Mark the axes and spin arrangements before fitting energies
For a real material, open the structure with the tested magnetization directions or spin pairs labelled, and inspect final site moments or spin density for every run. Plot directional energies or configuration energies with error scale and fit residuals, keeping the sign, cell, pair counting, and spin normalization visible. Compare with symmetry and published conventions through visual tools, specialist magnetic tools, and literature sources.
Start with real magnetic states: inspect the final moment vectors or spin density and retained high-precision energies for every direction or configuration. This repository supplies no real SOC or configuration series here. The optional ledger below is invented sign-convention arithmetic, not a magnetic result.
Use this fixture only to check ledger arithmetic after defining the real parent states. It separates two operations: ranking invented directional-energy rows for one fixed texture, and fitting invented two-site energies to the convention
Under this written convention, positive favours parallel moments. The built-in values are not outputs from an electronic-structure engine.
For a real calculation, the parent object must contain compatible total energies, final moment vectors, structure, SOC/noncollinear mode, crystallographic directions, Hamiltonian, numerical settings, and a declared normalization. This script supplies none of those material-specific inputs; it creates its invented ledger internally.
Purpose
If you want to check the written sign convention, run from the repository root:
python3 examples/practical-guides/anisotropy_exchange_ledger.pyThe command prints a Python result dictionary. Confirm normal exit and the final line Invented anisotropy/exchange fixture passed; it verifies arithmetic only. This establishes program completion for the fixture, not SOC or exchange convergence.
Inspect the directional-energy reduction
Read easy_direction_within_fixture and mae_relative_to_c. The first names the lowest built-in directional entry; the second subtracts the built-in c axis reference. Check that the values use the declared energy_unit normalization.
This ranking is conditional on the invented fixed-geometry entries. In real work, all directions must use the same magnetic texture, structure, SOC Hamiltonian, charge, potential data, k points, occupations, symmetry policy, and numerical precision. A scalar-relativistic energy cannot be mixed with an SOC energy, and a collinear energy cannot silently enter a noncollinear directional comparison.
Inspect the exchange reduction
Read heisenberg_convention before fitted_J. The fixture uses parallel energy and antiparallel energy , so their difference gives the fitted positive under the written two-site convention.
Then compare held_out_orthogonal_energy with predicted_orthogonal_energy. Equality checks one held-out reconstruction for this exactly determined invented model. It does not show that a real material follows a two-site Heisenberg Hamiltonian, that the fitted interaction range is unique, or that moment magnitudes remain fixed.
Continue with real energies only after compatibility checks
Quantum ESPRESSO pw.x documents noncollinear and spin—orbit inputs. VASP’s magnetic-anisotropy note and LSORBIT page describe SOC directional calculations and their numerical sensitivity. Follow the documentation for the selected code and version; this fixture does not provide or verify those input files.
For real MAE, reconstruct every directional difference from retained high-precision total energies and the declared normalization, then converge the difference and easy-direction ranking. For real exchange, retain the final moment map of every configuration, state the sign and pair-counting convention, fit enough independent configurations, and test held-out states. A passed script run does not replace any of those checks.
Execution verifies invented arithmetic, sign convention, normalization labels, and one held-out reconstruction only. It does not calculate a material MAE or exchange parameter, establish an easy axis, validate a magnetic Hamiltonian, predict a transition temperature, or support a scientific conclusion.