v0.6.1

Patch release adding reference and ground-truth quality checks.

Added:

• Reference and ground-truth comparison helpers through ReferenceValue, compare_to_reference(), compare_to_references(), and evidence_from_reference_comparison().
• Aggregated reference quality evidence for multi-field records, mapping predefined reference checks into consistency and anomaly-detection evidence.
• Documentation for scientific use, value-quality checks, and the distinction between payload integrity and measurement quality.

Verification:

• pytest: 39 passed
• mkdocs build --strict passed
• twine check passed for 0.6.1 wheel and sdist

v0.6.0

v0.6.0

Added

• Score explainability through ReliabilityScanner.score_breakdown().
• Evidence hashes, evidence snapshots, evidence confidence, and uncertainty metadata on scored records.
• SQL and document audit exports for policy decisions.
• Expanded SQL reliability metadata columns for evidence hash, evidence snapshot, confidence, and uncertainty.
• dri CLI for scanning JSON and JSONL records.
• Quarantine workflow and external data-quality mapping examples.
• Golden profile tests for default, scientific, and climate-record tier behavior.
• Dependency review workflow for pull requests.
• GitHub artifact attestations for release distributions.
• Documentation for custom profiles, score breakdowns, quarantine workflows, CLI usage, and release provenance.

Changed

• Updated packaging metadata to include explicit license files with the current setuptools backend baseline.
• Hardened GitHub Actions permissions to least-privilege read access where possible.
• Improved SQL boolean restoration for common database string values.

v0.5.0

v0.5.0

Added

• New core-features infographic in the README.
• Evidence templates for verified sensors, trusted APIs, cleaned datasets, user submissions, historical records, and climate station observations.
• Template registry helpers for creating ValidationEvidence without hand-tuning every score.
• Template API documentation and tests.

Verification

• pytest: 27 passed
• mkdocs build --strict: passed
• twine check: passed

v0.4.0

v0.4.0

Added

• Profile name and version metadata on each scored record.
• HMAC-SHA256 helpers for authenticated payload verification.
• Batch and streaming row scanning helpers for larger datasets.
• SQL DDL helpers for SQLite, PostgreSQL, MySQL, and DuckDB reliability columns.
• Optional extras for PostgreSQL, MySQL, DuckDB, MongoDB, Polars, and PyArrow integrations.
• Integration tests for SQL rows, document metadata, DataFrames, row dictionaries, and sequence payloads.

Verification

• pytest: 22 passed
• mkdocs build --strict: passed
• twine check: passed

v0.3.0

v0.3.0

Added

• Structured ReliabilityDecision results for policy audits and ingestion logs.
• Driver-neutral database helpers for SQL columns and document-store metadata.
• Typed SDK documentation for database usage and security guidance.
• CI coverage for Python 3.13 and 3.14.

Changed

• Refined SDK packaging so pandas, FastAPI, and uvicorn are optional extras.
• Updated README and docs for the v0.3.0 release.

v0.2.0

v0.2.0

Adds the first implemented Data Reliability Index scanning engine.

Added

• Reliability scanning engine for computing scores from validation evidence.
• Automatic trust-tier assignment from score and verification signals.
• Trace hash computation and verification support.
• ReliableData wrapper and policy resolution helper.
• Typed API documentation for the scanning engine.

Verification

• Local tests: 7 passed.
• Package artifacts: wheel and sdist built.
• Twine artifact check: passed.
• Docs build: mkdocs strict build passed.

Links

• Documentation: https://h3pdesign.github.io/data-reliability-index/
• PyPI: https://pypi.org/project/data-reliability-index/

Install after PyPI publishing is configured:

pip install data-reliability-index

v0.1.0

v0.1.0

Initial release of Data Reliability Index.

Added

• Pydantic reliability metadata and policy models.
• Pandas filtering helper.
• FastAPI ingestion example.
• PEP 561 typed package marker.
• MkDocs typed API documentation.
• CI checks for tests, package artifacts, and docs.
• PyPI trusted publishing workflow.

Links

• Documentation: https://h3pdesign.github.io/data-reliability-index/
• PyPI: https://pypi.org/project/data-reliability-index/

Install with:

pip install data-reliability-index