def test_arrayexpr_convert_indexed_to_array_and_back_to_matrix():
expr = a.T * b
elem = expr[0, 0]
cg = convert_indexed_to_array(elem)
assert cg == ArrayElement(
ArrayContraction(ArrayTensorProduct(a, b), (0, 2)), [0, 0])
expr = M[i, j] + N[i, j]
p1, p2 = _convert_indexed_to_array(expr)
assert convert_array_to_matrix(p1) == M + N
expr = M[i, j] + N[j, i]
p1, p2 = _convert_indexed_to_array(expr)
assert convert_array_to_matrix(p1) == M + N.T
expr = M[i, j] * N[k, l] + N[i, j] * M[k, l]
p1, p2 = _convert_indexed_to_array(expr)
assert convert_array_to_matrix(p1) == ArrayAdd(ArrayTensorProduct(M, N),
ArrayTensorProduct(N, M))
expr = (M * N * P)[i, j]
p1, p2 = _convert_indexed_to_array(expr)
assert convert_array_to_matrix(p1) == M * N * P
expr = Sum(M[i, j] * (N * P)[j, m], (j, 0, k - 1))
p1, p2 = _convert_indexed_to_array(expr)
assert convert_array_to_matrix(p1) == M * N * P
expr = Sum((P[j, m] + P[m, j]) * (M[i, j] * N[m, n] + N[i, j] * M[m, n]),
(j, 0, k - 1), (m, 0, k - 1))
p1, p2 = _convert_indexed_to_array(expr)
assert convert_array_to_matrix(
p1) == M * P * N + M * P.T * N + N * P * M + N * P.T * M
def test_arrayexpr_convert_array_element_to_array_expression():
A = ArraySymbol("A", (k, ))
B = ArraySymbol("B", (k, ))
s = Sum(A[i] * B[i], (i, 0, k - 1))
cg = convert_indexed_to_array(s)
assert cg == ArrayContraction(ArrayTensorProduct(A, B), (0, 1))
s = A[i] * B[i]
cg = convert_indexed_to_array(s)
assert cg == ArrayDiagonal(ArrayTensorProduct(A, B), (0, 1))
s = A[i] * B[j]
cg = convert_indexed_to_array(s, [i, j])
assert cg == ArrayTensorProduct(A, B)
cg = convert_indexed_to_array(s, [j, i])
assert cg == ArrayTensorProduct(B, A)
def test_arrayexpr_convert_indexed_to_array_broadcast():
A = ArraySymbol("A", (3, 3))
B = ArraySymbol("B", (3, 3))
expr = A[i, j] + B[k, l]
O2 = OneArray(3, 3)
expected = ArrayAdd(ArrayTensorProduct(A, O2), ArrayTensorProduct(O2, B))
assert convert_indexed_to_array(expr) == expected
assert convert_indexed_to_array(expr, [i, j, k, l]) == expected
assert convert_indexed_to_array(expr, [l, k, i, j]) == ArrayAdd(
PermuteDims(ArrayTensorProduct(O2, A), [1, 0, 2, 3]),
PermuteDims(ArrayTensorProduct(B, O2), [1, 0, 2, 3]))
expr = A[i, j] + B[j, k]
O1 = OneArray(3)
assert convert_indexed_to_array(expr, [i, j, k]) == ArrayAdd(
ArrayTensorProduct(A, O1), ArrayTensorProduct(O1, B))
C = ArraySymbol("C", (d0, d1))
D = ArraySymbol("D", (d3, d1))
expr = C[i, j] + D[k, j]
assert convert_indexed_to_array(expr, [i, j, k]) == ArrayAdd(
ArrayTensorProduct(C, OneArray(d3)),
PermuteDims(ArrayTensorProduct(OneArray(d0), D), [0, 2, 1]))
X = ArraySymbol("X", (5, 3))
expr = X[i, n] - X[j, n]
assert convert_indexed_to_array(expr, [i, j, n]) == ArrayAdd(
ArrayTensorProduct(-1, OneArray(5), X),
PermuteDims(ArrayTensorProduct(X, OneArray(5)), [0, 2, 1]))
raises(ValueError, lambda: convert_indexed_to_array(C[i, j] + D[i, j]))
def from_index_summation(expr,
first_index=None,
last_index=None,
dimensions=None):
r"""
Parse expression of matrices with explicitly summed indices into a
matrix expression without indices, if possible.
