"""Compressed Sparse Row matrix format"""
__docformat__ = "restructuredtext en"
__all__ = ['csr_array', 'csr_matrix', 'isspmatrix_csr']
import numpy as np
from ._matrix import spmatrix
from ._base import _spbase, sparray
from ._sparsetools import (csr_tocsc, csr_tobsr, csr_count_blocks,
get_csr_submatrix, csr_sample_values)
from ._sputils import upcast
from ._compressed import _cs_matrix
class _csr_base(_cs_matrix):
_format = 'csr'
_allow_nd = (1, 2)
def transpose(self, axes=None, copy=False):
if axes is not None and axes != (1, 0):
raise ValueError("Sparse arrays/matrices do not support "
"an 'axes' parameter because swapping "
"dimensions is the only logical permutation.")
if self.ndim == 1:
return self.copy() if copy else self
M, N = self.shape
return self._csc_container((self.data, self.indices,
self.indptr), shape=(N, M), copy=copy)
transpose.__doc__ = _spbase.transpose.__doc__
def tolil(self, copy=False):
if self.ndim != 2:
raise ValueError("Cannot convert a 1d sparse array to lil format")
lil = self._lil_container(self.shape, dtype=self.dtype)
self.sum_duplicates()
ptr,ind,dat = self.indptr,self.indices,self.data
rows, data = lil.rows, lil.data
for n in range(self.shape[0]):
start = ptr[n]
end = ptr[n+1]
rows[n] = ind[start:end].tolist()
data[n] = dat[start:end].tolist()
return lil
tolil.__doc__ = _spbase.tolil.__doc__
def tocsr(self, copy=False):
if copy:
return self.copy()
else:
return self
tocsr.__doc__ = _spbase.tocsr.__doc__
def tocoo(self, copy=False):
A = super().tocoo(copy=copy)
# CSR-to-COO conversion always preserves [non-]canonicity
# (indices sorting, presense of duplicate elements).
# Handled here instead of _cs_matrix because CSC-to-COO generally does not.
A.has_canonical_format = self.has_canonical_format
return A
tocoo.__doc__ = _spbase.tocoo.__doc__
def tocsc(self, copy=False):
if self.ndim != 2:
raise ValueError("Cannot convert a 1d sparse array to csc format")
M, N = self.shape
idx_dtype = self._get_index_dtype((self.indptr, self.indices),
maxval=max(self.nnz, M))
indptr = np.empty(N + 1, dtype=idx_dtype)
indices = np.empty(self.nnz, dtype=idx_dtype)
data = np.empty(self.nnz, dtype=upcast(self.dtype))
csr_tocsc(M, N,
self.indptr.astype(idx_dtype, copy=False),
self.indices.astype(idx_dtype, copy=False),
self.data,
indptr,
indices,
data)
A = self._csc_container((data, indices, indptr), shape=self.shape)
A.has_sorted_indices = True
return A
tocsc.__doc__ = _spbase.tocsc.__doc__
def tobsr(self, blocksize=None, copy=True):
if self.ndim != 2:
raise ValueError("Cannot convert a 1d sparse array to bsr format")
if blocksize is None:
from ._spfuncs import estimate_blocksize
return self.tobsr(blocksize=estimate_blocksize(self))
elif blocksize == (1,1):
arg1 = (self.data.reshape(-1,1,1),self.indices,self.indptr)
return self._bsr_container(arg1, shape=self.shape, copy=copy)
else:
R,C = blocksize
M,N = self.shape
if R < 1 or C < 1 or M % R != 0 or N % C != 0:
raise ValueError(f'invalid blocksize {blocksize}')
blks = csr_count_blocks(M,N,R,C,self.indptr,self.indices)
idx_dtype = self._get_index_dtype((self.indptr, self.indices),
maxval=max(N//C, blks))
indptr = np.empty(M//R+1, dtype=idx_dtype)
indices = np.empty(blks, dtype=idx_dtype)
data = np.zeros((blks,R,C), dtype=self.dtype)
csr_tobsr(M, N, R, C,
self.indptr.astype(idx_dtype, copy=False),
self.indices.astype(idx_dtype, copy=False),
self.data,
indptr, indices, data.ravel())
return self._bsr_container(
(data, indices, indptr), shape=self.shape
)
tobsr.__doc__ = _spbase.tobsr.__doc__
# these functions are used by the parent class (_cs_matrix)
# to remove redundancy between csc_matrix and csr_array
@staticmethod
def _swap(x):
"""swap the members of x if this is a column-oriented matrix
"""
return x
def __iter__(self):
if self.ndim == 1:
zero = self.dtype.type(0)
u = 0
for v, d in zip(self.indices, self.data):
for _ in range(v - u):
yield zero
yield d
u = v + 1
for _ in range(self.shape[0] - u):
yield zero
return
indptr = np.zeros(2, dtype=self.indptr.dtype)
# return 1d (sparray) or 2drow (spmatrix)
shape = self.shape[1:] if isinstance(self, sparray) else (1, self.shape[1])
i0 = 0
for i1 in self.indptr[1:]:
indptr[1] = i1 - i0
indices = self.indices[i0:i1]
data = self.data[i0:i1]
yield self.__class__((data, indices, indptr), shape=shape, copy=True)
i0 = i1
def _getrow(self, i):
"""Returns a copy of row i of the matrix, as a (1 x n)
CSR matrix (row vector).
