified row of a COO matrix between start (inclusive) and end (exclusive).
    static <T> SparseMatrixData<T>
    getSlice(Shape shape, T[] entries, int[] rowIndices, int[] colIndices, int rowStart, int rowEnd, int colStart, int colEnd)
    Extracts a specified slice from a sparse COO matrix.
    static <T> SparseMatrixData<T>
    getTriL(int diagOffset, Shape shape, T[] entries, int[] rowIndices, int[] colIndices)
    Gets the lower-triangular portion of a sparse COO matrix with a possible diagonal offset.
    static <T> SparseMatrixData<T>
    getTriU(int diagOffset, Shape shape, T[] entries, int[] rowIndices, int[] colIndices)
    Gets the upper-triangular portion of a sparse COO matrix with a possible diagonal offset.
    static <T> SparseMatrixData<T>
    setCol(Shape srcShape, T[] srcEntries, int[] rowIndices, int[] colIndices, int colIdx, int size, T[] col, int[] indices)
    Sets a column of a sparse matrix to the values in a sparse tensor.
    static <T> SparseMatrixData<T>
    setRow(Shape srcShape, T[] srcEntries, int[] rowIndices, int[] colIndices, int rowIdx, int size, T[] row, int[] indices)
    Sets a specified row of a real sparse COO matrix to the values in a sparse COO vector.
    static <T> SparseMatrixData<T>
    setSlice(Shape shape1, T[] src1Entries, int[] src1RowIndices, int[] src1ColIndices, Shape shape2, T[] src2Entries, int[] src2RowIndices, int[] src2ColIndices, int row, int col)
    Copies a sparse matrix and sets a slice of the sparse COO matrix to the data of another sparse COO matrix.

    Methods inherited from class java.lang.Object

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

  • Method Details

    • setRow

      public static <T> SparseMatrixData<T> setRow(Shape srcShape, T[] srcEntries, int[] rowIndices, int[] colIndices, int rowIdx, int size, T[] row, int[] indices)
      Sets a specified row of a real sparse COO matrix to the values in a sparse COO vector.
      Parameters:
      srcShape - Shape of the matrix to set row in.
      srcEntries - Non-zero data of the COO matrix.
      rowIndices - Non-zero row indices of the COO matrix.
      colIndices - Non-zero column indices of the COO matrix.
      rowIdx - Index of the row to set.
      size - Full size of the COO vector.
      row - Non-zero data of the COO vector containing new row values.
      indices - Non-zero indices of the COO vector containing new row values.
      Returns:
      Sparse matrix data containing the data for the COO matrix resulting from setting the specified row of the provided COO matrix to the provided COO vector.
      Throws:
      IllegalArgumentException - If srcShape.get(1) != size.

    • setCol

      public static <T> SparseMatrixData<T> setCol(Shape srcShape, T[] srcEntries, int[] rowIndices, int[] colIndices, int colIdx, int size, T[] col, int[] indices)
      Sets a column of a sparse matrix to the values in a sparse tensor.
      Parameters:
      srcShape - Shape of the matrix to set column in.
      srcEntries - Non-zero data of the COO matrix.
      rowIndices - Non-zero row indices of the COO matrix.
      colIndices - Non-zero column indices of the COO matrix.
      colIdx - Index of the column to set.
      size - Full size of the COO vector.
      col - Non-zero data of the COO vector containing new column values.
      indices - Non-zero indices of the COO vector containing new column values.
      Returns:
      A copy of the src matrix with the specified column set to the col sparse vector.
      Throws:
      IllegalArgumentException - If the src matrix does not have the same number of rows as total data in the col vector.

    • getCoo

      public static <V> V getCoo(V[] entries, int[] indices, int index)
      Gets the specified element from a sparse COO vector.
      Parameters:
      entries - Non-zero values of the sparse COO vector.
      indices - Non-zero indices of the sparse COO vector.
      index - Index of the value to get from the vector.
      Returns:
      The value in the sparse COO vector at the specified index if it exists. If the value is not found within the non-zero data, null will be returned.

    • getCoo

      public static <V> V getCoo(V[] entries, int[] rowIndices, int[] colIndices, int row, int col)
      Gets the specified element from a sparse COO matrix.
      Parameters:
      entries - Non-zero values of the sparse COO matrix from which to get element.
      rowIndices - Non-zero row indices for the sparse COO matrix.
      colIndices - Non-zero column indices for the sparse COO matrix.
      row - Row index of the value to get from the sparse matrix.
      col - Column index of the value to get from the sparse matrix.
      Returns:
      The value in the sparse COO matrix at the specified row and column indices if it exists. If the value is not found within the non-zero data, null will be returned.

    • getCoo

      public static <T> T getCoo(T[] entries, int[][] indices, int[] target)
      Gets an element of a sparse COO tensor at the specified index. If no non-zero value exists, then null is returned.
      Parameters:
      entries - Non-zero data of the COO tensor.
      indices - Non-zero indices o the COO tensor.
      target - Target index to search for in indices.
      Returns:
      The value in data which has an index matching the target. That is, if some idx is found such that Arrays.equals(indices[idx], target), then data[idx] is returned. If no such idx id found, then null is returned.

    • cooInsertNewValue

      public static <T> void cooInsertNewValue(T value, int[] index, T[] srcEntries, int[][] srcIndices, int insertionPoint, T[] destEntries, int[][] destIndices)
      Inserts a new value into a sparse COO tensor. This assumes there is no non-zero value already at the specified index.
      Parameters:
      value - Value to insert into the tensor.
      index - Non-zero index for new value.
      srcEntries - Non-zero data of the source tensor. Unmodified.
      srcIndices - Non-zero indices of the source tensor. Assumed to be rectangular. Unmodified.
      insertionPoint - Index in srcEntries and srcIndices to insert value and index.
      destEntries - Destination for storing the result of inserting the value into srcEntries.
      destIndices - Destination for storing the result of inserting the index into srcIndices.
      Throws:
      IllegalArgumentException - If destEntries.length != srcEntries.length + 1 or destIndices.length != srcIndices.length + 1
      IllegalArgumentException - If index.length != srcIndices[0].length.

    • cooInsertNewValue

      public static <T> void cooInsertNewValue(T value, int rowIdx, int colIdx, T[] srcEntries, int[] srcRowIndices, int[] srcColIndices, int insertionPoint, T[] destEntries, int[] destRowIndices, int[] destColIndices)
      Inserts a new value into a sparse COO matrix. This assumes there is no non-zero value at the specified row and column.
      Parameters:
      value - Value to insert into the matrix.
      rowIdx - index for the value to insert.
      srcEntries - Non-zero data of the source matrix. Unmodified.
      srcRowIndices - Non-zero row indices of the source matrix. Unmodified.
      insertionPoint - Index in srcEntries, srcRowIndices, srcColIndices to insert value, rowIdx, and colIdx
      destEntries - Destination for storing the result of inserting the value into srcEntries.
      destRowIndices - Destination for storing the result of inserting the rowIdx into srcRowIndices.
      destColIndices - Destination for storing the result of inserting the colIdx into srcColIndices.
      Throws:
      IllegalArgumentException - If destEntries.length != srcEntries.length + 1 or destIndices.length != srcRowIndices.length + 1 or destColIndices.length != srcColIndices.length + 1.

    • cooInsertNewValue

      public static <T> void cooInsertNewValue(T value, int index, T[] srcEntries, int[] srcIndices, int insertionPoint, T[] destEntries, int[] destIndices)
      Inserts a new value into a sparse COO vector. This assumes there is no non-zero value at the specified index.
      Parameters:
      value - Value to insert into the vector.
      index - Index for the value to insert.
      srcEntries - Non-zero data of the source vector. Unmodified.
      srcIndices - Non-zero indices of the source vector. Unmodified.
      insertionPoint - Index in srcEntries and srcIndices to insert value and index.
      destEntries - Destination for storing the result of inserting the value into srcEntries.
      destIndices - Destination for storing the result of inserting the index into srcIndices.
      Throws:
      IllegalArgumentException - If destEntries.length != srcEntries.length + 1 or destIndices.length != srcIndices.length + 1.

    • getCoo

      public static double getCoo(double[] entries, int[][] indices, int[] target)
      Gets an element of a sparse COO tensor at the specified index. If no non-zero value exists, then null is returned.
      Parameters:
      entries - Non-zero data of the COO tensor.
      indices - Non-zero indices o the COO tensor.
      target - Target index to search for in indices.
      Returns:
      The value in data which has an index matching the target. That is, if some idx is found such that Arrays.equals(indices[idx], target), then data[idx] is returned. If no such idx id found, then null is returned.

    • getCoo

      public static float getCoo(float[] entries, int[][] indices, int[] target)
      Gets an element of a sparse COO tensor at the specified index. If no non-zero value exists, then null is returned.
      Parameters:
      entries - Non-zero data of the COO tensor.
      indices - Non-zero indices o the COO tensor.
      target - Target index to search for in indices.
      Returns:
      The value in data which has an index matching the target. That is, if some idx is found such that Arrays.equals(indices[idx], target), then data[idx] is returned. If no such idx id found, then null is returned.

    • getCoo

      public static int getCoo(int[] entries, int[][] indices, int[] target)
      Gets an element of a sparse COO tensor at the specified index. If no non-zero value exists, then null is returned.
      Parameters:
      entries - Non-zero data of the COO tensor.
      indices - Non-zero indices o the COO tensor.
      target - Target index to search for in indices.
      Returns:
      The value in data which has an index matching the target. That is, if some idx is found such that Arrays.equals(indices[idx], target), then data[idx] is returned. If no such idx id found, then null is returned.

    • getTriU

      public static <T> SparseMatrixData<T> getTriU(int diagOffset, Shape shape, T[] entries, int[] rowIndices, int[] colIndices)
      Gets the upper-triangular portion of a sparse COO matrix with a possible diagonal offset. The remaining values will be zero.
      Parameters:
      diagOffset - Diagonal offset indicating which diagonal to extract values at or above.
      • If diagOffset == 0 then the properly upper-triangular portion of the matrix is extracted.
      • If diagOffset == k where k > 0 then the values at and above the kth super-diagonal.
      • If diagOffset == k where k < 0 then the values at and above the kth sub-diagonal.
      shape - Shape of the COO matrix.
      entries - Non-zero data of the COO matrix.
      rowIndices - Row indices of the COO matrix.
      colIndices - Column indices of the COO matrix.
      Returns:
      A data container containing the resulting upper-triangular non-zero data, row indices, and column indices.

    • getTriL

      public static <T> SparseMatrixData<T> getTriL(int diagOffset, Shape shape, T[] entries, int[] rowIndices, int[] colIndices)
      Gets the lower-triangular portion of a sparse COO matrix with a possible diagonal offset. The remaining values will be zero.
      Parameters:
      diagOffset - Diagonal offset indicating which diagonal to extract values at or below.
      • If diagOffset == 0 then the properly lower-triangular portion of the matrix is extracted.
      • If diagOffset == k where k > 0 then the values at and below the kth super-diagonal.
      • If diagOffset == k where k < 0 then the values at and below the kth sub-diagonal.
      shape - Shape of the COO matrix.
      entries - Non-zero data of the COO matrix.
      rowIndices - Row indices of the COO matrix.
      colIndices - Column indices of the COO matrix.
      Returns:
      A data container containing the resulting lower-triangular non-zero data, row indices, and column indices.

