optfuncs 0.0.3

Optimization benchmark functions and pytest-oriented evaluation helpers.

optfunc

optfunc provides differentiable benchmark functions for unconstrained optimization experiments, along with pytest-oriented checks you can reuse when validating an optimizer against known global minima.

Install

pip install "optdunc-hazy[torch-cpu]"

Then import it with:

import optfunc

Local development

uv build --no-sources
uv run pytest

Release flow

1. Update the package version in pyproject.toml, or run uv version <version> --frozen.
2. Create and push a Git tag named v<version>.
3. CNB will publish that tag to PyPI through the tag_push pipeline.

Benchmark definitions are adapted from SFU's optimization benchmark collection.

---

uv version z.x.y --frozen
rm -rf dist build src/*.egg-info
uv lock

git tag -a vz,x,y -m "Release vX.Y.Z"
git push origin vz,x,y