update global omp interface by beykyle · Pull Request #76 · beykyle/jitr

beykyle · 2026-03-04T06:12:21Z

No description provided.

Copilot

Pull request overview

This PR updates the optical model potential (OMP) interface by introducing a shared single-channel OMP base class, refactoring existing global OMP implementations to use it, and extending elastic-scattering workspaces to accept a separate Coulomb interaction term. Example notebooks are updated to match the new calling patterns.

Changes:

Add SingleChannelOpticalModel (and LocalOpticalPotential) to standardize central/spin-orbit/Coulomb term handling across OMPs.
Refactor WLH, KDUQ, and CHUQ modules to expose new model classes and return “params-by-term” tuples.
Extend IntegralWorkspace / DifferentialWorkspace APIs to accept optional Coulomb interactions and update notebooks accordingly.

Reviewed changes

Copilot reviewed 8 out of 15 changed files in this pull request and generated 9 comments.

Show a summary per file

File	Description
`src/jitr/xs/elastic.py`	Adds optional Coulomb interaction/args to S-matrix and XS computations; adds docstrings.
`src/jitr/optical_potentials/omp.py`	Introduces a base single-channel OMP interface and a local Woods–Saxon OMP implementation.
`src/jitr/optical_potentials/wlh.py`	Refactors WLH potential into a `SingleChannelOpticalModel` subclass; adjusts parameter helpers/returns.
`src/jitr/optical_potentials/kduq.py`	Refactors KDUQ potential similarly; consolidates posterior sampling and adjusts returns.
`src/jitr/optical_potentials/chuq.py`	Refactors CHUQ potential similarly; consolidates posterior sampling and adjusts returns.
`src/jitr/optical_potentials/__init__.py`	Exports the new OMP base classes and keeps module imports.
`examples/notebooks/mass_exploration.ipynb`	Updates notebook usage for new KDUQ sampling/parameter ordering and Coulomb handling.
`examples/notebooks/comparison_to_Runge_Kutta.ipynb`	Adds a new performance/validation notebook.
`examples/notebooks/chex_jitr_validation.ipynb`	Updates validation notebook to use new OMP interface and parameter passing.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-03-04T06:19:58Z

src/jitr/xs/elastic.py

    def xs(
        self,
        interaction_central,
        interaction_spin_orbit,
+        interaction_coulomb=None,
        args_central=None,
        args_spin_orbit=None,
-        angles=None,
+        args_coulomb=None,
    ):
        r"""
        returns the angle-integrated total, elastic and reaction cross sections in mb
        """
        splus, sminus = self.smatrix(
            interaction_central,
            interaction_spin_orbit,
+            interaction_coulomb,
            args_central,
            args_spin_orbit,
+            args_coulomb,
        )
        return integral_elastic_xs(self.kinematics.k, splus, sminus, self.ls)

    def transmission_coefficients(
        self,
        interaction_central,
        interaction_spin_orbit,
+        interaction_coulomb=None,
        args_central=None,
        args_spin_orbit=None,
-        angles=None,
+        args_coulomb=None,


IntegralWorkspace.xs and IntegralWorkspace.transmission_coefficients similarly insert interaction_coulomb before the existing args_* parameters, which is a positional API break for any existing callers. Making Coulomb inputs keyword-only or appending them after the existing args would avoid silently swapped arguments.

Copilot · 2026-03-04T06:19:59Z

src/jitr/xs/elastic.py

+        interaction_coulomb=None,
        args_central=None,
        args_spin_orbit=None,


DifferentialWorkspace.xs also changes the positional parameter order by adding interaction_coulomb before the existing args_* parameters. This can break existing positional callers in subtle ways; consider keyword-only Coulomb params or reordering to keep previous positional arguments stable.

Suggested change

interaction_coulomb=None,

args_central=None,

args_spin_orbit=None,

args_central=None,

args_spin_orbit=None,

interaction_coulomb=None,

Copilot · 2026-03-04T06:19:59Z

src/jitr/optical_potentials/omp.py

+    For use with jitr.xs.elastic.DifferentialWorkspace, which requires
+    as input functions U(r;params_{central}) and U_{so}(r;params_{so})
+    for the central and spin-orbit interactions, respectively. These are
+    provided as callables that take a radius and a tuple of parameters
+    and return a complex potential.
+
+    In general, the parameters for the central and spin-orbit
+    interactions may depend on the reaction and kinematics, so the class
+    includes a method central_and_spin_orbit_params that can be
+    implemented by subclasses to calculate the appropriate parameters
+    for U and U_{so}.
+
+    This can be as complicated as a full global optical model in which a
+    full set of model parameters determines (the `params` arg in
+    `centralcentral_and_spin_orbit_params`) determines the A,Z,E,...
+    dependence of params_{central} and params_{so}, or it can be
+    something more simple, like a fixed set of parameters applicable to
+    a specific reaction or set of reactions.
+    """
+
+    def __init__(
+        self,
+        params: list[str],
+        interaction_central: Callable[[float, tuple], complex],
+        interaction_spin_orbit: Callable[[float, tuple], complex],
+        interaction_coulomb: Callable[[float, tuple], complex] = None,
+    ):


The SingleChannelOpticalModel docstring states the interactions take (r, tuple_of_params), and the type hints use Callable[[float, tuple], complex], but the quadrature kernel calls potentials as f(r, *args) (varargs). Consider updating the docstring and loosening the type annotations (e.g., Callable[..., complex] or a Protocol) to match the actual calling convention and avoid misleading users/type-checkers.

