Releases: scikit-hep/pyhf
Releases · scikit-hep/pyhf
v0.4.4
This is a patch release from v0.4.3 → v0.4.4. Changes: - fix: Disallow cloudpickle v1.5.0 in 'tensorflow' extra (#915) - feat: Add precision information to tensorlib backends (#910) - ci: Verify build backend is compliant with PEP 517 in publishing workflow (#909) - fix: make pytorch_backend use configured dtypes (#906) - ci: Separate linting from testing (#903) - fix: Update infer.mle docstrings to match SciPy v1.5.0 (#904) - fix: Keep Sphinx at v3.0.X releases (#898) - fix: Explicitly avoid Sphinx v3.1.0 in 'docs' extra (#895) - ci: Use Python 3.8 in all steps of publishing distributions (#893) - docs: Add 1Lbb hepdata likelihood record (#889) - fix: Check if HEAD of origin/master is accessible from latest tag (#888)
v0.4.3
This is a patch release from v0.4.2 → v0.4.3. Changes: - fix: Set checkout GHA persist-credentials to false for push to succeed (#884) - feat: Add Patchset/Patch classes and Patchset CLI (#857) - fix: Avoid warning of 'app.add_stylesheet() is deprecated' (#885) - docs: Add simple examples for inspect, cls, and digest CLI (#877)
v0.4.2
This is a patch release from v0.4.1 → v0.4.2. Changes: - build: Require minimum SciPy version of v1.4.0 (#796) - ci: Use checkout v2.2.0 fetch-depth option to unshallow all (#880) - fix: Sanitize double quotes for Tag Creator (#875) - chore: Migrate metadata and setup options to setup.cfg (#873) - chore: Update setup-python action to v2 (#868) - fix: logging (#866) - docs: Update keywords, classifiers, and URLs in setup.py (#862) - ci: Add Python 3.8 to tests (#860) - docs: Add instructions for installing test releases from TestPyPI (#859) - feat: Add utility for computing object digests ("hashing") (#852) - fix: Remove dead OVH Binder build endpoint and add GESIS and Turing endpoints (#855) - chore: Update papermill to use v2 releases (#853) - docs: Improve discussion of publication workflow (#647) - feat: Add patchset schema and datadir pytest fixture (#847) - fix: Add AUTHORS to MANIFEST ignore (#851) - docs: Add docstrings and examples to infer fitting methods (#849) - docs: List the dev team as authors (#844) - ci: Use actions-gh-pages GitHub Action for deploy of docs (#843) - fix: Get full history for Tag Creator workflow (#842) - docs: Add FAQs on validation (#841) - refactor: Simplify Workspace::get_measurement and harmonize poi_name (#838) - fix: Have release versioning work in docs build (#839) - feat: Update jsonschema to non-alpha release (#837) - feat: Catch unknown kwargs while creating a model (#813) - revert: Remove disk space regression hack fix (#831) - chore: Update pypi-publish-action to non-alpha release (#834) - fix: Get full history and all tags for TestPyPI and PyPI publishing workflow (#833) - feat: Add curl to Docker image (#827) - chore: Use GitHub Actions checkout action v2 (#826) - chore: Remove redundant install of matplotlib from 'test' extra (#823) - fix: Replace use of raw directive in README.rst (#825) - docs: Use rST instead of Markdown for README and ROADMAP (#819) - docs: Add JAX to Python API docs (#816) - ci: Use the same Docker workflow for testing as publication (#812) - ci: Use official Docker GitHub Action for publication to Docker Hub (#808) - docs: Add citation from Run 2 Reinterpretation of LHC Results paper (#809) - chore: Avoid mutable defaults (#794) - docs: Add copy buttons to code snippets in docs with sphinx-copybutton (#721) - fix: Set correct workspace.version upon loading a workspace (#802) - refactor: Use pathlib over os.path for path (#798) - refactor: Use _TensorViewer in ParamViewer (#804) - build: Remove dependency on six package (#799) - chore: Use Debian based python:3.7-slim as base Docker image (#795) - chore: Clean up schemas and fix typos in docs about schemas (#793) - docs: Fix MANIFEST and put schemas back in documentation (#791) - feat: Add AsymptoticCalculator (#750) - fix: Removes invalid property in .zenodo.json (#789) - fix: Remove unneeded directory slash in .dockerignore (#788)
v0.4.1
This is a patch release from v0.4.0 → v0.4.1. Changes: - fix: Uncorrelated shape systematic with proper masking (#766) - feat: Add GPU enabled Dockerfile (#748) - test: Add test of independence of pyhf.infer from pyhf details (#768) - docs: Add published HEPData likelihoods to Use and Citations page (#781) - docs: Add list of media publications and make 'outreach' page (#778) - docs: Add Zenodo metadata config file (#779) - feat: Only allocate shapesys parameter if yield and uncrt are positive, nonzero (#775) - docs: remove misplaced equal sign (#776) - feat: Update required_parset to pass in sample data and modifier data (#774) - fix: Add key to workflow dictionary to make it valid JavaScript (#771) - fix: Minor typo in GitHub Actions removeLabels to removeLabel (#770) - docs: Add Stack Overflow question section (#759) - feat: Combine is taught about formal data (outer) join methods (#752) - docs: Add preferred BibTeX entry for citation (#758) - refactor: standalone spec parsing for parameters (#744) - chore: Use scikit-hep.org hosted badge SVG (#755) - docs: Add Scikit-HEP Project badge to README (#753) - chore: Run pyflakes over the full codebase (#737) - docs: Cover src/pyhf/contrib and src/pyhf/cli with pydocstest (#732)
v0.4.0
This is a minor release from v0.3.4 → v0.4.0. Changes: - feat: Drop Python 2.7 support (#704) - docs: Use pip best practices with python -m pip (#730) - fix: Public release canary should be using [backends] option (#728) - fix: iminuit optimiser is never verbose (#724)
v0.3.4
This is a patch release from v0.3.3 → v0.3.4. Changes: - feat: Add jax backend (#377)
v0.3.3
This is a patch release from v0.3.2 → v0.3.3. Changes: - feat: Migrate to TensorFlow v2.0 (#541) - chore: Make tests pass pyflakes (#718) - refactor: Use multistage Docker build to include only built libraries in Docker image (#717) - fix: Update nbformat to v4.4 for all tested notebooks (#723) - docs: Directly link to GitHub project page from docs website (#719) - docs: Apply pydocstyle to Workspace (#707)
v0.3.2
This is a patch release from v0.3.1 → v0.3.2. Changes: - feat: change qmu API and remove test_stat return (#712) - fix: Correct order of Codecov reporting in CI (#711) - feat: Add contrib module (#694) - refactor: Use MLE API (#690) - ci: workflow that runs bumpversion and pushes when issue filed (#650) - feat: double precision for ML backends (#701) - docs: Add docstrings for interpolators (#702) - docs: Add Altair example notebook (#699) - docs: Add docstrings for src/pyhf/infer (#698) - feat: Add ENTRYPOINT to pyhf Docker image (#700) - ci: Use Python 3 scipy Alpine image for base Docker image (#697) - ci: Add Codecov threshold tolerance (#693) - chore: Remove top level 'examples' dir (#688)
v0.3.1
Bump version: 0.3.0 → 0.3.1
v0.3.0
refactor: Create infer subpackage to hold our statistics code (#531) * Create a pyhf.infer subpackage * Move cli.stats to cli.infer * Bump version: 0.2.2 → 0.3.0