chore(deps): bump softprops/action-gh-release from 1 to 2#3
Open
dependabot[bot] wants to merge 71 commits into
Open
chore(deps): bump softprops/action-gh-release from 1 to 2#3dependabot[bot] wants to merge 71 commits into
dependabot[bot] wants to merge 71 commits into
Conversation
Added new files for adapter architecture: • metrics/base.py – defines MetricResult schema and MetricAdapter protocol • metrics/native_adapter.py – built-in fallback for fairness metrics Expanded fairlearn_adapter.py and aequitas_adapter.py with guarded imports Updated core.py to register adapters, enable auto-selection, and add caching Added unit tests for adapter selection and result schema Ensured backward compatibility with Phase 0 Result class These deliver: → FairnessAnalyzer wrapper with unified interface → Common result schema → Automatic backend selection and lightweight caching → Test coverage target ~80%
- Added metric implementations: • demographic_parity_difference (classification) • equalized_odds_difference (classification) • mae_parity_difference (regression) - Implemented intersectional metric support with min_group_size filter - Added intersectional utilities (build_intersectional_labels, min_group_mask, group_sizes) - Introduced metric_registry.json for easy extensibility - Fixed categorical comparison bug in min_group_mask (dtype coercion for unordered categoricals) - Added unit tests for small-group exclusion, missing values, and intersectional logic - Added benchmarking script (benchmark_metrics_100k.py) for runtime profiling (~100k rows) - Verified statistical and runtime correctness for classification and regression metrics These deliver: → Fully functional metric engine with classification, regression, and intersectional support → Safe handling of categorical data types → Benchmarks and tests confirming performance and reliability
Summary: -Implemented bootstrap confidence intervals (percentile & BCa) with improved two-sample resampling logic for correct coverage. -Added Bayesian credible interval (Beta–Binomial) for small-sample fairness evaluation. -Implemented multiple comparison corrections (Bonferroni & Benjamini–Hochberg FDR). -Added effect size computations (risk ratio, Cohen’s d placeholder). -Enhanced unit tests for all statistical components using synthetic datasets. Results: -All tests passed successfully. -Empirical coverage for the bootstrap percentile method verified within expected 0.85–0.95 range. -Current test coverage: ~53% total, with 90%+ coverage for native_adapter, fairlearn_adapter, and core metrics. -Phase 3 functionality validated—ready to proceed to Phase 4 integration.
…d statistical validation
- Added MLflow logging utility (log_fairness_metrics) - Integrated pytest plugin with assert_fairness() - Extended CLI with command for fairness checks - Implemented Markdown reporting function (to_markdown_report) - Enhanced CI with multi-OS testing and coverage - Configured pre-commit hooks for linting and sample-check placeholder All core workflows now operational and passing tests.
…prepare v0.1.0 release - Added full end-to-end (E2E) system tests validating metric computation, CI integration, and Markdown reporting - Introduced deterministic test fixture () for reproducibility - Created demo notebook () showcasing metric computation, CI confidence intervals, and MLflow tracking - Updated README with usage tutorial and CLI validation examples - Added CHANGELOG.md and version bump to v0.1.0 - Verified workflow integration and test coverage for release candidate
…correctness, reproducibility, and usability
- Added visualization of fairness metrics (DP, EO) with bootstrap confidence intervals. - Updated demo notebook and documentation for reproducibility. - Verified end-to-end CLI, reporting, and MLflow integration. - Finalized readiness for v0.1.0 release. - Added bootstrap confidence intervals and effect size computation to CLI fairness validation. - Created visualization scripts for demographic parity and equalized odds with 95% CIs. - Updated demo.ipynb for end-to-end workflow demonstration (metrics → CI → MLflow tracking). - Verified reproducibility and usability of fairness measurement module. - Generated report_with_ci.md artifact as validation output.
…e record, updated README and changelog for release readiness; validated full workflow integration and reporting.
…record, updated README and changelog for release readiness; validated full workflow integration and reporting.
- Added new package under with initial submodules: - detectors/: representation, disparity, proxy, and report stubs - transformers/: reweighting, disparate impact remover, and utils stubs - config/: YAML loader (Phase 0 version) - orchestration/: builder and registry for dynamic pipeline construction - Expanded CLI to include command for running stub detectors via config and CSV input - Added example , fixture CSV, and smoke test for pipeline import wiring - Updated to include PyYAML dependency - Establishes foundation for Phase 1 (functional bias detection and config validation)
Implements the foundational architecture for the Fairness Pipeline Module, covering Phase 1 (scaffolding and configuration) and Phase 2 (detection and execution logic). Details: - Added pipeline package structure with submodules: • config – typed YAML loader (PipelineConfig, PipelineStep) • detectors – bias detection primitives • transformers – InstanceReweighting, DisparateImpactRemover • orchestration – pipeline builder, execution engine - Implemented run_detectors, build_pipeline, and apply_pipeline in orchestration/engine.py. - Added BiasReport container for consistent access (.meta) and JSON-safe serialization. - Integrated CLI compatibility with Mapping interface and to_dict() handling. - Added smoke test (tests/pipeline/test_smoke_pipeline.py) and config validation. - Fixed CLI artifact serialization and test assertions for Phase 2. Testing: - Verified pytest smoke test passes end-to-end. - Confirmed cmd_pipeline_run writes valid JSON artifacts.
Merge: Pipeline module
…n, and reweighing support ### Summary - Implemented Phase 3 (configuration + detector orchestration refinement) - Enhanced to return typed objects. - Added JSON serialization () for detector reports. - Expanded logic for safer transformer instantiation with injected defaults. - Implemented Phase 4 (CLI + end-to-end integration) - Integrated in CLI to execute full pipeline runs from config and CSV input. - Added artifact outputs (, , ) and improved error handling. - Updated to include , , and aliases. - Added fallback import handling for spelling variants (e.g., reweighting / reweighing). - Ensured smart default injection for and when omitted in YAML. ### Improvements - Unified bias detection output into a consistent schema. - Enabled reproducible runs via YAML-driven configuration. - Improved pipeline usability through flexible transformer naming and automatic default handling. - Simplified CLI experience with full artifact generation and path validation.
- Implemented ReductionsWrapper for scikit-learn with DemographicParity constraint - Added FairnessRegularizer and LagrangianFairnessTrainer for PyTorch - Introduced GroupFairnessCalibrator and ParetoFrontier visualization tools - Integrated module tests for sklearn_, torch_, postproc, and viz subpackages - Updated CLI to include training operations and maintained compatibility with Python 3.12.5
refactor(pipeline,cli,monitoring): unify builder, add config validation, refactor validate; packaging + docs; tests Pipeline/orchestration Consolidate builder into pipeline/orchestration/engine.py; remove builder.py Centralize transformer registry; keep single build_pipeline/apply_pipeline/run_detectors API Config Add minimal schema validation in pipeline/config/loader.py (profiles supported) Export PipelineConfig/PipelineStep/load_config via pipeline/config/__init__.py CLI Refactor validate into small helpers (CSV parsing, sensitive normalization, evaluation) Parameterize CI/effect sizes via FairnessAnalyzer (adds ci_samples) and flags Lazy-load heavy deps inside subcommands (fairlearn, torch, etc.) Add return type hints for public CLI functions Monitoring Introduce monitoring/config.py with MonitoringSettings, DriftConfig, ReportConfig Update drift engine and dashboard to use shared settings; persist config alongside artifacts Packaging Move runtime deps to [project.dependencies] in pyproject.toml Add extras: training, monitoring, adapters; add requirements.in/requirements-dev.in Simplify requirements.txt to pip-tools workflow Documentation & DX Rename CHANEGELOG.md → CHANGELOG.md Enrich README: minimal pipeline YAML + config schema; extras + pip-compile instructions Add .pre-commit-config.yaml (ruff/black/isort/nbstripout) Typing/Naming Normalize “sensitive” terminology; add explicit return types in CLI and orchestration Tests Add tests/pipeline/test_config_loader.py (schema/profiles) Extend pipeline smoke to assert sample weights from YAML Add CLI validation test for --with-ci and --with-effects Add monitoring tests for dashboard stubs and drift engine (with optional deps auto-skip)
Corrected percentile CI scaling so bootstrap intervals now use percent units expected by NumPy and covered the regression with a direct unit test of _percentile_ci behavior. Reworked the MLflow logging helper to emit string payloads (preferring mlflow.log_dict, falling back safely) and added a stubbed unit test to ensure structured artifacts log without runtime errors. Enabled confidence intervals and effect sizes by default across the analyzer APIs and verified the richer outputs with targeted tests covering classification and regression metrics.