Copilot · 2026-03-04T06:19:59Z

src/jitr/optical_potentials/wlh.py

        uso,
        rso * A ** (1.0 / 3.0),
        aso,
    )
-    return coulomb_params, central_params, spin_orbit_params
+    return central_params, spin_orbit_params, coulomb_params


calculate_params now returns (central_params, spin_orbit_params, coulomb_params), which changes the return ordering from the previous Coulomb-first convention used elsewhere in the module history. This is an API-breaking change for any downstream code unpacking Global.get_params(...) results; consider preserving the old order (or adding a compatibility wrapper / named return) to avoid silent misassignment bugs.

Copilot · 2026-03-04T06:20:00Z

src/jitr/optical_potentials/chuq.py

    coulomb_params = (Z * Zp, RC)

-    return coulomb_params, central_params, spin_orbit_params
+    return central_params, spin_orbit_params, coulomb_params


calculate_params now returns (central_params, spin_orbit_params, coulomb_params). This changes the ordering from the earlier Coulomb-first return convention and can break any external code unpacking Global.get_params(...) results. Consider preserving the old order or returning a named structure to make ordering explicit and backward-compatible.

Suggested change

return central_params, spin_orbit_params, coulomb_params

return coulomb_params, central_params, spin_orbit_params

Copilot · 2026-03-04T06:20:00Z

src/jitr/optical_potentials/omp.py

+        central_params = ()
+        spin_orbit_params = ()
+        coulomb_params = ()
+        return central_params, spin_orbit_params, coulomb_params


params_by_term raises NotImplementedError, but the function then defines and returns empty tuples afterward (unreachable code). This should be removed to avoid confusion, or the method should be implemented to return real defaults if that's intended.

Suggested change

central_params = ()

spin_orbit_params = ()

coulomb_params = ()

return central_params, spin_orbit_params, coulomb_params

Copilot · 2026-03-04T06:20:00Z

src/jitr/optical_potentials/kduq.py

    spin_orbit_params = (
        vso,
        rso * A ** (1.0 / 3.0),
        aso,
        wso,
        rwso * A ** (1.0 / 3.0),
        awso,
    )

-    return coulomb_params, central_params, spin_orbit_params
+    return central_params, spin_orbit_params, coulomb_params


calculate_params now returns (central_params, spin_orbit_params, coulomb_params). If any downstream code (including callers of Global.get_params) still expects the previous return ordering, it will silently unpack the wrong values. Consider keeping the original ordering or providing an explicit, backward-compatible API (e.g., namedtuple/dataclass or a separate function) to prevent misassignment bugs.

Copilot · 2026-03-04T06:20:00Z

examples/notebooks/mass_exploration.ipynb

+    "        tcoeff_kduq[i, j, 0, :] = tplus\n",
+    "        tcoeff_kduq[i, j, 1, :] = tminus"


In the BMA loop the results are written into tcoeff_kduq[...] instead of tcoeff_kduq_bma[...]. This overwrites the non-BMA results and leaves tcoeff_kduq_bma unused/empty for downstream analysis.

Suggested change

" tcoeff_kduq[i, j, 0, :] = tplus\n",

" tcoeff_kduq[i, j, 1, :] = tminus"

" tcoeff_kduq_bma[i, j, 0, :] = tplus\n",

" tcoeff_kduq_bma[i, j, 1, :] = tminus"

Copilot · 2026-03-04T06:20:01Z

src/jitr/xs/elastic.py

+        interaction_coulomb=None,
        args_central=None,
        args_spin_orbit=None,


IntegralWorkspace.smatrix adds interaction_coulomb as a positional parameter between the existing interaction functions and the existing args_* parameters. Any existing callers passing args_central/args_spin_orbit positionally will now be misinterpreted as Coulomb inputs. Consider making the new Coulomb parameters keyword-only (e.g., using *) or moving them after the existing args_* parameters to preserve backward-compatible positional calling.

Suggested change

interaction_coulomb=None,

args_central=None,

args_spin_orbit=None,

args_central=None,

args_spin_orbit=None,

interaction_coulomb=None,

update global omp interface

bb8ac60

beykyle requested a review from Copilot March 4, 2026 06:12

Copilot started reviewing on behalf of beykyle March 4, 2026 06:12 View session

Copilot AI reviewed Mar 4, 2026

View reviewed changes

beykyle added 8 commits March 18, 2026 16:09

add volume integral calculation example

00dc0a3

Merge branch 'feature-volume-integral' into feature-omp-interface

c20e72e

add volume integrals

a51cf21

formatting

b2eed9d

return last pw

ba7e287

add more frame conversion tests

b3afdb7

fix bug in frame conversion

7965aaf

update examples

da6aee0

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

update global omp interface#76

update global omp interface#76
beykyle wants to merge 9 commits intomainfrom
feature-omp-interface

beykyle commented Mar 4, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Mar 4, 2026

Uh oh!

Copilot AI Mar 4, 2026

Uh oh!

Copilot AI Mar 4, 2026

Uh oh!

Copilot AI Mar 4, 2026

Uh oh!

Copilot AI Mar 4, 2026

Uh oh!

Copilot AI Mar 4, 2026

Uh oh!

Copilot AI Mar 4, 2026

Uh oh!

Copilot AI Mar 4, 2026

Uh oh!

Copilot AI Mar 4, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

	return central_params, spin_orbit_params, coulomb_params
	return coulomb_params, central_params, spin_orbit_params

		" tcoeff_kduq[i, j, 0, :] = tplus\n",
		" tcoeff_kduq[i, j, 1, :] = tminus"

Conversation

beykyle commented Mar 4, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Mar 4, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 4, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 4, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 4, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 4, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 4, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 4, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 4, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 4, 2026

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants