You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

499 lines
15 KiB

from sympy.assumptions.ask import ask, Q
from sympy.assumptions.refine import handlers_dict
from sympy.core import Basic, sympify, S
from sympy.core.mul import mul, Mul
from sympy.core.numbers import Number, Integer
from sympy.core.symbol import Dummy
from sympy.functions import adjoint
from sympy.strategies import (rm_id, unpack, typed, flatten, exhaust,
do_one, new)
from sympy.matrices.common import NonInvertibleMatrixError
from sympy.matrices.matrices import MatrixBase
from sympy.utilities.exceptions import sympy_deprecation_warning
from sympy.matrices.expressions._shape import validate_matmul_integer as validate
from .inverse import Inverse
from .matexpr import MatrixExpr
from .matpow import MatPow
from .transpose import transpose
from .permutation import PermutationMatrix
from .special import ZeroMatrix, Identity, GenericIdentity, OneMatrix
# XXX: MatMul should perhaps not subclass directly from Mul
class MatMul(MatrixExpr, Mul):
"""
A product of matrix expressions
Examples
========
>>> from sympy import MatMul, MatrixSymbol
>>> A = MatrixSymbol('A', 5, 4)
>>> B = MatrixSymbol('B', 4, 3)
>>> C = MatrixSymbol('C', 3, 6)
>>> MatMul(A, B, C)
A*B*C
"""
is_MatMul = True
identity = GenericIdentity()
def __new__(cls, *args, evaluate=False, check=None, _sympify=True):
if not args:
return cls.identity
# This must be removed aggressively in the constructor to avoid
# TypeErrors from GenericIdentity().shape
args = list(filter(lambda i: cls.identity != i, args))
if _sympify:
args = list(map(sympify, args))
obj = Basic.__new__(cls, *args)
factor, matrices = obj.as_coeff_matrices()
if check is not None:
sympy_deprecation_warning(
"Passing check to MatMul is deprecated and the check argument will be removed in a future version.",
deprecated_since_version="1.11",
active_deprecations_target='remove-check-argument-from-matrix-operations')
if check is not False:
validate(*matrices)
if not matrices:
# Should it be
#
# return Basic.__neq__(cls, factor, GenericIdentity()) ?
return factor
if evaluate:
return cls._evaluate(obj)
return obj
@classmethod
def _evaluate(cls, expr):
return canonicalize(expr)
@property
def shape(self):
matrices = [arg for arg in self.args if arg.is_Matrix]
return (matrices[0].rows, matrices[-1].cols)
def _entry(self, i, j, expand=True, **kwargs):
# Avoid cyclic imports
from sympy.concrete.summations import Sum
from sympy.matrices.immutable import ImmutableMatrix
coeff, matrices = self.as_coeff_matrices()
if len(matrices) == 1: # situation like 2*X, matmul is just X
return coeff * matrices[0][i, j]
indices = [None]*(len(matrices) + 1)
ind_ranges = [None]*(len(matrices) - 1)
indices[0] = i
indices[-1] = j
def f():
counter = 1
while True:
yield Dummy("i_%i" % counter)
counter += 1
dummy_generator = kwargs.get("dummy_generator", f())
for i in range(1, len(matrices)):
indices[i] = next(dummy_generator)
for i, arg in enumerate(matrices[:-1]):
ind_ranges[i] = arg.shape[1] - 1
matrices = [arg._entry(indices[i], indices[i+1], dummy_generator=dummy_generator) for i, arg in enumerate(matrices)]
expr_in_sum = Mul.fromiter(matrices)
if any(v.has(ImmutableMatrix) for v in matrices):
expand = True
result = coeff*Sum(
expr_in_sum,
*zip(indices[1:-1], [0]*len(ind_ranges), ind_ranges)
)
# Don't waste time in result.doit() if the sum bounds are symbolic
if not any(isinstance(v, (Integer, int)) for v in ind_ranges):
expand = False
return result.doit() if expand else result
def as_coeff_matrices(self):
scalars = [x for x in self.args if not x.is_Matrix]
matrices = [x for x in self.args if x.is_Matrix]
coeff = Mul(*scalars)
if coeff.is_commutative is False:
raise NotImplementedError("noncommutative scalars in MatMul are not supported.")
return coeff, matrices
def as_coeff_mmul(self):
coeff, matrices = self.as_coeff_matrices()
return coeff, MatMul(*matrices)
def expand(self, **kwargs):
expanded = super(MatMul, self).expand(**kwargs)
return self._evaluate(expanded)
def _eval_transpose(self):
"""Transposition of matrix multiplication.
Notes
=====
The following rules are applied.
