A Python-embedded modeling language for convex optimization problems.
Full Changelog: https://github.com/cvxpy/cvxpy/compare/v1.3.3...v1.3.4
This is a patch release for 1.4. The patch includes a variety of fixes:
Full Changelog: https://github.com/cvxpy/cvxpy/compare/v1.4.2...v1.4.3
This is a patch release for 1.4. The patch includes a variety of fixes:
SCIP return solution if [total]nodelimit hit (https://github.com/cvxpy/cvxpy/pull/2279) Update swig autogenerated files to swig 4.1.1 (https://github.com/cvxpy/cvxpy/pull/2273) Update contributing docs to reflect usage of ruff over flake8/isort (https://github.com/cvxpy/cvxpy/pull/2302) Change some O(n) lookups in scip_conif.py to O(1) (https://github.com/cvxpy/cvxpy/pull/2313) Avoid sparse matrix multiply for identity matrix (https://github.com/cvxpy/cvxpy/pull/2315) Restore Mosek 9 compatibility (https://github.com/cvxpy/cvxpy/pull/2325) Reformatting the installation page (https://github.com/cvxpy/cvxpy/pull/2268) Updating NAG website links in doc (https://github.com/cvxpy/cvxpy/pull/2299) Fix Gurobi interface issues (https://github.com/cvxpy/cvxpy/pull/2300) Use new NumPy API (https://github.com/cvxpy/cvxpy/pull/2319) Fixing links to cvx_short_course (https://github.com/cvxpy/cvxpy/pull/2264)
This is a patch release for 1.4. The patch includes fixes from several contributors:
@goulart-paul Change conic solver preference order (#2259) @maxschaller Include use_quad_obj in cache (#2262)
This release is consistent with our semantic versioning guarantee. It comes packed with many new features, bug fixes, and performance improvements. This version of CVXPY supports Python 3.8 through 3.12, and is our first release that supports Python 3.12. While working on the next release, we continue to officially support CVXPY 1.3 and 1.4.
convolve
mean
outer
ptp
std
var
vec_to_upper_tri
.sum(), .max(), and .mean()
PIQP
PowerConeND now supports extracting its dual variablesreshape now supports using -1 as a dimension, with the same meaning it has in NumPyperspective atom now supports s=0
CVXPY has used ECOS as the default solver for many years; however, it has known issues with performance and numerical stability in edge cases. Recently, a new solver, Clarabel, that improves the algorithm and implementation of ECOS has been under development.
In 1.5, CVXPY plans to start using Clarabel instead of ECOS by default for some categories of problems. In 1.6, we plan to no longer install ECOS as a CVXPY dependency. We have no plans to remove support for calling ECOS as a solver. As part of this transition, in 1.4 CVXPY will raise a warning whenever ECOS is called by default. We encourage you to try and use Clarabel instead, but if you're dependent on ECOS's exact behavior please explicitly specify it as a solver.
conv deprecationThe CVXPY atom conv is inconsistent with NumPy's convolve functions. We are deprecating it, but have no plans to remove it in the short term. We encourage all users to use the CVXPY atom convolve instead.
NonPos deprecationThe NonPos cone uses the opposite dual variable sign convention as the rest of the CVXPY cones and a constraint of NonPos(expr) is the same as a constraint on NonNeg(-expr). We are deprecating NonPos, but have no plans to remove it in the short term. We encourage users to switch to using NonNeg.
This release would not have been possible without the contributions of many CVXPY users and developers. Across 29 contributors and 116 PRs, we would like to thank the following people for their contributions to this release (in alphabetical order):
Special shoutout to our Google Summer of Code interns @aryamanjeendgar and @Transurgeon!
This is a patch release for 1.1. The patch includes fixes from many contributors:
This is a patch release for 1.2. The patch includes fixes from many contributors:
This is a patch release for 1.3. The patch includes fixes from many contributors:
This is a patch release for 1.1. The patch includes fixes from many contributors:
@SteveDiamond Fix cvxpy base deployment https://github.com/cvxpy/cvxpy/pull/2071 @Transurgeon Typo fix in documentation https://github.com/cvxpy/cvxpy/pull/2044 @SteveDiamond Fix Mosek power cone https://github.com/cvxpy/cvxpy/pull/2043 @aszekMosek Clean up Mosek parameter settings https://github.com/cvxpy/cvxpy/pull/2029 @jlchen0 Fix p-norm issue with DGP problems https://github.com/cvxpy/cvxpy/pull/2028 @SteveDiamond Unpin setuptools https://github.com/cvxpy/cvxpy/pull/2022 @samuel-adekunle Update atomic functions docs https://github.com/cvxpy/cvxpy/pull/2016 @phschiele Build wheels on push https://github.com/cvxpy/cvxpy/pull/2008 @SteveDiamond Fix issue with matrix multiplication involving quad https://github.com/cvxpy/cvxpy/pull/2067 @PTNobel Adds error on floats as indices https://github.com/cvxpy/cvxpy/pull/2058 @h-vetinari Fixes for 1.3.0 https://github.com/cvxpy/cvxpy/pull/1998
This is a patch release for 1.2. The patch includes fixes from many contributors:
@SteveDiamond Fix cvxpy base deployment https://github.com/cvxpy/cvxpy/pull/2071 @Transurgeon Typo fix in documentation https://github.com/cvxpy/cvxpy/pull/2044 @SteveDiamond Fix Mosek power cone https://github.com/cvxpy/cvxpy/pull/2043 @aszekMosek Clean up Mosek parameter settings https://github.com/cvxpy/cvxpy/pull/2029 @jlchen0 Fix p-norm issue with DGP problems https://github.com/cvxpy/cvxpy/pull/2028 @SteveDiamond Unpin setuptools https://github.com/cvxpy/cvxpy/pull/2022 @samuel-adekunle Update atomic functions docs https://github.com/cvxpy/cvxpy/pull/2016 @phschiele Build wheels on push https://github.com/cvxpy/cvxpy/pull/2008 @SteveDiamond Fix issue with matrix multiplication involving quad https://github.com/cvxpy/cvxpy/pull/2067 @PTNobel Adds error on floats as indices https://github.com/cvxpy/cvxpy/pull/2058 @rileyjmurray Remove ill-posed xexp test case https://github.com/cvxpy/cvxpy/pull/2001 @h-vetinari Fixes for 1.3.0 https://github.com/cvxpy/cvxpy/pull/1998