Skip to content

Adds option to weight nonlinear fitting in emittance calc. #371

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
dylanmkennedy wants to merge 2 commits into
base: main
Choose a base branch
from
Open

Adds option to weight nonlinear fitting in emittance calc. #371

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
9a84e85
Adds option to weight nonlinear fitting in emittance calc by inverse …
dylanmkennedy May 19, 2026
c01fe5f
Black formatting.
dylanmkennedy May 19, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
19 changes: 17 additions & 2 deletions lcls_tools/common/data/emittance.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,6 +11,7 @@ def compute_emit_bmag(
beamsize_squared: np.ndarray,
rmat: np.ndarray,
twiss_design: np.ndarray = None,
weighted_fit: bool = False,
maxiter: int = None,
):
"""
Expand Down Expand Up @@ -40,6 +41,10 @@ def compute_emit_bmag(
to each measurement in the respective batch for the calculation of Bmag
(useful for quad scans).

weighted_fit : bool, default = False
Boolean specifying whether or not to weight the fitting loss function by the inverse beam
size. If True, smaller beam sizes will be prioritized in the non-linear curve fitting.

maxiter : int, optional
Maximum number of iterations to perform in nonlinear fitting (minimization algorithm).

Expand Down Expand Up @@ -98,8 +103,14 @@ def loss_torch(params):
params = torch.reshape(params, [*beamsize_squared.shape[:-2], 3])
sig = torch.stack(beam_matrix_tuple(params), dim=-1).unsqueeze(-1)
# sig should now be shape batchshape x 3 x 1 (column vectors)
if weighted_fit:
weights = 1.0 / beamsize_squared.sqrt()
else:
weights = 1.0
total_abs_error = (
(torch.sqrt(amat @ sig) - torch.sqrt(beamsize_squared)).abs().nansum()
(weights * (torch.sqrt(amat @ sig) - torch.sqrt(beamsize_squared)))
.abs()
.nansum()
)
return total_abs_error

Expand All @@ -119,8 +130,12 @@ def loss(params):
params = np.reshape(params, [*beamsize_squared.shape[:-2], 3])
sig = np.expand_dims(np.stack(beam_matrix_tuple(params), axis=-1), axis=-1)
# sig should now be shape batchshape x 3 x 1 (column vectors)
if weighted_fit:
weights = 1.0 / np.sqrt(beamsize_squared)
else:
weights = 1.0
total_abs_error = np.nansum(
np.abs(np.sqrt(amat @ sig) - np.sqrt(beamsize_squared))
np.abs(weights * (np.sqrt(amat @ sig) - np.sqrt(beamsize_squared)))
)
return total_abs_error

Expand Down
103 changes: 103 additions & 0 deletions tests/unit_tests/lcls_tools/common/data/test_emittance.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -54,6 +54,109 @@ def test_emittance_calc(self):
result["bmag"], np.array([[1.0, 1.0, 1.0], [1.0, 1.0, 1.0]]), rtol=1e-2
)

def test_emittance_calc_with_weighted_fit(self):
# set up test values
beamsize_squared = np.array(
[
[[0.00299705], [0.12034662], [0.03792509]],
[[0.00259244], [0.23703413], [0.09598636]],
]
)

rmat = np.array(
[
[
[[-0.24254679, 1.04600053], [-1.17140665, 0.92885986]],
[[-1.16646303, 0.71788384], [-0.8039517, -0.36251133]],
[[-0.55801204, -0.55330746], [0.61964408, -1.17765613]],
],
[
[[-0.24254679, 1.04600053], [-1.17140665, 0.92885986]],
[[-1.16646303, 0.71788384], [-0.8039517, -0.36251133]],
[[-0.55801204, -0.55330746], [0.61964408, -1.17765613]],
],
]
)

twiss_design = np.array(
[
[
[0.29970473, -1.58494282],
[12.03466156, -9.17146332],
[3.79250871, 3.01307483],
],
[
[0.1296221, -0.66578778],
[11.85170667, -9.88871364],
[4.79931781, 2.87869131],
],
]
)

# compute emittance & bmag
result = compute_emit_bmag(
beamsize_squared=beamsize_squared,
rmat=rmat,
twiss_design=twiss_design,
weighted_fit=True,
)

# compare results with ground-truth
assert np.allclose(result["emittance"], np.array([[0.01], [0.02]]), rtol=1e-2)
assert np.allclose(
result["bmag"], np.array([[1.0, 1.0, 1.0], [1.0, 1.0, 1.0]]), rtol=1e-2
)

def test_emittance_calc_with_nans(self):
# set up test values
beamsize_squared = np.array(
[
[[0.00299705], [0.12034662], [0.03792509]],
[[0.00259244], [0.23703413], [0.09598636]],
]
)

rmat = np.array(
[
[
[[-0.24254679, 1.04600053], [-1.17140665, 0.92885986]],
[[-1.16646303, 0.71788384], [-0.8039517, -0.36251133]],
[[-0.55801204, -0.55330746], [0.61964408, -1.17765613]],
],
[
[[-0.24254679, 1.04600053], [-1.17140665, 0.92885986]],
[[-1.16646303, 0.71788384], [-0.8039517, -0.36251133]],
[[-0.55801204, -0.55330746], [0.61964408, -1.17765613]],
],
]
)

twiss_design = np.array(
[
[
[0.29970473, -1.58494282],
[12.03466156, -9.17146332],
[3.79250871, 3.01307483],
],
[
[0.1296221, -0.66578778],
[11.85170667, -9.88871364],
[4.79931781, 2.87869131],
],
]
)

# compute emittance & bmag
result = compute_emit_bmag(
beamsize_squared=beamsize_squared, rmat=rmat, twiss_design=twiss_design
)

# compare results with ground-truth
assert np.allclose(result["emittance"], np.array([[0.01], [0.02]]), rtol=1e-2)
assert np.allclose(
result["bmag"], np.array([[1.0, 1.0, 1.0], [1.0, 1.0, 1.0]]), rtol=1e-2
)

def test_emittance_calc_with_broadcast_twiss(self):
# set up test values
beamsize_squared = np.array(
Expand Down
Loading