Transposition for matrix multiplied with another matrix:
`\\left(A B\\right)^{T} = B^{T} A^{T}`
Transposition for matrix multiplied with scalar:
`\\left(c A\\right)^{T} = c A^{T}`
References
==========
.. [1] https://en.wikipedia.org/wiki/Transpose
"""
coeff, matrices = self.as_coeff_matrices()
return MatMul(
coeff, *[transpose(arg) for arg in matrices[::-1]]).doit()
def _eval_adjoint(self):
return MatMul(*[adjoint(arg) for arg in self.args[::-1]]).doit()
def _eval_trace(self):
factor, mmul = self.as_coeff_mmul()
if factor != 1:
from .trace import trace
return factor * trace(mmul.doit())
else:
raise NotImplementedError("Can't simplify any further")
def _eval_determinant(self):
from sympy.matrices.expressions.determinant import Determinant
factor, matrices = self.as_coeff_matrices()
square_matrices = only_squares(*matrices)
return factor**self.rows * Mul(*list(map(Determinant, square_matrices)))
def _eval_inverse(self):
if all(arg.is_square for arg in self.args if isinstance(arg, MatrixExpr)):
return MatMul(*(
arg.inverse() if isinstance(arg, MatrixExpr) else arg**-1
for arg in self.args[::-1]
)
).doit()
return Inverse(self)
def doit(self, **hints):
deep = hints.get('deep', True)
if deep:
args = tuple(arg.doit(**hints) for arg in self.args)
else:
args = self.args
# treat scalar*MatrixSymbol or scalar*MatPow separately
expr = canonicalize(MatMul(*args))
return expr
# Needed for partial compatibility with Mul
def args_cnc(self, cset=False, warn=True, **kwargs):
coeff_c = [x for x in self.args if x.is_commutative]
coeff_nc = [x for x in self.args if not x.is_commutative]
if cset:
clen = len(coeff_c)
coeff_c = set(coeff_c)
if clen and warn and len(coeff_c) != clen:
raise ValueError('repeated commutative arguments: %s' %
[ci for ci in coeff_c if list(self.args).count(ci) > 1])
return [coeff_c, coeff_nc]
def _eval_derivative_matrix_lines(self, x):
from .transpose import Transpose
with_x_ind = [i for i, arg in enumerate(self.args) if arg.has(x)]
lines = []
for ind in with_x_ind:
left_args = self.args[:ind]
right_args = self.args[ind+1:]
if right_args:
right_mat = MatMul.fromiter(right_args)
else:
right_mat = Identity(self.shape[1])
if left_args:
left_rev = MatMul.fromiter([Transpose(i).doit() if i.is_Matrix else i for i in reversed(left_args)])
else:
left_rev = Identity(self.shape[0])
d = self.args[ind]._eval_derivative_matrix_lines(x)
for i in d:
i.append_first(left_rev)
i.append_second(right_mat)
lines.append(i)
return lines
mul.register_handlerclass((Mul, MatMul), MatMul)
# Rules
def newmul(*args):
if args[0] == 1:
args = args[1:]
return new(MatMul, *args)
def any_zeros(mul):
if any(arg.is_zero or (arg.is_Matrix and arg.is_ZeroMatrix)
for arg in mul.args):
matrices = [arg for arg in mul.args if arg.is_Matrix]
return ZeroMatrix(matrices[0].rows, matrices[-1].cols)
return mul
def merge_explicit(matmul):
""" Merge explicit MatrixBase arguments
>>> from sympy import MatrixSymbol, Matrix, MatMul, pprint
>>> from sympy.matrices.expressions.matmul import merge_explicit
>>> A = MatrixSymbol('A', 2, 2)
>>> B = Matrix([[1, 1], [1, 1]])
>>> C = Matrix([[1, 2], [3, 4]])
>>> X = MatMul(A, B, C)
>>> pprint(X)
[1 1] [1 2]
A*[ ]*[ ]
[1 1] [3 4]
>>> pprint(merge_explicit(X))
[4 6]
A*[ ]
[4 6]
>>> X = MatMul(B, A, C)
>>> pprint(X)
[1 1] [1 2]
[ ]*A*[ ]
[1 1] [3 4]
>>> pprint(merge_explicit(X))
[1 1] [1 2]
[ ]*A*[ ]
[1 1] [3 4]
"""
if not any(isinstance(arg, MatrixBase) for arg in matmul.args):
return matmul
newargs = []
last = matmul.args[0]
for arg in matmul.args[1:]:
if isinstance(arg, (MatrixBase, Number)) and isinstance(last, (MatrixBase, Number)):
last = last * arg
else:
newargs.append(last)
last = arg
newargs.append(last)
return MatMul(*newargs)
def remove_ids(mul):
""" Remove Identities from a MatMul
This is a modified version of sympy.strategies.rm_id.
This is necesssary because MatMul may contain both MatrixExprs and Exprs
as args.