- Restructure README with enhanced PyPI installation and quick start - Add comprehensive API reference (docs/api.md) - Add integration guide with usage patterns and best practices (docs/integration_guide.md) Completes Phase 4 documentation requirements for production deployment.
…y (Phase 5) Add comprehensive VERSIONING.md documenting semantic versioning strategy, backward compatibility guarantees, deprecation policy, and migration guides. Update README and API docs to reference the new versioning document. Implements Phase 5 of production deployment plan.
…e 6) Add comprehensive test coverage for: - Package installation and CLI entry points - Public API validation across all modules - Backward compatibility verification Includes 54 tests covering installation, API structure, and compatibility. Tests gracefully handle optional dependencies and skip when missing. Completes Phase 6 of production deployment implementation plan.
- Add test_detector_stubs.py with tests for DisparityDetector, ProxyDetector, and RepresentationDetector - Verify stub implementations return correct result types with empty dictionaries - Tests cover main functionality and edge cases (None values, multiple attributes) - Part of Phase 1 test coverage improvement plan
Implement Phase 2 reporting tests to improve coverage from 8% to 70%+.
Add 96 test cases covering all reporting functions and edge cases.
Test Coverage Added:
- to_markdown_report(): 14 tests for various metric types, CI values,
effect sizes, and edge cases (NaN, empty results, missing fields)
- Helper functions: 25 tests total
- _fmt_ci(): 7 tests for CI formatting with None/empty values
- _coerce(): 5 tests for dataclass/dict/primitive coercion
- _compute_group_rates(): 6 tests for group rate computation with
min_group_size filtering
- _assess_severity(): 7 tests for severity assessment across thresholds
- _interpret_metric_value(): 6 tests for metric interpretation
- _analyze_training_convergence(): 13 tests for various history patterns,
convergence detection, and edge cases (insufficient data, missing keys)
- _generate_recommendations(): 12 tests for all lifecycle stages
(data_stage, training_stage, evaluation_stage, deployment_stage)
- _prepare_report_data(): 11 tests for auto-computation of performance
metrics, convergence analysis, and weight statistics
- generate_training_fairness_report(): 13 tests for minimal/full data,
file output, JSON serialization, and status handling
All tests pass and follow existing test patterns in the codebase.
- Add 34 tests for AequitasAdapter (init, available, _mask_small_groups, demographic_parity_difference, equalized_odds_difference) - Add 37 tests for bootstrap functions (_percentile_ci, _z, _phi, bootstrap_ci, bca_ci) - Add 2 tests for CLI cmd_sample_check command - Fix bootstrap.py: use scipy.stats.norm.ppf instead of numpy.erfcinv - Fix bca_ci() call to use keyword argument for level parameter Implements Phase 3 and Phase 4 test coverage improvements from plan. All 77 tests passing.
- Add CLI helper function tests (test_cli_helpers.py) - Add training command tests (test_training_commands.py) - Add run-pipeline workflow tests (test_run_pipeline.py) - Extend dashboard tests with comprehensive coverage for init, plotting, and reporting methods
Rename local variables to avoid name conflicts in class definitions. This fixes NameError exceptions that were preventing test execution and coverage verification. All tests now pass and Phase 1 module coverage verification is complete.
