to construct from three arrays:
            1. data[:]   the entries of the matrix, in any order
            2. i[:]      the row indices of the matrix entries
            3. j[:]      the column indices of the matrix entries

        Where ``A[i[k], j[k]] = data[k]``.  When shape is not
        specified, it is inferred from the index arrays

Attributes
----------
dtype : dtype
    Data type of the matrix
shape : 2-tuple
    Shape of the matrix
ndim : int
    Number of dimensions (this is always 2)
nnz
size
data
    COO format data array of the matrix
row
    COO format row index array of the matrix
col
    COO format column index array of the matrix
has_canonical_format : bool
    Whether the matrix has sorted indices and no duplicates
format
T

Notes
-----

Sparse matrices can be used in arithmetic operations: they support
addition, subtraction, multiplication, division, and matrix power.

Advantages of the COO format
    - facilitates fast conversion among sparse formats
    - permits duplicate entries (see example)
    - very fast conversion to and from CSR/CSC formats

Disadvantages of the COO format
    - does not directly support:
        + arithmetic operations
        + slicing

Intended Usage
    - COO is a fast format for constructing sparse matrices
    - Once a COO matrix has been constructed, convert to CSR or
      CSC format for fast arithmetic and matrix vector operations
    - By default when converting to CSR or CSC format, duplicate (i,j)
      entries will be summed together.  This facilitates efficient
      construction of finite element matrices and the like. (see example)

Canonical format
    - Entries and coordinates sorted by row, then column.
    - There are no duplicate entries (i.e. duplicate (i,j) locations)
    - Data arrays MAY have explicit zeros.

Examples
--------

>>> # Constructing an empty matrix
>>> import numpy as np
>>> from scipy.sparse import coo_matrix
>>> coo_matrix((3, 4), dtype=np.int8).toarray()
array([[0, 0, 0, 0],
       [0, 0, 0, 0],
       [0, 0, 0, 0]], dtype=int8)

>>> # Constructing a matrix using ijv format
>>> row  = np.array([0, 3, 1, 0])
>>> col  = np.array([0, 3, 1, 2])
>>> data = np.array([4, 5, 7, 9])
>>> coo_matrix((data, (row, col)), shape=(4, 4)).toarray()
array([[4, 0, 9, 0],
       [0, 7, 0, 0],
       [0, 0, 0, 0],
       [0, 0, 0, 5]])

>>> # Constructing a matrix with duplicate coordinates
>>> row  = np.array([0, 0, 1, 3, 1, 0, 0])
>>> col  = np.array([0, 2, 1, 3, 1, 0, 0])
>>> data = np.array([1, 1, 1, 1, 1, 1, 1])
>>> coo = coo_matrix((data, (row, col)), shape=(4, 4))
>>> # Duplicate coordinates are maintained until implicitly or explicitly summed
>>> np.max(coo.data)
1
>>> coo.toarray()
array([[3, 0, 1, 0],
       [0, 2, 0, 0],
       [0, 0, 0, 0],
       [0, 0, 0, 1]])

c