    • setSlice

      public static <T> SparseMatrixData<T> setSlice(Shape shape1, T[] src1Entries, int[] src1RowIndices, int[] src1ColIndices, Shape shape2, T[] src2Entries, int[] src2RowIndices, int[] src2ColIndices, int row, int col)
      Copies a sparse matrix and sets a slice of the sparse COO matrix to the data of another sparse COO matrix.
      Parameters:
      shape1 - Shape of the first matrix.
      src1Entries - Non-zero data of the matrix to set slice within.
      src1RowIndices - Row indices of the matrix to set slice within.
      src1ColIndices - Column indices of the matrix to set slice within.
      shape2 - Shape of the first matrix.
      src2Entries - Non-zero data of the matrix to copy into the specified slice.
      src2RowIndices - Row indices of the matrix to copy into the specified slice.
      src2ColIndices - Column indices of the matrix to copy into the specified slice.
      row - Starting row index of slice.
      col - Starting column index of slice.
      Returns:
      A sparse data container containing the result of setting the slice in the source matrix.
      Throws:
      IndexOutOfBoundsException - If the values matrix does not fit in the src matrix given the row and column index.

    • getSlice

      public static <T> SparseMatrixData<T> getSlice(Shape shape, T[] entries, int[] rowIndices, int[] colIndices, int rowStart, int rowEnd, int colStart, int colEnd)
      Extracts a specified slice from a sparse COO matrix.
      Parameters:
      shape - Shape of the COO matrix.
      entries - Non-zero data of the COO matrix.
      rowIndices - Row indices of the COO matrix.
      colIndices - Column indices of the COO matrix.
      rowStart - Starting row index of the slice (inclusive).
      rowEnd - Ending row index of the slice (exclusive).
      colStart - Staring column index of a slice (inclusive).
      colEnd - Ending column index of the slice (exclusive).
      Returns:
      A sparse data container containing the specified slice extracted from the COO matrix.
      Throws:
      IndexOutOfBoundsException - If the specified slice does not fit into the matrix.

    • getDiag

      public static <T> SparseVectorData<T> getDiag(Shape shape, T[] entries, int[] rowIndices, int[] colIndices, int diagOffset)
      Gets the elements of a COO matrix along the specified diagonal.
      Parameters:
      shape - Shape of the COO matrix.
      entries - Non-zero data of the COO matrix.
      rowIndices - Non-zero row indices of the COO matrix.
      colIndices - Non-zero column indices of the COO matrix.
      diagOffset - The diagonal to get within the COO matrix.
      • If diagOffset == 0: Then the elements of the principle diagonal are collected.
      • If diagOffset < 0: Then the elements of the sub-diagonal diagOffset below the principle diagonal are collected.
      • If diagOffset > 0: Then the elements of the super-diagonal diagOffset above the principle diagonal are collected.
      Returns:
      A sparse vector data object containing the non-zero data and indices along the specified diagonal of the COO matrix.

    • getRow

      public static <T> SparseVectorData<T> getRow(Shape shape, T[] entries, int[] rowIndices, int[] colIndices, int rowIdx, int start, int end)
      Gets a specified row of a COO matrix between start (inclusive) and end (exclusive).
      Parameters:
      shape - Shape of the COO matrix.
      entries - Non-zero data of the COO matrix.
      rowIndices - Non-zero row indices of the COO matrix.
      colIndices - Non-zero column indices of the COO matrix.
      rowIdx - Index of the row of this matrix to get.
      colStart - Starting column of the row (inclusive).
      colEnd - Ending column of the row (exclusive).
      Returns:
      The row at index rowIdx of this matrix between the start and end indices.
      Throws:
      IndexOutOfBoundsException - If either end are start out of bounds for the shape of this matrix.
      IllegalArgumentException - If end is less than start.

    • getCol

      public static <T> SparseVectorData<T> getCol(Shape shape, T[] entries, int[] rowIndices, int[] colIndices, int colIdx, int start, int end)
      Gets a specified column of a COO matrix between start (inclusive) and end (exclusive).
      Parameters:
      shape - Shape of the COO matrix.
      entries - Non-zero data of the COO matrix.
      rowIndices - Non-zero row indices of the COO matrix.
      colIndices - Non-zero column indices of the COO matrix.
      colIdx - Index of the column of this matrix to get.
      colStart - Starting column of the row (inclusive).
      colEnd - Ending column of the row (exclusive).
      Returns:
      The column at index colIdx of this matrix between the start and end indices.
      Throws:
      IndexOutOfBoundsException - If either end are start out of bounds for the shape of this matrix.
      IllegalArgumentException - If end is less than start.