"""
if self.ndim == 1:
if i not in (0, -1):
raise IndexError(f'index ({i}) out of range')
return self.reshape((1, self.shape[0]), copy=True)
M, N = self.shape
i = int(i)
if i < 0:
i += M
if i < 0 or i >= M:
raise IndexError(f'index ({i}) out of range')
indptr, indices, data = get_csr_submatrix(
M, N, self.indptr, self.indices, self.data, i, i + 1, 0, N)
return self.__class__((data, indices, indptr), shape=(1, N),
dtype=self.dtype, copy=False)
def _getcol(self, i):
"""Returns a copy of column i. A (m x 1) sparse array (column vector).
"""
if self.ndim == 1:
raise ValueError("getcol not provided for 1d arrays. Use indexing A[j]")
M, N = self.shape
i = int(i)
if i < 0:
i += N
if i < 0 or i >= N:
raise IndexError(f'index ({i}) out of range')
indptr, indices, data = get_csr_submatrix(
M, N, self.indptr, self.indices, self.data, 0, M, i, i + 1)
return self.__class__((data, indices, indptr), shape=(M, 1),
dtype=self.dtype, copy=False)
def _get_int(self, idx):
spot = np.flatnonzero(self.indices == idx)
if spot.size:
return self.data[spot[0]]
return self.data.dtype.type(0)
def _get_slice(self, idx):
if idx == slice(None):
return self.copy()
if idx.step in (1, None):
ret = self._get_submatrix(0, idx, copy=True)
return ret.reshape(ret.shape[-1])
return self._minor_slice(idx)
def _get_array(self, idx):
idx_dtype = self._get_index_dtype(self.indices)
idx = np.asarray(idx, dtype=idx_dtype)
if idx.size == 0:
return self.__class__([], dtype=self.dtype)
M, N = 1, self.shape[0]
row = np.zeros_like(idx, dtype=idx_dtype)
col = np.asarray(idx, dtype=idx_dtype)
val = np.empty(row.size, dtype=self.dtype)
csr_sample_values(M, N, self.indptr, self.indices, self.data,
row.size, row, col, val)
new_shape = col.shape if col.shape[0] > 1 else (col.shape[0],)
return self.__class__(val.reshape(new_shape))
def _get_intXarray(self, row, col):
return self._getrow(row)._minor_index_fancy(col)
def _get_intXslice(self, row, col):
if col.step in (1, None):
return self._get_submatrix(row, col, copy=True)
# TODO: uncomment this once it's faster:
# return self._getrow(row)._minor_slice(col)
M, N = self.shape
start, stop, stride = col.indices(N)
ii, jj = self.indptr[row:row+2]
row_indices = self.indices[ii:jj]
row_data = self.data[ii:jj]
if stride > 0:
ind = (row_indices >= start) & (row_indices < stop)
else:
ind = (row_indices <= start) & (row_indices > stop)
if abs(stride) > 1:
ind &= (row_indices - start) % stride == 0
row_indices = (row_indices[ind] - start) // stride
row_data = row_data[ind]
row_indptr = np.array([0, len(row_indices)])
if stride < 0:
row_data = row_data[::-1]
row_indices = abs(row_indices[::-1])
shape = (1, max(0, int(np.ceil(float(stop - start) / stride))))
return self.__class__((row_data, row_indices, row_indptr), shape=shape,
dtype=self.dtype, copy=False)
def _get_sliceXint(self, row, col):
if row.step in (1, None):
return self._get_submatrix(row, col, copy=True)
return self._major_slice(row)._get_submatrix(minor=col)
def _get_sliceXarray(self, row, col):
return self._major_slice(row)._minor_index_fancy(col)
def _get_arrayXint(self, row, col):
res = self._major_index_fancy(row)._get_submatrix(minor=col)
if row.ndim > 1:
return res.reshape(row.shape)
return res
def _get_arrayXslice(self, row, col):
if col.step not in (1, None):
col = np.arange(*col.indices(self.shape[1]))
return self._get_arrayXarray(row, col)
return self._major_index_fancy(row)._get_submatrix(minor=col)
def _set_int(self, idx, x):
self._set_many(0, idx, x)
def _set_array(self, idx, x):
x = np.broadcast_to(x, idx.shape)
self._set_many(np.zeros_like(idx), idx, x)
def isspmatrix_csr(x):
"""Is `x` of csr_matrix type?
