Antropy v0.2.2

Summary — v0.2.2 (April 2026)

Permutation entropy

Support for 2D arrays

• perm_entropy now accepts 2-D arrays of shape (n_channels, n_times) for order ∈ {3, 4}, computing entropy for every row in a single vectorised call.

Speed-up:

• perm_entropy fast path (only for order 3 or 4): ordinal patterns are now encoded via pairwise comparison bit-keys and counted with np.bincount rather than argsort + np.unique, giving a major speed-up
• perm_entropy general path (order > 4): the embedded matrix is now built as a zero-copy as_strided view and the hash is computed via a single @ product instead of an element-wise multiply + sum (faster).

Bugfixes

• perm_entropy now applies a small positional epsilon jitter before sorting, making results fully deterministic across platforms and dtypes. Previously, integer-typed signals (and float signals with exact ties) could give different outputs depending on the platform's unstable-sort tie-breaking behaviour.
• Normalized perm_entropy output is clipped to [0, 1] to prevent returning -0.0 for perfectly regular signals.

Docs

• Fixed doctest failures in spectral_entropy, num_zerocross, and hjorth_params caused by NumPy's new scalar repr (np.float64(x) / np.int64(x) instead of bare literals).
• Added a Big O complexity column to the performance tables, with footnotes for perm_entropy, svd_entropy, higuchi_fd, spectral_entropy, and app_entropy / sample_entropy.
• Performance tables now show timings at both 1 000 and 10 000 samples.
• Added perm_entropy docstring warning about tie-breaking and the jitter strategy.
• Added benchmarks/benchmark_all.py script to reproduce all table timings.
• Rewrote docs/index.rst to match README.rst in content and style.

Maintenance

• Clean up docs/conf.py (remove redundant sys.path manipulation) and tighten pyproject.toml (remove stale extras, fix editable-install metadata).

What's Changed

• DOC: Better README by @raphaelvallat in #49
• MAINT: Fix imports and clean up project configuration by @raphaelvallat in #50
• [v0.2.2] perm_entropy improvements; better doc/doctest by @raphaelvallat in #51

Full Changelog: v0.2.1...v0.2.2

Antropy v0.2.1

See full changelog at https://raphaelvallat.com/antropy/changelog.html

What's Changed

• DOC: Fix and improve docstrings across entropy, fractal, and utils modules by @raphaelvallat in #44
• CI: deploy documentation to GitHub Pages via GitHub Actions by @raphaelvallat in #45
• Code quality and API improvements by @raphaelvallat in #46
• FIX: correct off-by-one in Higuchi FD inner loop by @raphaelvallat in #47
• Release 0.2.1 by @raphaelvallat in #48

Full Changelog: v0.2.0...v0.2.1

AntroPy v0.2.0

v0.2.0 (March 2026)

Build & CI

• Drop Python 3.9 (EOL), add Python 3.13 support. Minimum is now Python 3.10.
• Switch from pip to uv in all GitHub Actions workflows.
• Add explicit minimum versions for core dependencies: numpy>=1.22.4, scipy>=1.8.0, scikit-learn>=1.2.0.
• Migrate [project.optional-dependencies] to PEP 735 [dependency-groups].
• Bump setuptools>=80.0.
• Split CI into three jobs: test-core (3 platforms × 4 Python versions), test-dependency-combinations (4 dep combos from minimum to latest), and coverage.
• Fix test-dependency-combinations job: separate antropy install (--no-deps) from test-dependency install so that pytest’s own dependencies (e.g. pluggy) are always resolved.
• Fix Codecov upload to use ${{ secrets.CODECOV_TOKEN }} instead of a hardcoded token.
• Switch Ruff workflow from astral-sh/ruff-action@v1 to uvx ruff via astral-sh/setup-uv@v7.
• Extend Ruff rules: add W (pycodestyle warnings) and NPY (NumPy rules).

