Skip to content

Cross-Validation and Accuracy

Use this page to choose a starting configuration and the validation checks for a reported RF result. Start with the Recommended Configuration, check the exact Support Boundaries, and use the Benchmarks for per-case evidence.

Start with the uniform Cartesian Yee RF/FDTD configuration for:

  • cavity and waveguide-style RF problems;
  • patch-style resonance workflows;
  • probes and Harminv resonance extraction;
  • selected port and S-parameter calculations; and
  • benchmarked far-field calculations.

The Recommended Configuration page lists the applicable boundaries, sources, ports, observables, and required checks. A successful run does not replace a mesh-convergence study or comparison with an analytic or independent RF reference.

Calculations with narrower validated limits

Section titled “Calculations with narrower validated limits”

The following calculations require their matching API and documented geometry, mesh, frequency, normalization, and placement limits:

  • rectangular waveguide S-matrices via compute_waveguide_s_matrix(...);
  • specialized microstrip-line S-matrices via compute_msl_s_matrix(...);
  • lumped/wire S-parameters via the matching add_port(...) calculator;
  • one-port coaxial transmission-line reflection via compute_coaxial_line_reflection(...); and
  • differentiable design loops that use proxy objectives and then validate the final RF observable with the applicable convergence and reference checks.

Examples and regression tests do not expand those validated limits. See Support Boundaries for the current restrictions and the S-parameter guide for calculator-specific checks.

ExamplePurpose
validation/crossval/05_patch_antenna.pypatch-antenna cross-check
validation/crossval/11_waveguide_port_wr90.pyrectangular waveguide-port reference workflow
examples/inverse_design/multilayer_ar_coating.pyinverse-design example; validate the optimized result independently

Record the mesh, boundary, source or port family, calculator, frequency range, normalization, and comparison criterion used for the result. If a configuration falls outside the listed limits, do not describe its RF accuracy as validated.