def sparse_reverse_permute(A, rperm=(), cperm=(), safe=True):
"""
Performs a reverse permutations of the rows and columns of a sparse CSR/CSC matrix or Qobj
according to the permutation arrays rperm and cperm, respectively. Here, the permutation
arrays specify the order of the rows and columns used to permute the original array/Qobj.
Parameters
----------
A : qobj, csr_matrix, csc_matrix
Input matrix.
rperm : array_like of integers
Array of row permutations.
cperm : array_like of integers
Array of column permutations.
safe : bool
Check structure of permutation arrays.
Returns
-------
perm_csr : csr_matrix, csc_matrix
CSR or CSC matrix with permuted rows/columns.
"""
rperm = np.asarray(rperm, dtype=np.int32)
cperm = np.asarray(cperm, dtype=np.int32)
nrows = A.shape[0]
ncols = A.shape[1]
if len(rperm)==0:
rperm = np.arange(nrows, dtype=np.int32)
if len(cperm)==0:
cperm = np.arange(ncols, dtype=np.int32)
if safe:
if len(np.setdiff1d(rperm, np.arange(nrows)))!=0:
raise Exception('Invalid row permutation array.')
if len(np.setdiff1d(cperm, np.arange(ncols)))!=0:
raise Exception('Invalid column permutation array.')
shp = A.shape
if A.__class__.__name__=='Qobj':
kind = 'csr'
dt = complex
data, ind, ptr = _sparse_reverse_permute(
A.data, A.indices, A.indptr,
nrows, ncols, rperm, cperm, 0)
else:
kind=A.getformat()
if kind=='csr':
flag = 0
elif kind=='csc':
flag = 1
else:
raise Exception('Input must be Qobj, CSR, or CSC matrix.')
data, ind, ptr = _sparse_reverse_permute(A.data, A.indices, A.indptr,
nrows, ncols, rperm, cperm, flag)
if kind=='csr':
return sp.csr_matrix((data, ind, ptr), shape=shp, dtype=data.dtype)
elif kind=='csc':
return sp.csc_matrix((data, ind, ptr), shape=shp, dtype=data.dtype)
def sparse_reverse_permute(A, rperm=(), cperm=(), safe=True):
"""
Performs a reverse permutations of the rows and columns of a sparse CSR/CSC matrix or Qobj
according to the permutation arrays rperm and cperm, respectively. Here, the permutation
arrays specify the order of the rows and columns used to permute the original array/Qobj.
Parameters
----------
A : qobj, csr_matrix, csc_matrix
Input matrix.
rperm : array_like of integers
Array of row permutations.
cperm : array_like of integers
Array of column permutations.
safe : bool
Check structure of permutation arrays.
Returns
-------
perm_csr : csr_matrix, csc_matrix
CSR or CSC matrix with permuted rows/columns.
"""
rperm = np.asarray(rperm, dtype=np.int32)
cperm = np.asarray(cperm, dtype=np.int32)
nrows = A.shape[0]
ncols = A.shape[1]
if len(rperm) == 0:
rperm = np.arange(nrows, dtype=np.int32)
if len(cperm) == 0:
cperm = np.arange(ncols, dtype=np.int32)
if safe:
if len(np.setdiff1d(rperm, np.arange(nrows))) != 0:
raise Exception('Invalid row permutation array.')
if len(np.setdiff1d(cperm, np.arange(ncols))) != 0:
raise Exception('Invalid column permutation array.')
shp = A.shape
if A.__class__.__name__ == 'Qobj':
kind = 'csr'
dt = complex
data, ind, ptr = _sparse_reverse_permute(A.data, A.indices, A.indptr,
nrows, ncols, rperm, cperm, 0)
else:
kind = A.getformat()
if kind == 'csr':
flag = 0
elif kind == 'csc':
flag = 1
else:
raise Exception('Input must be Qobj, CSR, or CSC matrix.')
data, ind, ptr = _sparse_reverse_permute(A.data, A.indices, A.indptr,
nrows, ncols, rperm, cperm,
flag)
if kind == 'csr':
return sp.csr_matrix((data, ind, ptr), shape=shp, dtype=data.dtype)
elif kind == 'csc':
return sp.csc_matrix((data, ind, ptr), shape=shp, dtype=data.dtype)