Issues

• Explainers: Adverse Impact at Each Hiring Stage, Compounding Disparities, and Differential Validity

  documentationhelp wanted
  yakew7/Fair-Code#1119 · Filter by author yakew7 opened on Oct 21, 2026

• Add a lint check that every explainer has the required sections

  enhancementgood first issue
  yakew7/Fair-Code#1118 · Filter by author yakew7 opened on Oct 21, 2026

• Explainers: Feature Importance by Group, Partial Dependence by Group, and Explanation Disagreement

  documentationhelp wanted
  yakew7/Fair-Code#1117 · Filter by author yakew7 opened on Oct 21, 2026

• Explainers: Seed and Split Sensitivity, Cross-Validation for Fairness Metrics, and Building a Test Set That Can Detect a Gap

  documentationhelp wanted
  yakew7/Fair-Code#1116 · Filter by author yakew7 opened on Oct 21, 2026

• Explainers: Race as a Variable in Clinical Models, Utilization as a Label, and Discharge Leakage in Readmission Models

  documentationhelp wanted
  yakew7/Fair-Code#1115 · Filter by author yakew7 opened on Oct 21, 2026

• Explainers: Adverse Action Notices, Geographic Proxies and Redlining, and Risk-Based Pricing Gaps

  documentationhelp wanted
  yakew7/Fair-Code#1114 · Filter by author yakew7 opened on Oct 21, 2026

• profile() and compare() raise AttributeError and KeyError when given the wrong kind of object

  enhancementgood first issue
  yakew7/Fair-Code#1113 · Filter by author yakew7 opened on Oct 21, 2026

• load_audit("Nope") fails with a bare FileNotFoundError that does not say an audit folder was expected

  enhancementgood first issue
  yakew7/Fair-Code#1112 · Filter by author yakew7 opened on Oct 21, 2026

• scripts/freeze_paper_results.py freeze() is untested (39% coverage)

  enhancementhelp wanted
  yakew7/Fair-Code#1111 · Filter by author yakew7 opened on Oct 21, 2026

• scripts/build_explainers.py is only 38% covered and has no test for the page-head output

  enhancementhelp wanted
  yakew7/Fair-Code#1110 · Filter by author yakew7 opened on Oct 21, 2026

• scripts/check_broken_links.py has no tests (0% coverage)

  enhancementgood first issue
  yakew7/Fair-Code#1109 · Filter by author yakew7 opened on Oct 21, 2026

• Four explainers have no Limitations section: proxy-variables, sampling-bias, neural-networks, automation-bias

  documentationgood first issue
  yakew7/Fair-Code#1108 · Filter by author yakew7 opened on Oct 21, 2026

• 14 explainers have no link from any other explainer, and 28 have one or none

  documentationhelp wanted
  yakew7/Fair-Code#1107 · Filter by author yakew7 opened on Oct 21, 2026

• Profile flags say "only 1 rows" instead of "only 1 row"

  buggood first issue
  yakew7/Fair-Code#1106 · Filter by author yakew7 opened on Oct 21, 2026
  2

  

• 11 explainer page titles are longer than 65 characters (up to 99) and get cut in search results

  enhancementgood first issue
  yakew7/Fair-Code#1105 · Filter by author yakew7 opened on Oct 21, 2026

• 65 of 82 explainer pages have a meta description longer than 160 characters (up to 634)

  enhancementhelp wanted
  yakew7/Fair-Code#1104 · Filter by author yakew7 opened on Oct 21, 2026

• Explainers: Randomized Post-Processing, Fairness-Accuracy Pareto Frontiers and Choosing a Mitigation Strategy

  documentationhelp wanted
  yakew7/Fair-Code#1102 · Filter by author yakew7 opened on Oct 21, 2026

• Explainers: Causal Graphs for Fairness and Path-Specific Effects

  documentationhelp wanted
  yakew7/Fair-Code#1101 · Filter by author yakew7 opened on Oct 21, 2026

• Explainers: Monitoring Fairness in Production, Label Shift and Fairness, and Delayed Feedback

  documentationhelp wanted
  yakew7/Fair-Code#1100 · Filter by author yakew7 opened on Oct 21, 2026

• Explainers: Minimum Detectable Gap, Small-Group Shrinkage and Confidence Intervals for Ratios

  documentationhelp wanted
  yakew7/Fair-Code#1099 · Filter by author yakew7 opened on Oct 21, 2026

• Run the Python code blocks in the explainers as a smoke test

  enhancementhelp wanted
  yakew7/Fair-Code#1098 · Filter by author yakew7 opened on Oct 21, 2026
  2

  

• disparate_impact_ratio explodes when the advantaged selection rate is near zero (COMPAS gradient boosting: 1,250, interval up to 879,158)

  bughelp wanted
  yakew7/Fair-Code#1094 · Filter by author yakew7 opened on Oct 21, 2026

• Refresh results/ with a full benchmark rerun on the current requirements-lock.txt

  enhancementhelp wanted
  yakew7/Fair-Code#1068 · Filter by author yakew7 opened on Oct 18, 2026
  2

  

• Numeric column headers keep their type in Python (the integer 2020) and are text in the web engine

  bughelp wanted
  yakew7/Fair-Code#1045 · Filter by author yakew7 opened on Oct 17, 2026
  2

• JSON string cells "NA" and "" are missing in the web engine and ordinary values in Python

  bughelp wanted
  yakew7/Fair-Code#1036 · Filter by author yakew7 opened on Oct 17, 2026
  2