def __init__(self, blocks):
blocks = np.array(blocks)
assert isinstance(blocks, np.ndarray) and blocks.ndim == 2
self._blocks = blocks
assert all(
isinstance(op, OperatorInterface) for op in self._operators())
assert all(
any(self._blocks[i, j] is not None
for j in range(self._blocks.shape[1]))
for i in range(self._blocks.shape[0]))
assert all(
any(self._blocks[i, j] is not None
for i in range(self._blocks.shape[0]))
for j in range(self._blocks.shape[1]))
source_types = [None for j in range(self._blocks.shape[1])]
range_types = [None for i in range(self._blocks.shape[0])]
for (i, j), op in np.ndenumerate(self._blocks):
if op is not None:
assert source_types[j] is None or op.source == source_types[j]
source_types[j] = op.source
assert range_types[i] is None or op.range == range_types[i]
range_types[i] = op.range
self.source = VectorSpace(BlockVectorArray, tuple(source_types))
self.range = VectorSpace(BlockVectorArray, tuple(range_types))
self._source_dims = tuple(space.dim for space in self.source.subtype)
self._range_dims = tuple(space.dim for space in self.range.subtype)
self.num_source_blocks = len(source_types)
self.num_range_blocks = len(range_types)
self.linear = all(op.linear for op in self._operators())
self.build_parameter_type(inherits=list(self._operators()))
def __init__(self, op):
assert isinstance(op, DiffusionOperator)
self._impl = op
self.source = VectorSpace(ListVectorArray,
(WrappedVector, op.dim_source))
self.range = VectorSpace(ListVectorArray,
(WrappedVector, op.dim_range))
self.linear = True
def __init__(self, matrix, functional=False, vector=False, solver_options=None, name=None):
assert not (functional and vector)
super().__init__(matrix, solver_options=solver_options, name=name)
if not vector:
self.source = VectorSpace(ListVectorArray, (NumpyVector, matrix.shape[1]))
if not functional:
self.range = VectorSpace(ListVectorArray, (NumpyVector, matrix.shape[0]))
self.functional = functional
self.vector = vector
def __init__(self,
obj_id,
operators,
functionals,
vector_operators,
products=None,
pickle_subtypes=True,
array_type=MPIVectorArray):
d = mpi.get_object(obj_id)
visualizer = MPIVisualizer(obj_id)
super().__init__(operators,
functionals,
vector_operators,
products=products,
visualizer=visualizer,
cache_region=None,
name=d.name)
self.obj_id = obj_id
subtypes = mpi.call(_MPIDiscretization_get_subtypes, obj_id,
pickle_subtypes)
if all(subtype == subtypes[0] for subtype in subtypes):
subtypes = (subtypes[0], )
self.solution_space = VectorSpace(array_type,
(d.solution_space.type, subtypes))
self.build_parameter_type(inherits=(d, ))
self.parameter_space = d.parameter_space
def __init__(self, matrix, functional=False, vector=False, solver_options=None, name=None):
assert not (functional and vector)
super(NumpyListVectorArrayMatrixOperator, self).__init__(matrix, solver_options=solver_options, name=name)
if not vector:
self.source = VectorSpace(ListVectorArray, (NumpyVector, matrix.shape[1]))
if not functional:
self.range = VectorSpace(ListVectorArray, (NumpyVector, matrix.shape[0]))
self.functional = functional
self.vector = vector
def __init__(self,
obj_id,
functional=False,
vector=False,
with_apply2=False,
pickle_subtypes=True,
array_type=MPIVectorArray):
assert not (functional and vector)
self.obj_id = obj_id
self.op = op = mpi.get_object(obj_id)
self.functional = functional
self.vector = vector
self.with_apply2 = with_apply2
self.pickle_subtypes = pickle_subtypes
self.array_type = array_type
self.linear = op.linear
self.name = op.name
self.build_parameter_type(inherits=(op, ))
if vector:
self.source = NumpyVectorSpace(1)
assert self.source == op.source
else:
subtypes = mpi.call(_MPIOperator_get_source_subtypes, obj_id,
pickle_subtypes)
if all(subtype == subtypes[0] for subtype in subtypes):
subtypes = (subtypes[0], )
self.source = VectorSpace(array_type, (op.source.type, subtypes))
if functional:
self.range = NumpyVectorSpace(1)
assert self.range == op.range
else:
subtypes = mpi.call(_MPIOperator_get_range_subtypes, obj_id,
pickle_subtypes)
if all(subtype == subtypes[0] for subtype in subtypes):
subtypes = (subtypes[0], )
self.range = VectorSpace(array_type, (op.range.type, subtypes))
def FenicsVectorSpace(V):
return VectorSpace(ListVectorArray, (FenicsVector, V))
class NumpyListVectorArrayMatrixOperator(NumpyMatrixOperator):
"""Variant of |NumpyMatrixOperator| using |ListVectorArray| instead of |NumpyVectorArray|."""
def __init__(self, matrix, functional=False, vector=False, solver_options=None, name=None):
assert not (functional and vector)
super(NumpyListVectorArrayMatrixOperator, self).__init__(matrix, solver_options=solver_options, name=name)
if not vector:
self.source = VectorSpace(ListVectorArray, (NumpyVector, matrix.shape[1]))
if not functional:
self.range = VectorSpace(ListVectorArray, (NumpyVector, matrix.shape[0]))
self.functional = functional
self.vector = vector
def apply(self, U, ind=None, mu=None):
assert U in self.source
assert U.check_ind(ind)
if self.vector:
V = super(NumpyListVectorArrayMatrixOperator, self).apply(U, ind=ind, mu=mu)
return ListVectorArray([NumpyVector(v, copy=False) for v in V.data], subtype=self.range.subtype)
if ind is None:
vectors = U._list
elif isinstance(ind, Number):
vectors = [U._list[ind]]
else:
vectors = (U._list[i] for i in ind)
V = [self._matrix.dot(v._array) for v in vectors]
if self.functional:
return NumpyVectorArray(V) if len(V) > 0 else self.range.empty()
else:
return ListVectorArray([NumpyVector(v, copy=False) for v in V], subtype=self.range.subtype)
def apply_adjoint(self, U, ind=None, mu=None, source_product=None, range_product=None):
raise NotImplementedError
def apply_inverse(self, V, ind=None, mu=None, least_squares=False):
assert V in self.range
assert V.check_ind(ind)
assert not self.functional and not self.vector
if V.dim == 0:
if self.source.dim == 0 and least_squares:
return ListVectorArray(
[NumpyVector(np.zeros(0), copy=False) for _ in range(V.len_ind(ind))], subtype=self.source.subtype
)
else:
raise InversionError
options = (
self.solver_options.get("inverse") if self.solver_options else "least_squares" if least_squares else None
)
if options and not least_squares:
solver_type = options if isinstance(options, str) else options["type"]
if solver_type.startswith("least_squares"):
self.logger.warn('Least squares solver selected but "least_squares == False"')
if ind is None:
vectors = V._list
elif isinstance(ind, Number):
vectors = [V._list[ind]]
else:
vectors = (V._list[i] for i in ind)
try:
return ListVectorArray(
[
NumpyVector(
_apply_inverse(self._matrix, v._array.reshape((1, -1)), options=options).ravel(), copy=False
)
for v in vectors
],
subtype=self.source.subtype,
)
except InversionError as e:
if least_squares and options:
solver_type = options if isinstance(options, str) else options["type"]
if not solver_type.startswith("least_squares"):
msg = (
str(e)
+ "\nNote: linear solver was selected for solving least squares problem (maybe not invertible?)"
)
raise InversionError(msg)
raise e
def as_vector(self, mu=None):
if self.source.dim != 1 and self.range.dim != 1:
raise TypeError("This operator does not represent a vector or linear functional.")
return ListVectorArray([NumpyVector(self._matrix.ravel(), copy=True)])
def assemble_lincomb(self, operators, coefficients, solver_options=None, name=None):
lincomb = super(NumpyListVectorArrayMatrixOperator, self).assemble_lincomb(operators, coefficients)
if lincomb is None:
return None
else:
return NumpyListVectorArrayMatrixOperator(lincomb._matrix, solver_options=solver_options, name=name)
def NumpyVectorSpace(dim):
"""Shorthand for |VectorSpace| `(NumpyVectorArray, dim)`."""
return VectorSpace(NumpyVectorArray, dim)
class NumpyListVectorArrayMatrixOperator(NumpyMatrixOperator):
"""Variant of |NumpyMatrixOperator| using |ListVectorArray| instead of |NumpyVectorArray|."""
def __init__(self, matrix, functional=False, vector=False, solver_options=None, name=None):
assert not (functional and vector)
super().__init__(matrix, solver_options=solver_options, name=name)
if not vector:
self.source = VectorSpace(ListVectorArray, (NumpyVector, matrix.shape[1]))
if not functional:
self.range = VectorSpace(ListVectorArray, (NumpyVector, matrix.shape[0]))
self.functional = functional
self.vector = vector
def apply(self, U, ind=None, mu=None):
assert U in self.source
assert U.check_ind(ind)
if self.vector:
V = super().apply(U, ind=ind, mu=mu)
return ListVectorArray([NumpyVector(v, copy=False) for v in V.data],
subtype=self.range.subtype)
if ind is None:
vectors = U._list
elif isinstance(ind, Number):
vectors = [U._list[ind]]
else:
vectors = (U._list[i] for i in ind)
V = [self._matrix.dot(v._array) for v in vectors]
if self.functional:
return NumpyVectorArray(V) if len(V) > 0 else self.range.empty()
else:
return ListVectorArray([NumpyVector(v, copy=False) for v in V], subtype=self.range.subtype)
def apply_adjoint(self, U, ind=None, mu=None, source_product=None, range_product=None):
raise NotImplementedError
def apply_inverse(self, V, ind=None, mu=None, least_squares=False):
assert V in self.range
assert V.check_ind(ind)
assert not self.functional and not self.vector
if V.dim == 0:
if self.source.dim == 0 and least_squares:
return ListVectorArray([NumpyVector(np.zeros(0), copy=False) for _ in range(V.len_ind(ind))],
subtype=self.source.subtype)
else:
raise InversionError
options = (self.solver_options.get('inverse') if self.solver_options else
'least_squares' if least_squares else
None)
if options and not least_squares:
solver_type = options if isinstance(options, str) else options['type']
if solver_type.startswith('least_squares'):
self.logger.warn('Least squares solver selected but "least_squares == False"')
if ind is None:
vectors = V._list
elif isinstance(ind, Number):
vectors = [V._list[ind]]
else:
vectors = (V._list[i] for i in ind)
try:
return ListVectorArray([NumpyVector(_apply_inverse(self._matrix, v._array.reshape((1, -1)),
options=options).ravel(),
copy=False)
for v in vectors],
subtype=self.source.subtype)
except InversionError as e:
if least_squares and options:
solver_type = options if isinstance(options, str) else options['type']
if not solver_type.startswith('least_squares'):
msg = str(e) \
+ '\nNote: linear solver was selected for solving least squares problem (maybe not invertible?)'
raise InversionError(msg)
raise e
def as_vector(self, mu=None):
if self.source.dim != 1 and self.range.dim != 1:
raise TypeError('This operator does not represent a vector or linear functional.')
return ListVectorArray([NumpyVector(self._matrix.ravel(), copy=True)])
def assemble_lincomb(self, operators, coefficients, solver_options=None, name=None):
lincomb = super().assemble_lincomb(operators, coefficients)
if lincomb is None:
return None
else:
return NumpyListVectorArrayMatrixOperator(lincomb._matrix, solver_options=solver_options, name=name)
class NumpyListVectorArrayMatrixOperator(NumpyMatrixOperator):
"""Variant of |NumpyMatrixOperator| using |ListVectorArray| instead of |NumpyVectorArray|."""
def __init__(self, matrix, functional=False, vector=False, name=None):
assert not (functional and vector)
super(NumpyListVectorArrayMatrixOperator, self).__init__(matrix, name)
if not vector:
self.source = VectorSpace(ListVectorArray, (NumpyVector, matrix.shape[1]))
if not functional:
self.range = VectorSpace(ListVectorArray, (NumpyVector, matrix.shape[0]))
self.functional = functional
self.vector = vector
def apply(self, U, ind=None, mu=None):
assert U in self.source
assert U.check_ind(ind)
if self.vector:
V = super(NumpyListVectorArrayMatrixOperator, self).apply(U, ind=ind, mu=mu)
return ListVectorArray([NumpyVector(v, copy=False) for v in V.data],
subtype=self.range.subtype)
if ind is None:
vectors = U._list
elif isinstance(ind, Number):
vectors = [U._list[ind]]
else:
vectors = (U._list[i] for i in ind)
V = [self._matrix.dot(v._array) for v in vectors]
if self.functional:
return NumpyVectorArray(V) if len(V) > 0 else self.range.empty()
else:
return ListVectorArray([NumpyVector(v, copy=False) for v in V], subtype=self.range.subtype)
def apply_adjoint(self, U, ind=None, mu=None, source_product=None, range_product=None):
raise NotImplementedError
def apply_inverse(self, U, ind=None, mu=None, options=None):
assert U in self.range
assert U.check_ind(ind)
assert not self.functional and not self.vector
if U.dim == 0:
if (self.source.dim == 0
or isinstance(options, str) and options.startswith('least_squares')
or isinstance(options, dict) and options['type'].startswith('least_squares')):
return ListVectorArray([NumpyVector(np.zeros(0), copy=False) for _ in range(U.len_ind(ind))],
subtype=self.source.subtype)
else:
raise InversionError
if ind is None:
vectors = U._list
elif isinstance(ind, Number):
vectors = [U._list[ind]]
else:
vectors = (U._list[i] for i in ind)
return ListVectorArray([NumpyVector(_apply_inverse(self._matrix, v._array, options=options), copy=False)
for v in vectors],
subtype=self.source.subtype)
def as_vector(self, mu=None):
if self.source.dim != 1 and self.range.dim != 1:
raise TypeError('This operator does not represent a vector or linear functional.')
return ListVectorArray([NumpyVector(self._matrix.ravel(), copy=True)])
def assemble_lincomb(self, operators, coefficients, name=None):
lincomb = super(NumpyListVectorArrayMatrixOperator, self).assemble_lincomb(operators, coefficients)
if lincomb is None:
return None
else:
return NumpyListVectorArrayMatrixOperator(lincomb._matrix, name=name)