Paper Reference Archive Parity Report
This page documents the evaluation-only report builder used to compare the
saved repository paper-reference archives under models/paper_reference.
Build a parity report across the saved RCIM paper-reference archives.
- scripts.paper_reimplementation.rcim_ml_compensation.reference_family_vs_feedforward.run_paper_reference_archive_parity_report.load_yaml_dictionary(yaml_path)[source]
Load one YAML dictionary from disk.
- Parameters:
yaml_path (Path)
- Return type:
dict[str, Any]
- scripts.paper_reimplementation.rcim_ml_compensation.reference_family_vs_feedforward.run_paper_reference_archive_parity_report.find_latest_track2_validation_summary()[source]
Find the newest TE Curve Verification Pipeline validation summary with paper-reference metrics.
- Return type:
Path
- scripts.paper_reimplementation.rcim_ml_compensation.reference_family_vs_feedforward.run_paper_reference_archive_parity_report.resolve_candidate_key(source_label, family_label, surface_label)[source]
Resolve the TE Curve Verification Pipeline candidate id for one paper-reference archive model.
- Parameters:
source_label (str)
family_label (str)
surface_label (str)
- Return type:
str
- scripts.paper_reimplementation.rcim_ml_compensation.reference_family_vs_feedforward.run_paper_reference_archive_parity_report.resolve_composite_candidate_key(source_label, surface_label)[source]
Resolve the composed best-candidate id for one source and surface.
- Parameters:
source_label (str)
surface_label (str)
- Return type:
str | None
- scripts.paper_reimplementation.rcim_ml_compensation.reference_family_vs_feedforward.run_paper_reference_archive_parity_report.build_candidate_lookup(track2_summary)[source]
Build a lookup from candidate id to candidate metadata.
- Parameters:
track2_summary (dict[str, Any])
- Return type:
dict[str, dict[str, Any]]
- scripts.paper_reimplementation.rcim_ml_compensation.reference_family_vs_feedforward.run_paper_reference_archive_parity_report.resolve_direction_metric(track2_summary, candidate_id, direction_label)[source]
Resolve the direction-filtered TE Curve Verification Pipeline curve metric for one candidate.
- Parameters:
track2_summary (dict[str, Any])
candidate_id (str)
direction_label (str)
- Return type:
dict[str, float] | None
- scripts.paper_reimplementation.rcim_ml_compensation.reference_family_vs_feedforward.run_paper_reference_archive_parity_report.resolve_target_metric(track2_summary, candidate_id)[source]
Resolve the target-level metric summary for one reference-bank candidate.
- Parameters:
track2_summary (dict[str, Any])
candidate_id (str)
- Return type:
dict[str, float] | None
- scripts.paper_reimplementation.rcim_ml_compensation.reference_family_vs_feedforward.run_paper_reference_archive_parity_report.build_curve_metric_row_list(track2_summary)[source]
Build family-level TE Curve Verification Pipeline curve metric rows for paper-reference archives.
- Parameters:
track2_summary (dict[str, Any])
- Return type:
list[dict[str, Any]]
- scripts.paper_reimplementation.rcim_ml_compensation.reference_family_vs_feedforward.run_paper_reference_archive_parity_report.build_target_metric_row_list(track2_summary)[source]
Build harmonic-target metric rows for paper-reference archives.
- Parameters:
track2_summary (dict[str, Any])
- Return type:
list[dict[str, Any]]
- scripts.paper_reimplementation.rcim_ml_compensation.reference_family_vs_feedforward.run_paper_reference_archive_parity_report.write_csv(row_list, csv_path)[source]
Write a homogeneous row list to CSV.
- Parameters:
row_list (list[dict[str, Any]])
csv_path (Path)
- Return type:
None
- scripts.paper_reimplementation.rcim_ml_compensation.reference_family_vs_feedforward.run_paper_reference_archive_parity_report.build_row_lookup(row_list)[source]
Build a source/family/surface lookup for metric rows.
- Parameters:
row_list (list[dict[str, Any]])
- Return type:
dict[tuple[str, str, str], dict[str, Any]]
- scripts.paper_reimplementation.rcim_ml_compensation.reference_family_vs_feedforward.run_paper_reference_archive_parity_report.classify_pairwise_similarity(delta_mean_percentage_error_pct)[source]
Classify archive similarity from the TE Curve Verification Pipeline MPE delta.
- Parameters:
delta_mean_percentage_error_pct (float)
- Return type:
str
- scripts.paper_reimplementation.rcim_ml_compensation.reference_family_vs_feedforward.run_paper_reference_archive_parity_report.build_pairwise_comparison_row_list(curve_metric_row_list, target_metric_row_list)[source]
Build same-family pairwise comparisons across archive groups.
- Parameters:
curve_metric_row_list (list[dict[str, Any]])
target_metric_row_list (list[dict[str, Any]])
- Return type:
list[dict[str, Any]]
- scripts.paper_reimplementation.rcim_ml_compensation.reference_family_vs_feedforward.run_paper_reference_archive_parity_report.select_rows(row_list, direction_label, source_label)[source]
Select rows for one direction and source.
- Parameters:
row_list (list[dict[str, Any]])
direction_label (str)
source_label (str)
- Return type:
list[dict[str, Any]]
- scripts.paper_reimplementation.rcim_ml_compensation.reference_family_vs_feedforward.run_paper_reference_archive_parity_report.find_best_row(row_list, direction_label, source_label, include_composite=False)[source]
Find the best curve row for one source and direction.
- Parameters:
row_list (list[dict[str, Any]])
direction_label (str)
source_label (str)
include_composite (bool)
- Return type:
dict[str, Any]
- scripts.paper_reimplementation.rcim_ml_compensation.reference_family_vs_feedforward.run_paper_reference_archive_parity_report.format_float(value, decimal_count=3)[source]
Format one floating-point value for Markdown tables.
- Parameters:
value (float)
decimal_count (int)
- Return type:
str
- scripts.paper_reimplementation.rcim_ml_compensation.reference_family_vs_feedforward.run_paper_reference_archive_parity_report.build_curve_table(row_list)[source]
Build one Markdown table for TE Curve Verification Pipeline curve rows.
- Parameters:
row_list (list[dict[str, Any]])
- Return type:
list[str]
- scripts.paper_reimplementation.rcim_ml_compensation.reference_family_vs_feedforward.run_paper_reference_archive_parity_report.build_target_table(row_list)[source]
Build one Markdown table for target-level rows.
- Parameters:
row_list (list[dict[str, Any]])
- Return type:
list[str]
- scripts.paper_reimplementation.rcim_ml_compensation.reference_family_vs_feedforward.run_paper_reference_archive_parity_report.select_pairwise_rows(row_list, comparison_id)[source]
Select pairwise rows for one comparison id.
- Parameters:
row_list (list[dict[str, Any]])
comparison_id (str)
- Return type:
list[dict[str, Any]]
- scripts.paper_reimplementation.rcim_ml_compensation.reference_family_vs_feedforward.run_paper_reference_archive_parity_report.build_pairwise_table(row_list)[source]
Build one Markdown table for same-family archive pairwise rows.
- Parameters:
row_list (list[dict[str, Any]])
- Return type:
list[str]
- scripts.paper_reimplementation.rcim_ml_compensation.reference_family_vs_feedforward.run_paper_reference_archive_parity_report.build_pairwise_verdict_summary(row_list)[source]
Build compact count summary for pairwise verdict classes.
- Parameters:
row_list (list[dict[str, Any]])
- Return type:
list[str]
- scripts.paper_reimplementation.rcim_ml_compensation.reference_family_vs_feedforward.run_paper_reference_archive_parity_report.build_canonical_report_markdown(summary_dictionary)[source]
Build the canonical paper-reference archive parity Markdown report.
- Parameters:
summary_dictionary (dict[str, Any])
- Return type:
str
- scripts.paper_reimplementation.rcim_ml_compensation.reference_family_vs_feedforward.run_paper_reference_archive_parity_report.build_archive_parity_summary(track2_summary_path, output_suffix)[source]
Build and persist the archive parity summary and report.
- Parameters:
track2_summary_path (Path)
output_suffix (str)
- Return type:
dict[str, Any]