.. warning::
SciPy sparse is shifting from a sparse matrix interface to a sparse
array interface. In the next few releases we expect to deprecate the
sparse matrix interface. For documentation of the matrix
interface, see the :ref:`spmatrix interface docs <spmatrix_api>`.
For guidance on converting existing code to sparse arrays, see
:ref:`Migration from spmatrix to sparray <migration_to_sparray>`.
Parameters
----------
x
object to check for being a csr matrix
Returns
-------
bool
True if `x` is a csr matrix, False otherwise
Examples
--------
>>> from scipy.sparse import csr_array, csr_matrix, coo_matrix, isspmatrix_csr
>>> isspmatrix_csr(csr_matrix([[5]]))
True
>>> isspmatrix_csr(csr_array([[5]]))
False
>>> isspmatrix_csr(coo_matrix([[5]]))
False
"""
return isinstance(x, csr_matrix)
# This namespace class separates array from matrix with isinstance
class csr_array(_csr_base, sparray):
"""
Compressed Sparse Row array.
This can be instantiated in several ways:
csr_array(D)
where D is a 2-D ndarray
csr_array(S)
with another sparse array or matrix S (equivalent to S.tocsr())
csr_array((M, N), [dtype])
to construct an empty array with shape (M, N)
dtype is optional, defaulting to dtype='d'.
csr_array((data, (row_ind, col_ind)), [shape=(M, N)])
where ``data``, ``row_ind`` and ``col_ind`` satisfy the
relationship ``a[row_ind[k], col_ind[k]] = data[k]``.
csr_array((data, indices, indptr), [shape=(M, N)])
is the standard CSR representation where the column indices for
row i are stored in ``indices[indptr[i]:indptr[i+1]]`` and their
corresponding values are stored in ``data[indptr[i]:indptr[i+1]]``.
If the shape parameter is not supplied, the array dimensions
are inferred from the index arrays.
Attributes
----------
data : ndarray
CSR format data array of the array
indices : ndarray
CSR format index array of the array
indptr : ndarray
CSR format index pointer array of the array
has_sorted_indices : bool
Whether indices are sorted
has_canonical_format : bool
Whether indices are sorted and no duplicate entries exist
dtype : dtype
Data type of the array
shape : 2-tuple
Shape of the array
ndim : int
Number of dimensions (this is always 2)
format : str
Three letter code for the format of the array storage, e.g. 'csr'
nnz : int
Number of values stored in the array
size : int
Number of values stored in the array
T : csr_array
The transpose of the array
mT : csr_array
The matrix transpose of the array
Notes
-----
Sparse arrays can be used in arithmetic operations: they support
addition, subtraction, multiplication, division, and matrix power.
Advantages of the CSR format
- efficient arithmetic operations CSR + CSR, CSR * CSR, etc.
- efficient row slicing
- fast matrix vector products
Disadvantages of the CSR format
- slow column slicing operations (consider CSC)
- changes to the sparsity structure are expensive (consider LIL or DOK)
Canonical Format
- Within each row, indices are sorted by column.
- There are no duplicate entries.
Examples
--------
>>> import numpy as np
>>> from scipy.sparse import csr_array
>>> csr_array((3, 4), dtype=np.int8).toarray()
array([[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]], dtype=int8)
>>> row = np.array([0, 0, 1, 2, 2, 2])
>>> col = np.array([0, 2, 2, 0, 1, 2])
>>> data = np.array([1, 2, 3, 4, 5, 6])
>>> csr_array((data, (row, col)), shape=(3, 3)).toarray()
array([[1, 0, 2],
[0, 0, 3],
[4, 5, 6]])
>>> indptr = np.array([0, 2, 3, 6])
>>> indices = np.array([0, 2, 2, 0, 1, 2])
>>> data = np.array([1, 2, 3, 4, 5, 6])
>>> csr_array((data, indices, indptr), shape=(3, 3)).toarray()
array([[1, 0, 2],
[0, 0, 3],
[4, 5, 6]])
Duplicate entries are summed together:
>>> row = np.array([0, 1, 2, 0])
>>> col = np.array([0, 1, 1, 0])
>>> data = np.array([1, 2, 4, 8])
>>> csr_array((data, (row, col)), shape=(3, 3)).toarray()
array([[9, 0, 0],
[0, 2, 0],
[0, 4, 0]])
As an example of how to construct a CSR array incrementally,
the following snippet builds a term-document array from texts:
>>> docs = [["hello", "world", "hello"], ["goodbye", "cruel", "world"]]
>>> indptr = [0]
>>> indices = []
>>> data = []
>>> vocabulary = {}
>>> for d in docs:
... for term in d:
... index = vocabulary.setdefault(term, len(vocabulary))
... indices.append(index)
... data.append(1)
... indptr.append(len(indices))
...
