Skip to content

Release and reproducibility

Report both versions

For reproducibility, report both:

  • the installed eyeprocesspy version; and
  • eyeprocesspy.__r_reference_version__.

For release 0.1.0 the frozen R reference is 0.11.1.

Release gate

The public release is built only after the following checks are green:

uv sync --extra dev
uv run ruff check src tests
uv run pytest
uv run pytest --cov=eyeprocesspy --cov-branch --cov-report=term-missing --cov-fail-under=100
uv build
uv run python -m twine check dist/*

CI additionally verifies the frozen R oracle and runs the Python test matrix on Ubuntu, Windows and macOS with Python 3.11–3.14.

For v0.1.0, the controlling release evidence is 1,458 passing tests, 23,085 / 23,085 statements, and 9,680 / 9,680 branches covered.

Published artifacts

Release 0.1.0 is available through three coordinated publication surfaces:

  • GitHub Release: https://github.com/stefanosbalaskas/eyeprocesspy/releases/tag/v0.1.0
  • PyPI: https://pypi.org/project/eyeprocesspy/0.1.0/
  • Zenodo archive: https://doi.org/10.5281/zenodo.22285167

The release contains both a source distribution (eyeprocesspy-0.1.0.tar.gz) and a universal Python wheel (eyeprocesspy-0.1.0-py3-none-any.whl). The GitHub release is the source-control release of record, PyPI provides the installable distribution, and Zenodo provides the immutable research-software archive and DOI.

PyPI

The repository uses GitHub Actions OpenID Connect / PyPI Trusted Publishing. No long-lived PyPI API token is stored in the repository.

Install the release with:

pip install eyeprocesspy

or pin the exact initial release:

pip install eyeprocesspy==0.1.0

Zenodo

Zenodo archived v0.1.0 and minted the release DOI 10.5281/zenodo.22285167. The DOI is recorded in CITATION.cff and surfaced from the README. .zenodo.json remains the repository-side metadata source for future GitHub-to-Zenodo releases.

Integrity

The package contains validation-manifest and freeze utilities for study-level evidence. Native RDS remains R-specific; Python persistence formats are labelled honestly and are not disguised as RDS.