[Stage: Week 1 - Release Preparation] - Update README with 645 tests and 87% coverage badge - Add ROADMAP.md with feature timeline - Create performance benchmarks for pipeline and bootstrap operations Next steps: Complete v0.5.1 release, add CI coverage gate, critical security fixes
changing artifact name to have unique id per job
Release v0.5.1: Test coverage improvements, security fixes, and CI enhancements - Update version to 0.5.1 in pyproject.toml and __init__.py - Update CHANGELOG.md with 645 tests and 87% coverage improvements - Update README.md with current metrics and version - Add coverage gate to CI (fail if < 85%, upload artifacts) - Create ROADMAP.md with planned features for 0.6.0-1.0.0 - Critical security fixes: filelock>=3.20.3, urllib3>=2.6.3, pip>=25.3 - Fix test_installation.py to expect version 0.5.1
- Fix: Make training module imports lazy to allow core package import without optional deps - Add: Structured exception types (DataValidationError, DependencyError) with context and suggestions - Add: Performance documentation (PERFORMANCE.md) and automated release workflow - Improve: Error messages with actionable suggestions and installation commands - Enhance: CI/CD with performance benchmarking and automated releases All tests passing. No breaking changes to public APIs.
… logging, and feedback collection - Security: Update dependencies, add security workflow and policy - Performance: Add performance test suite and profiling tools - Logging: Implement structured logging infrastructure - Feedback: Add GitHub issue templates and feedback process - Update CHANGELOG.md for v0.5.3 - Update release workflow to use modern action
CI/CD error ERROR: Could not find a version that satisfies the requirement virtualenv>=20.36.2 fix removing non-runtime dependency
Add property-based testing with Hypothesis - Expand integration tests with comprehensive edge case coverage - Set up Sphinx documentation site with GitHub Pages deployment - Configure Dependabot for automated dependency security updates - Add security review process and monthly automated reviews - Update documentation (CHANGELOG, README) for v0.5.4
Remove invalid update-types field from groups section. Dependabot automatically groups security updates, so explicit update-types configuration is not needed.
Bumps [softprops/action-gh-release](https://github.com/softprops/action-gh-release) from 1 to 2. - [Release notes](https://github.com/softprops/action-gh-release/releases) - [Changelog](https://github.com/softprops/action-gh-release/blob/master/CHANGELOG.md) - [Commits](softprops/action-gh-release@v1...v2) --- updated-dependencies: - dependency-name: softprops/action-gh-release dependency-version: '2' dependency-type: direct:production update-type: version-update:semver-major ... Signed-off-by: dependabot[bot] <support@github.com>
dependabot
Bot
added
dependencies
JobCollins
force-pushed
the
dependabot/github_actions/softprops/action-gh-release-2
branch
from
5af5e33 to
4ae72ab
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Bumps softprops/action-gh-release from 1 to 2.
Release notes
Sourced from softprops/action-gh-release's releases.
Changelog
Sourced from softprops/action-gh-release's changelog.
Commits
a06a81arelease 2.5.07da8983feat: mark release as draft until all artifacts are uploaded (#692)8797328chore(deps): bump actions/checkout in the github-actions group (#689)1bfc62achore(deps): bump the npm group across 1 directory with 5 updates (#697)5be0e66release 2.4.2af658b4feat: Ensure generated release notes cannot be over 125000 characters (#684)237aaccchore: bump node to 24.11.000362bechore(deps): bump the npm group with 5 updates (#687)0adea5achore(deps): bump the npm group with 3 updates (#686)aa05f9dchore(deps): bump actions/setup-node from 5.0.0 to 6.0.0 in the github-action...Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting
@dependabot rebase.Dependabot commands and options
You can trigger Dependabot actions by commenting on this PR:
@dependabot rebasewill rebase this PR@dependabot recreatewill recreate this PR, overwriting any edits that have been made to it@dependabot mergewill merge this PR after your CI passes on it@dependabot squash and mergewill squash and merge this PR after your CI passes on it@dependabot cancel mergewill cancel a previously requested merge and block automerging@dependabot reopenwill reopen this PR if it is closed@dependabot closewill close this PR and stop Dependabot recreating it. You can achieve the same result by closing it manually@dependabot show <dependency name> ignore conditionswill show all of the ignore conditions of the specified dependency@dependabot ignore this major versionwill close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself)@dependabot ignore this minor versionwill close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself)@dependabot ignore this dependencywill close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)