Phase 1 Polynomial-Fourier Benchmark

Run the Wave 5.2 Phase 1 Polynomial-Fourier analytical benchmark.

class scripts.analysis.polynomial_fourier_benchmark.run_phase1_polynomial_fourier_benchmark.CurveRecord(condition_id, split, direction, source_path, nominal_speed_rpm, nominal_torque_nm, nominal_temperature_deg_c, measured_speed_rpm, measured_torque_nm, measured_temperature_deg_c, theta_deg, te_deg)[source]

Bases: object

One leakage-safe directional benchmark curve.

Parameters:

  • condition_id (str)

  • split (str)

  • direction (str)

  • source_path (str)

  • nominal_speed_rpm (float)

  • nominal_torque_nm (float)

  • nominal_temperature_deg_c (float)

  • measured_speed_rpm (float)

  • measured_torque_nm (float)

  • measured_temperature_deg_c (float)

  • theta_deg (ndarray)

  • te_deg (ndarray)

condition_id: str
split: str
direction: str
source_path: str
nominal_speed_rpm: float
nominal_torque_nm: float
nominal_temperature_deg_c: float
measured_speed_rpm: float
measured_torque_nm: float
measured_temperature_deg_c: float
theta_deg: ndarray
te_deg: ndarray
operating_features()[source]

Return signed torque, absolute speed, and temperature.

Return type:

ndarray

onnx_features()[source]

Return the exact recovered MATLAB input order.

Return type:

ndarray

scripts.analysis.polynomial_fourier_benchmark.run_phase1_polynomial_fourier_benchmark.parse_arguments()[source]

Parse command-line arguments.

Return type:

Namespace

scripts.analysis.polynomial_fourier_benchmark.run_phase1_polynomial_fourier_benchmark.load_yaml(path)[source]

Load one YAML mapping.

Parameters:

path (Path)

Return type:

dict[str, Any]

scripts.analysis.polynomial_fourier_benchmark.run_phase1_polynomial_fourier_benchmark.load_csv_rows(path)[source]

Load one CSV as dictionaries.

Parameters:

path (Path)

Return type:

list[dict[str, str]]

scripts.analysis.polynomial_fourier_benchmark.run_phase1_polynomial_fourier_benchmark.resolve_repository_path(path_text)[source]

Resolve a repository-relative path.

Parameters:

path_text (str)

Return type:

Path

scripts.analysis.polynomial_fourier_benchmark.run_phase1_polynomial_fourier_benchmark.load_curve_records(configuration, manifest)[source]

Load and uniformly resample every eligible directional curve.

Parameters:

  • configuration (dict[str, Any])

  • manifest (dict[str, Any])

Return type:

list[CurveRecord]

scripts.analysis.polynomial_fourier_benchmark.run_phase1_polynomial_fourier_benchmark.fit_surface_map(curve_record_list, harmonic_order_list)[source]

Fit one separate quadratic coefficient surface per direction.

Parameters:

  • curve_record_list (list[CurveRecord])

  • harmonic_order_list (list[int])

Return type:

dict[str, QuadraticCoefficientSurface]

scripts.analysis.polynomial_fourier_benchmark.run_phase1_polynomial_fourier_benchmark.curve_metrics(measured_curve, predicted_curve)[source]

Compute curve-first raw, shape, offset, derivative, and closure metrics.

Parameters:

  • measured_curve (ndarray)

  • predicted_curve (ndarray)

Return type:

dict[str, float]

scripts.analysis.polynomial_fourier_benchmark.run_phase1_polynomial_fourier_benchmark.harmonic_error_metrics(measured_curve, predicted_curve, harmonic_order_list)[source]

Summarize retained-order amplitude and circular phase errors.

Parameters:

  • measured_curve (ndarray)

  • predicted_curve (ndarray)

  • harmonic_order_list (list[int])

Return type:

dict[str, float]

scripts.analysis.polynomial_fourier_benchmark.run_phase1_polynomial_fourier_benchmark.evaluate_benchmark(configuration, curve_record_list, surface_map_by_model, onnx_predictor, plc_parameters)[source]

Evaluate all analytical formulations on all common curves.