This transformation expressed in mathematical notation:
`\sum_{j=0}^{N-1} A_{i,j} B_{j,k} \Longrightarrow \mathbf{A}\cdot \mathbf{B}`
Optional parameter ``first_index``: specify which free index to use as
the index starting the expression.
Examples
========
>>> from sympy import MatrixSymbol, MatrixExpr, Sum
>>> from sympy.abc import i, j, k, l, N
>>> A = MatrixSymbol("A", N, N)
>>> B = MatrixSymbol("B", N, N)
>>> expr = Sum(A[i, j]*B[j, k], (j, 0, N-1))
>>> MatrixExpr.from_index_summation(expr)
A*B
Transposition is detected:
>>> expr = Sum(A[j, i]*B[j, k], (j, 0, N-1))
>>> MatrixExpr.from_index_summation(expr)
A.T*B
Detect the trace:
>>> expr = Sum(A[i, i], (i, 0, N-1))
>>> MatrixExpr.from_index_summation(expr)
Trace(A)
More complicated expressions:
>>> expr = Sum(A[i, j]*B[k, j]*A[l, k], (j, 0, N-1), (k, 0, N-1))
>>> MatrixExpr.from_index_summation(expr)
A*B.T*A.T
"""
from sympy.tensor.array.expressions.conv_indexed_to_array import convert_indexed_to_array
from sympy.tensor.array.expressions.conv_array_to_matrix import convert_array_to_matrix
first_indices = []
if first_index is not None:
first_indices.append(first_index)
if last_index is not None:
first_indices.append(last_index)
arr = convert_indexed_to_array(expr, first_indices=first_indices)
return convert_array_to_matrix(arr)
def test_arrayexpr_convert_indexed_to_array_expression():
s = Sum(A[i] * B[i], (i, 0, 3))
cg = convert_indexed_to_array(s)
assert cg == ArrayContraction(ArrayTensorProduct(A, B), (0, 1))
expr = M * N
result = ArrayContraction(ArrayTensorProduct(M, N), (1, 2))
elem = expr[i, j]
assert convert_indexed_to_array(elem) == result
expr = M * N * M
elem = expr[i, j]
result = _array_contraction(_array_tensor_product(M, M, N), (1, 4), (2, 5))
cg = convert_indexed_to_array(elem)
assert cg == result
cg = convert_indexed_to_array((M * N * P)[i, j])
assert cg == _array_contraction(ArrayTensorProduct(M, N, P), (1, 2),
(3, 4))
cg = convert_indexed_to_array((M * N.T * P)[i, j])
assert cg == _array_contraction(ArrayTensorProduct(M, N, P), (1, 3),
(2, 4))
expr = -2 * M * N
elem = expr[i, j]
cg = convert_indexed_to_array(elem)
assert cg == ArrayContraction(ArrayTensorProduct(-2, M, N), (1, 2))
def test_arrayexpr_convert_indexed_to_array_out_of_bounds():
expr = Sum(M[i, i], (i, 0, 4))
raises(ValueError, lambda: convert_indexed_to_array(expr))
expr = Sum(M[i, i], (i, 0, k))
raises(ValueError, lambda: convert_indexed_to_array(expr))
expr = Sum(M[i, i], (i, 1, k - 1))
raises(ValueError, lambda: convert_indexed_to_array(expr))
expr = Sum(M[i, j] * N[j, m], (j, 0, 4))
raises(ValueError, lambda: convert_indexed_to_array(expr))
expr = Sum(M[i, j] * N[j, m], (j, 0, k))
raises(ValueError, lambda: convert_indexed_to_array(expr))
expr = Sum(M[i, j] * N[j, m], (j, 1, k - 1))
raises(ValueError, lambda: convert_indexed_to_array(expr))
def _arrayify(self, indexed):
from sympy.tensor.array.expressions.conv_indexed_to_array import convert_indexed_to_array
try:
return convert_indexed_to_array(indexed)
except Exception:
return indexed
