HarmonicConvergenceResult#

HarmonicConvergenceResult is the reusable output of HarmonicTransform.converge_parameters(). It keeps calibration separate from production evaluation and preserves the selected parameters together with sampling and quadrature diagnostics.

class quartic2d.HarmonicConvergenceResult(sampling, quadrature, method, interpolator, rtol, atol, q_tail_rtol)#

Calibration result returned by HarmonicTransform.converge_parameters().

The result separates an explicit convergence study from later production transforms. It stores the selected numerical parameters together with the q-support, q-grid, and radial-quadrature diagnostics used to select them.

The requested rtol is a self-convergence criterion, not an independent reference-error guarantee. On each refinement comparison, acceptance requires both

\[\frac{\|f^{(n)}-f^{(n-1)}\|_2} {\|f^{(n-1)}\|_2}\leq \mathrm{rtol}\]

and

\[\|f^{(n)}-f^{(n-1)}\|_\infty \leq \mathrm{atol} +\mathrm{rtol}\,\|f^{(n-1)}\|_\infty.\]

The relative \(L^\infty\) change is recorded as a diagnostic but is not an additional acceptance condition. q_tail_rtol is separate and limits the estimated relative \(L^2\) norm omitted beyond the selected q_max.

Parameters:

  • sampling (object) – Structured q-support and q-grid convergence record.

  • quadrature (object) – Structured radial-quadrature convergence record.

  • method (str) – Selected radial Hankel quadrature method.

  • interpolator (str) – Radial interpolator used during calibration.

  • rtol (float) – Relative and absolute self-convergence thresholds.

  • atol (float) – Relative and absolute self-convergence thresholds.

  • q_tail_rtol (float) – Relative L2 tail budget used to select momentum support.

converged#

True only when q support, q-grid interpolation, and radial quadrature satisfy their configured criteria.

Type:

bool

parameters#

Constructor keyword arguments selected for a production HarmonicTransform.

Type:

dict

Notes

Use transform() to build a production transform without rerunning the convergence study. Use to_dict() when the calibration record should be stored with numerical results or provenance metadata.

Common operations#

plot_convergence([title])

Plot the convergence record used to select production parameters.

transform(decomposition, *[, check])

Build a production transform from the selected parameters.

to_dict()

Return a JSON-serializable calibration record.

The parameters property can be inspected directly, while to_dict() is intended for provenance records and machine-readable output.