See Also
========
sympy.strategies.rm_id
"""
# Separate Exprs from MatrixExprs in args
factor, mmul = mul.as_coeff_mmul()
# Apply standard rm_id for MatMuls
result = rm_id(lambda x: x.is_Identity is True)(mmul)
if result != mmul:
return newmul(factor, *result.args) # Recombine and return
else:
return mul
def factor_in_front(mul):
factor, matrices = mul.as_coeff_matrices()
if factor != 1:
return newmul(factor, *matrices)
return mul
def combine_powers(mul):
r"""Combine consecutive powers with the same base into one, e.g.
$$A \times A^2 \Rightarrow A^3$$
This also cancels out the possible matrix inverses using the
knowledgebase of :class:`~.Inverse`, e.g.,
$$ Y \times X \times X^{-1} \Rightarrow Y $$
"""
factor, args = mul.as_coeff_matrices()
new_args = [args[0]]
for i in range(1, len(args)):
A = new_args[-1]
B = args[i]
if isinstance(B, Inverse) and isinstance(B.arg, MatMul):
Bargs = B.arg.args
l = len(Bargs)
if list(Bargs) == new_args[-l:]:
new_args = new_args[:-l] + [Identity(B.shape[0])]
continue
if isinstance(A, Inverse) and isinstance(A.arg, MatMul):
Aargs = A.arg.args
l = len(Aargs)
if list(Aargs) == args[i:i+l]:
identity = Identity(A.shape[0])
new_args[-1] = identity
for j in range(i, i+l):
args[j] = identity
continue
if A.is_square == False or B.is_square == False:
new_args.append(B)
continue
if isinstance(A, MatPow):
A_base, A_exp = A.args
else:
A_base, A_exp = A, S.One
if isinstance(B, MatPow):
B_base, B_exp = B.args
else:
B_base, B_exp = B, S.One
if A_base == B_base:
new_exp = A_exp + B_exp
new_args[-1] = MatPow(A_base, new_exp).doit(deep=False)
continue
elif not isinstance(B_base, MatrixBase):
try:
B_base_inv = B_base.inverse()
except NonInvertibleMatrixError:
B_base_inv = None
if B_base_inv is not None and A_base == B_base_inv:
new_exp = A_exp - B_exp
new_args[-1] = MatPow(A_base, new_exp).doit(deep=False)
continue
new_args.append(B)
return newmul(factor, *new_args)
def combine_permutations(mul):
"""Refine products of permutation matrices as the products of cycles.
"""
args = mul.args
l = len(args)
if l < 2:
return mul
result = [args[0]]
for i in range(1, l):
A = result[-1]
B = args[i]
if isinstance(A, PermutationMatrix) and \
isinstance(B, PermutationMatrix):
cycle_1 = A.args[0]
cycle_2 = B.args[0]
result[-1] = PermutationMatrix(cycle_1 * cycle_2)
else:
result.append(B)
return MatMul(*result)
def combine_one_matrices(mul):
"""
Combine products of OneMatrix
e.g. OneMatrix(2, 3) * OneMatrix(3, 4) -> 3 * OneMatrix(2, 4)
"""
factor, args = mul.as_coeff_matrices()
new_args = [args[0]]
for B in args[1:]:
A = new_args[-1]
if not isinstance(A, OneMatrix) or not isinstance(B, OneMatrix):
new_args.append(B)
continue
new_args.pop()
new_args.append(OneMatrix(A.shape[0], B.shape[1]))
factor *= A.shape[1]
return newmul(factor, *new_args)
def distribute_monom(mul):
"""
Simplify MatMul expressions but distributing
rational term to MatMul.
e.g. 2*(A+B) -> 2*A + 2*B
"""
args = mul.args
if len(args) == 2:
from .matadd import MatAdd
if args[0].is_MatAdd and args[1].is_Rational:
return MatAdd(*[MatMul(mat, args[1]).doit() for mat in args[0].args])
if args[1].is_MatAdd and args[0].is_Rational:
return MatAdd(*[MatMul(args[0], mat).doit() for mat in args[1].args])
return mul
rules = (
distribute_monom, any_zeros, remove_ids, combine_one_matrices, combine_powers, unpack, rm_id(lambda x: x == 1),
merge_explicit, factor_in_front, flatten, combine_permutations)
canonicalize = exhaust(typed({MatMul: do_one(*rules)}))
def only_squares(*matrices):
"""factor matrices only if they are square"""
if matrices[0].rows != matrices[-1].cols:
raise RuntimeError("Invalid matrices being multiplied")
out = []
start = 0
for i, M in enumerate(matrices):
if M.cols == matrices[start].rows:
out.append(MatMul(*matrices[start:i+1]).doit())
start = i+1
return out
def refine_MatMul(expr, assumptions):
"""
>>> from sympy import MatrixSymbol, Q, assuming, refine
>>> X = MatrixSymbol('X', 2, 2)
>>> expr = X * X.T
>>> print(expr)
X*X.T
>>> with assuming(Q.orthogonal(X)):
... print(refine(expr))
I
"""
newargs = []
exprargs = []
for args in expr.args:
if args.is_Matrix:
exprargs.append(args)
else:
newargs.append(args)
last = exprargs[0]
for arg in exprargs[1:]:
if arg == last.T and ask(Q.orthogonal(arg), assumptions):
last = Identity(arg.shape[0])
elif arg == last.conjugate() and ask(Q.unitary(arg), assumptions):
last = Identity(arg.shape[0])
else:
newargs.append(last)
last = arg
newargs.append(last)
return MatMul(*newargs)
handlers_dict['MatMul'] = refine_MatMul