Tests

• Increase test coverage from ~54 % to 100 %.
• Add tests/test_utils.py covering all branches of the _embed helper (1-D and 2-D paths, all error conditions).
• Add edge-case tests: sample_entropy returning inf (m-length matches exist but no (m+1)-length matches); detrended_fluctuation returning NaN for a constant signal; spectral_entropy raising on an invalid method string.
• Set NUMBA_DISABLE_JIT=1 in the coverage CI job so coverage.py can instrument Numba JIT function bodies; other CI jobs still exercise real compiled code.

Bug fixes

• Fix antropy.higuchi_fd() returning a ValueError (math domain error) on constant or integer-typed input arrays: log(0) is now guarded to return -inf, matching Numba’s IEEE 754 behaviour.

Docs

• Switch documentation theme from sphinx_bootstrap_theme to pydata-sphinx-theme (dark/light toggle, GitHub icon, improved layout).
• Fix three broken intersphinx URLs: NumPy, SciPy, and MNE-Python.
• Add sphinx.ext.mathjax for LaTeX math rendering in docstrings.
• Add contributing.rst guide.
• Fix stale :py:func: cross-references in changelog entries v0.1.1–v0.1.3 (entropy.XXX → antropy.XXX).
• Fix typo in v0.1.6 changelog (“Fox for KDTree” → “Fix for KDTree”).
• Modernize README.rst and docs/index.rst: add PyPI, conda-forge, downloads, and Ruff badges; add uv installation instructions; fix broken links.

v0.1.9

What's Changed

• MAINT: Remove stochastic hard dependency by @larsoner in #41

New Contributors

• @larsoner made their first contribution in #41

Full Changelog: v0.1.8...v0.1.9

v0.1.8

What's Changed

• Modern python packaging + v0.1.8 by @raphaelvallat in #38

Full Changelog: v0.1.7...v0.1.8

v0.1.7

What's Changed

• update SampEn and AppEn to allow for variable tolerance values by @JAC28 in #32
• Update Katz_FD by @PiethonProgram in #36
• Release v0.1.7 by @raphaelvallat in #37

New Contributors

• @JAC28 made their first contribution in #32
• @PiethonProgram made their first contribution in #36

Full Changelog: v0.1.6...v0.1.7

v0.1.6

This release requires numba >= 0.57.

What's Changed

• Allow readonly arrays in numba jit signature by @jvdd in #23
• Improved sample entropy kernel by @FirefoxMetzger in #25
• fix: KDTree.valid_metrics is method since sklearn 1.3 by @jvdd in #30
• Replaced Python 3.7 with 3.10 in Github Actions
• Fix pandas deprecation in notebooks

New Contributors

• @FirefoxMetzger made their first contribution in #25

Full Changelog: v0.1.5...v0.1.6

v0.1.5

This is a minor release.

What's Changed

• Handle the limit of p = 0 in p log2 p by @jftsang in #3
• Correlation between entropy/FD metrics for data traces from Hodgin-Huxley model by @antelk in #5
• Fix docstrings and rerun by @antelk in #7
• Improve performance in _xlog2x by @jftsang in #8
• Prevent invalid operations in xlogx by @guiweber in #11
• Allow readonly arrays for higuchi_fd by @jvdd in #13
• modify the _embed function to fit the 2d input by @cheliu-computation in #15
• Fixed division by zero in linear regresion function (with test) by @Arritmic in #21
• Add conda install instructions by @raphaelvallat in #19

New Contributors

• @jftsang made their first contribution in #3
• @antelk made their first contribution in #5
• @guiweber made their first contribution in #11
• @jvdd made their first contribution in #13
• @cheliu-computation made their first contribution in #15
• @Arritmic made their first contribution in #21
• @raphaelvallat made their first contribution in #19

Full Changelog: v0.1.4...v0.1.5

v0.1.4

This new release includes a faster implementation of the LZ complexity algorithm (see #1).