>>> csr_array((data, indices, indptr), dtype=int).toarray()
array([[2, 1, 0, 0],
[0, 1, 1, 1]])
""" # numpydoc ignore=PR01
[docs]
class csr_matrix(spmatrix, _csr_base):
"""
Compressed Sparse Row matrix.
.. warning::
SciPy sparse is shifting from a sparse matrix interface to a sparse
array interface. In the next few releases we expect to deprecate the
sparse matrix interface. For documentation of the matrix
interface, see the :ref:`spmatrix interface docs <spmatrix_api>`.
For guidance on converting existing code to sparse arrays, see
:ref:`Migration from spmatrix to sparray <migration_to_sparray>`.
This can be instantiated in several ways:
csr_matrix(D)
where D is a 2-D ndarray
csr_matrix(S)
with another sparse array or matrix S (equivalent to S.tocsr())
csr_matrix((M, N), [dtype])
to construct an empty matrix with shape (M, N)
dtype is optional, defaulting to dtype='d'.
csr_matrix((data, (row_ind, col_ind)), [shape=(M, N)])
where ``data``, ``row_ind`` and ``col_ind`` satisfy the
relationship ``a[row_ind[k], col_ind[k]] = data[k]``.
csr_matrix((data, indices, indptr), [shape=(M, N)])
is the standard CSR representation where the column indices for
row i are stored in ``indices[indptr[i]:indptr[i+1]]`` and their
corresponding values are stored in ``data[indptr[i]:indptr[i+1]]``.
If the shape parameter is not supplied, the matrix dimensions
are inferred from the index arrays.
Attributes
----------
data : ndarray
CSR format data array of the matrix
indices : ndarray
CSR format index array of the matrix
indptr : ndarray
CSR format index pointer array of the matrix
has_sorted_indices : bool
Whether indices are sorted
has_canonical_format : bool
Whether indices are sorted and no duplicate entries exist
dtype : dtype
Data type of the matrix
shape : 2-tuple
Shape of the matrix
ndim : int
Number of dimensions (this is always 2)
format : str
Three letter code for the format of the matrix storage, e.g. 'csr'
nnz : int
Number of values stored in the matrix
size : int
Number of values stored in the matrix
T : csr_matrix
The transpose of the matrix
mT : csr_matrix
The matrix transpose
Notes
-----
Sparse matrices can be used in arithmetic operations: they support
addition, subtraction, multiplication, division, and matrix power.
Advantages of the CSR format
- efficient arithmetic operations CSR + CSR, CSR * CSR, etc.
- efficient row slicing
- fast matrix vector products
Disadvantages of the CSR format
- slow column slicing operations (consider CSC)
- changes to the sparsity structure are expensive (consider LIL or DOK)
Canonical Format
- Within each row, indices are sorted by column.
- There are no duplicate entries.
Examples
--------
>>> import numpy as np
>>> from scipy.sparse import csr_matrix
>>> csr_matrix((3, 4), dtype=np.int8).toarray()
array([[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]], dtype=int8)
>>> row = np.array([0, 0, 1, 2, 2, 2])
>>> col = np.array([0, 2, 2, 0, 1, 2])
>>> data = np.array([1, 2, 3, 4, 5, 6])
>>> csr_matrix((data, (row, col)), shape=(3, 3)).toarray()
array([[1, 0, 2],
[0, 0, 3],
[4, 5, 6]])
>>> indptr = np.array([0, 2, 3, 6])
>>> indices = np.array([0, 2, 2, 0, 1, 2])
>>> data = np.array([1, 2, 3, 4, 5, 6])
>>> csr_matrix((data, indices, indptr), shape=(3, 3)).toarray()
array([[1, 0, 2],
[0, 0, 3],
[4, 5, 6]])
Duplicate entries are summed together:
>>> row = np.array([0, 1, 2, 0])
>>> col = np.array([0, 1, 1, 0])
>>> data = np.array([1, 2, 4, 8])
>>> csr_matrix((data, (row, col)), shape=(3, 3)).toarray()
array([[9, 0, 0],
[0, 2, 0],
[0, 4, 0]])
As an example of how to construct a CSR matrix incrementally,
the following snippet builds a term-document matrix from texts:
>>> docs = [["hello", "world", "hello"], ["goodbye", "cruel", "world"]]
>>> indptr = [0]
>>> indices = []
>>> data = []
>>> vocabulary = {}
>>> for d in docs:
... for term in d:
... index = vocabulary.setdefault(term, len(vocabulary))
... indices.append(index)
... data.append(1)
... indptr.append(len(indices))
...
>>> csr_matrix((data, indices, indptr), dtype=int).toarray()
array([[2, 1, 0, 0],
[0, 1, 1, 1]])
"""