Parameters:

  • configuration (dict[str, Any])

  • curve_record_list (list[CurveRecord])

  • surface_map_by_model (dict[str, dict[str, QuadraticCoefficientSurface]])

  • onnx_predictor (RecoveredMatlabOnnxPredictor)

  • plc_parameters (Any)

Return type:

tuple[list[dict[str, Any]], dict[tuple[str, str, str], ndarray]]

scripts.analysis.polynomial_fourier_benchmark.run_phase1_polynomial_fourier_benchmark.aggregate_metrics(metric_row_list)[source]

Aggregate every numeric metric by model, split, and direction.

Parameters:

metric_row_list (list[dict[str, Any]])

Return type:

list[dict[str, Any]]

scripts.analysis.polynomial_fourier_benchmark.run_phase1_polynomial_fourier_benchmark.select_phase1_models(aggregate_row_list)[source]

Select an analytical reference and a structurally different comparator.

Parameters:

aggregate_row_list (list[dict[str, Any]])

Return type:

dict[str, Any]

scripts.analysis.polynomial_fourier_benchmark.run_phase1_polynomial_fourier_benchmark.false_value()[source]

Return a YAML-safe explicit false value.

Return type:

bool

scripts.analysis.polynomial_fourier_benchmark.run_phase1_polynomial_fourier_benchmark.run_deterministic_tests(onnx_predictor, plc_parameters)[source]

Run deterministic analytical identity and contract tests.

Parameters:

  • onnx_predictor (RecoveredMatlabOnnxPredictor)

  • plc_parameters (Any)

Return type:

dict[str, Any]

scripts.analysis.polynomial_fourier_benchmark.run_phase1_polynomial_fourier_benchmark.evaluate_matlab_examples(configuration, onnx_predictor)[source]

Evaluate recovered ONNX inference against the five MATLAB Fw examples.

Parameters:

  • configuration (dict[str, Any])

  • onnx_predictor (RecoveredMatlabOnnxPredictor)

Return type:

list[dict[str, Any]]

scripts.analysis.polynomial_fourier_benchmark.run_phase1_polynomial_fourier_benchmark.build_preprocessing_rows(configuration, curve_record_list)[source]

Audit the Bauer preprocessing chain on the held-out test curves.

Parameters:

  • configuration (dict[str, Any])

  • curve_record_list (list[CurveRecord])

Return type:

list[dict[str, Any]]

scripts.analysis.polynomial_fourier_benchmark.run_phase1_polynomial_fourier_benchmark.write_csv_rows(path, row_list)[source]

Write stable dictionary rows to CSV.

Parameters:

  • path (Path)

  • row_list (list[dict[str, Any]])

Return type:

None

scripts.analysis.polynomial_fourier_benchmark.run_phase1_polynomial_fourier_benchmark.surface_to_payload(surface)[source]

Serialize one fitted surface with explicit coefficients.

Parameters:

surface (QuadraticCoefficientSurface)

Return type:

dict[str, Any]

scripts.analysis.polynomial_fourier_benchmark.run_phase1_polynomial_fourier_benchmark.create_representative_plots(configuration, curve_record_list, metric_row_list, prediction_cache, selected_model_map)[source]

Create median and worst held-out plots for both directions.

Parameters:

  • configuration (dict[str, Any])

  • curve_record_list (list[CurveRecord])

  • metric_row_list (list[dict[str, Any]])

  • prediction_cache (dict[tuple[str, str, str], ndarray])

  • selected_model_map (dict[str, Any])

Return type:

list[str]

scripts.analysis.polynomial_fourier_benchmark.run_phase1_polynomial_fourier_benchmark.write_report(configuration, summary_payload, aggregate_row_list)[source]

Write the canonical Phase 1 analytical report.

Parameters:

  • configuration (dict[str, Any])

  • summary_payload (dict[str, Any])

  • aggregate_row_list (list[dict[str, Any]])

Return type:

None

scripts.analysis.polynomial_fourier_benchmark.run_phase1_polynomial_fourier_benchmark.main()[source]

Run Phase 1 and write its complete evidence package.

Return type:

None