Changelog

1.6.0 (2025-06-24)

Features

  • add CV method to Slope class (#37) (a4752da)

  • add plotting for cv and path objects (#38) (70c5bef)

  • enable feature screening toggling (16a9820)

  • set update_clusters to True by default (b34ff38)

1.5.0 (2025-06-16)

Features

  • enable randomized CD type (a4abef1)

  • update libslope to v3.0.0 (040ecc6)

1.4.1 (2025-06-11)

Bug Fixes

  • correctly pass update_clusters flag (60d544e)

1.4.0 (2025-06-11)

Features

  • add update_clusters to Slope() (de4d706)

1.3.1 (2025-06-11)

Bug Fixes

  • update libslope to v2.8.2 (c085652)

1.3.0 (2025-06-08)

Features

  • update libslope to v2.8.0 (02a5a1b)

1.2.0 (2025-05-30)

Features

  • update libslope to v2.7.0 (9387915)

1.1.1 (2025-05-30)

Reverts

  • “ci: only trigger release on workflow_dispatch” (359a31c)

1.1.0 (2025-05-27)

Features

  • update libslope to v2.6.0 (899bfce)

Bug Fixes

1.0.1 (2025-04-23)

Bug Fixes

  • actually pass lambda_type to c++ api (4903081)

1.0.0 (2025-04-23)

⚠ BREAKING CHANGES

  • path: allow setting maximum clusters for path method

Features

  • path: allow setting maximum clusters for path method (3eb28c4)

0.7.0 (2025-04-23)

⚠ BREAKING CHANGES

  • change default of theta1 to 1.0

Features

  • change default of theta1 to 1.0 (7eb0930)

  • make lambda type selectable (f516a3e)

0.6.0 (2025-04-22)

Features

  • enable setting solver via solver (70eb844)

0.5.1 (2025-04-22)

Bug Fixes

0.5.0 (2025-04-22)

Features

  • add path early stopping criteria (a6b690a)

0.4.0 (2025-04-22)

Features

  • add poisson, logistic, multinomial losses (9858ae7)

  • enable setting the q parameter (05734fb)

0.3.1 (2025-04-11)

Bug Fixes

  • don’t use value() since it does not exist on early macos (034c9af)

Reverts

  • “ci: add pytest to cibuildwheel” (631f2cb)

  • “ci: update ci workflows: (6bbfedd)

0.3.0 (2025-04-10)

⚠ BREAKING CHANGES

  • change default centering and scaling to "none"

  • drop support for python 3.7 and 3.8

  • deprecate standardize in Slope estimator

  • change loss value parameterization

  • update libslope to v0.21.0

  • allow only single float input for alpha

Features

  • add slope path (#31) (f407d8c)

  • allow only single float input for alpha (3f65287)

  • change default centering and scaling to "none" (7e27318)

  • change loss value parameterization (3f65287)

  • deprecate standardize in Slope estimator (3f65287)

  • update Eigen to 3.4.0 (5ce41d2)

  • update libslope to v0.21.0 (3f65287)

  • update libslope to v0.30.0 (307f36b)

  • update libslope to v0.30.1 (f012ba6)

Bug Fixes

  • add missing sources to pybind11 module (a08be2c)

  • fix convergence issues with solver (3f65287)

Dependencies

  • cet scikit-build-core minversion to 0.10 and remove cmake ver (bd63a29)

  • drop support for python 3.7 and 3.8 (139e0d8)

Documentation

  • document as compliantwith python 3.13 (87e9219)

  • update copyright year (cd707c4)

0.2.3 (2024-02-07)

Bug Fixes

  • deps: update to libslope 0.3.1 (bac6f93)

  • update libslope to 0.3.2 (c059f6a)

Documentation

  • make logo work better for mobile phones (765077a)

0.2.2 (2023-12-14)

Reverts

  • “ci: build documentation once through ci to have something on the page” (118a416)

Documentation

  • add changelog to docs (defdaae)

  • add codecov badge to readme (7ac3148)

  • add file-level docstring from estimators.py (fa98821)

  • add logos and fix duplicate documentation (50b29eb)

  • add source code link in docs (4b61204)

  • don’t use sphinx-apidoc (4c168ac)

  • fix autosummary documentation (87d5cf3)

  • fix SLOPE -> sortedl1, and write about convcommits (15c1166)

  • fix typo and pip instructions (4e78858)

  • specify main branch for badge in readme (abdc18e)

  • update logo (7ad646c)

  • update readme with badges and pypi info (c3edd6e)

  • use ivar for napoleon (c6d8522)

  • use myst parser (0288739)

0.2.1 (2023-12-11)

Documentation

0.2.0 (2023-12-11)

Features

  • add max_iter and tol parameters to Slope (46ec82d)

  • make estimator fully scikit-learn compliant (and test it) (0e236fe)

  • support sparse X input (8d2cf08)

  • support standardization and intercept (99f9aee)

  • update to libslope 0.2.0 (f5ecaad)

Bug Fixes

  • return intercept as float (5e83de1)

  • update to libslope 0.3.0 (905424f)

Documentation

  • add a few more classifiers (42ad653)

  • add docstring for module (7294370)

  • add some additional classifiers (e649ffd)

  • add some additional files for pyproject.toml (a9ea635)

  • add some basic information on readme (8f4c4e0)

  • clean up documentation and add copybutton (4e51a29)

  • document att attributes (4b881f2)

  • document tests (027a2da)

  • don’t say what defaults are in documentation (8fa20cd)

  • fix documentation for slope (1069aff)

  • fix linting error for estimators.py (eba539b)

  • fix sorting of python version (9ef1433)

  • increment header in readme (b94c9b1)

  • polish readme a bit (4cceb1b)

  • switch to furo theme (6006fd9)

  • update authors and make version dynamic (a29a16b)

  • use static version instead (6071f9d)

0.1.0 (2023-11-30)

Features

  • add a docs workflow (7981b47)

  • add basic package infrastructure (e53a245)

  • add initial c++ source for libslope (a118ed4)

  • add requirements to extras for building documentation (a5362bc)

  • change to sklearn-compatible interface (41d4836)

  • make cpp function return a tuple with actual estimates (3c3ff24)

Bug Fixes

  • fix test to work with new api (3192911)

  • move installation dependencies to setup.py (2506902)

  • use scikit-learn, not sklearn, as dependency (dc78b49)

Documentation

  • add initial docs (b4a5bac)

  • change estimators to models (24c57ec)

  • change to sphinx build/source structure (ed9ab31)

  • remove documentation stubs from cpp file (69bf98c)

  • run apidoc for correct location (4246e41)