lattice/pyerrors
1
0
Fork 0
mirror of https://github.com/fjosw/pyerrors.git synced 2025-05-15 12:03:42 +02:00
Code Issues Projects Releases Packages Wiki Activity

fix: linalg.jack_matmul now works with arrays of arbitrary shape

Browse source
This commit is contained in:
Fabian Joswig 2021-11-30 11:52:53 +00:00
parent 443e4eb74e
commit 1fa3bc1291
1 changed files with 16 additions and 16 deletions
Download patch file Download diff file Expand all files Collapse all files

32
pyerrors/linalg.py
View file

@ -186,49 +186,49 @@ def jack_matmul(*operands):
For large matrices this is considerably faster compared to matmul. For large matrices this is considerably faster compared to matmul.
""" """
if any(isinstance(o[0, 0], CObs) for o in operands): if any(isinstance(o.flat[0], CObs) for o in operands):
name = operands[0][0, 0].real.names[0] name = operands[0].flat[0].real.names[0]
idl = operands[0][0, 0].real.idl[name] idl = operands[0].flat[0].real.idl[name]
def _exp_to_jack(matrix): def _exp_to_jack(matrix):
base_matrix = np.empty_like(matrix) base_matrix = np.empty_like(matrix)
for (n, m), entry in np.ndenumerate(matrix): for index, entry in np.ndenumerate(matrix):
base_matrix[n, m] = entry.real.export_jackknife() + 1j * entry.imag.export_jackknife() base_matrix[index] = entry.real.export_jackknife() + 1j * entry.imag.export_jackknife()
return base_matrix return base_matrix
def _imp_from_jack(matrix): def _imp_from_jack(matrix):
base_matrix = np.empty_like(matrix) base_matrix = np.empty_like(matrix)
for (n, m), entry in np.ndenumerate(matrix): for index, entry in np.ndenumerate(matrix):
base_matrix[n, m] = CObs(import_jackknife(entry.real, name, [idl]), base_matrix[index] = CObs(import_jackknife(entry.real, name, [idl]),
import_jackknife(entry.imag, name, [idl])) import_jackknife(entry.imag, name, [idl]))
return base_matrix return base_matrix
r = _exp_to_jack(operands[0]) r = _exp_to_jack(operands[0])
for op in operands[1:]: for op in operands[1:]:
if isinstance(op[0, 0], CObs): if isinstance(op.flat[0], CObs):
r = r @ _exp_to_jack(op) r = r @ _exp_to_jack(op)
else: else:
r = r @ op r = r @ op
return _imp_from_jack(r) return _imp_from_jack(r)
else: else:
name = operands[0][0, 0].names[0] name = operands[0].flat[0].names[0]
idl = operands[0][0, 0].idl[name] idl = operands[0].flat[0].idl[name]
def _exp_to_jack(matrix): def _exp_to_jack(matrix):
base_matrix = np.empty_like(matrix) base_matrix = np.empty_like(matrix)
for (n, m), entry in np.ndenumerate(matrix): for index, entry in np.ndenumerate(matrix):
base_matrix[n, m] = entry.export_jackknife() base_matrix[index] = entry.export_jackknife()
return base_matrix return base_matrix
def _imp_from_jack(matrix): def _imp_from_jack(matrix):
base_matrix = np.empty_like(matrix) base_matrix = np.empty_like(matrix)
for (n, m), entry in np.ndenumerate(matrix): for index, entry in np.ndenumerate(matrix):
base_matrix[n, m] = import_jackknife(entry, name, [idl]) base_matrix[index] = import_jackknife(entry, name, [idl])
return base_matrix return base_matrix
r = _exp_to_jack(operands[0]) r = _exp_to_jack(operands[0])
for op in operands[1:]: for op in operands[1:]:
if isinstance(op[0, 0], Obs): if isinstance(op.flat[0], Obs):
r = r @ _exp_to_jack(op) r = r @ _exp_to_jack(op)
else: else:
r = r @ op r = r @ op