v0.7.0

What's Changed

• Move from Pybind11 to nanobind by @ManifoldFR in #345
• CMake: fix use of system cereal by @nim65s in #352
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci in #353
• Fix build issue on windows by @fabinsch in #357
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci in #355
• Release GIL when doing standalone solves by @jberg5 in #359
• [reopen2] Release GIL when doing standalone solves by @fabinsch in #363
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci in #365
• release: updates by @nim65s in #361
• ci: setup miniconda with conda-forge and remove defaults by @fabinsch in #366

New Contributors

• @jberg5 made their first contribution in #359

Full Changelog: v0.6.7...v0.7.0

ProxSuite 0.6.7

What's Changed

Added

• Fix mu update function for PrimalLDLT backend (#349)
• Allow use of installed pybind11, cereal and jrl-cmakemodules via cmake
• Add compatibility with jrl-cmakemodules workspace (#339)
• Specifically mention that timings are in microseconds (#340)
• Fix cereal include directory in cmake (#340)
• Extend doc with hint for conda installation from source (#340)

Fixed

• Fix inequality constraints return in QPLayer (#343)

Changed

• Refactor Python examples with a new "util.py" file (#347)

Full Changelog: v0.6.7...v0.6.7

ProxSuite 0.6.6

What's Changed

Fixed

• Fix infeasibility detection and add a unit test (#328)

Full Changelog: v0.6.5...v0.6.6

ProxSuite 0.6.5

[0.6.5] - 2024-05-31

Added

• Pip wheels for Python 3.12 and stop support Python 3.7 (#324)

Fixed

• Fixes compilation issue with GCC 14 on Arch (#322)

What's Changed

• Change from torch.Tensor to torch.empty or torch.tensor and specify type explicitly (#308)
• Fix handling of batch of inequality constraints in QPFunctionFn_infeas. The derivations in qplayer was done for single-sided constraints, that's the reason for the concatenation but the expansion of batchsize dimension was not working properly (#308)
• Switch from self-hosted runner for macos-14-ARM to runner from github (#306)
• Fix missing cassert for some compilers (#316)

Full Changelog: v0.6.4...v0.6.5

ProxSuite 0.6.4

What's Changed

Fixed

• Changed primal_infeasibility_solving to False for feasible QPs (#302)

Full Changelog: v0.6.3...v0.6.4

ProxSuite 0.6.3

What's Changed

Fixed

• Fix Python tests with scipy>=1.12 (#299)

Full Changelog: v0.6.2...v0.6.3

ProxSuite 0.6.2

What's Changed

Fixed

• Fix Windows build (#290)
• Fix math formulae in documentation (#294)
• Restore correc values for infeasibility (#292)
• Handles CPU/GPU transfer in QPFunctionFn's backward function (#297)

Full Changelog: v0.6.1...v0.6.2

ProxSuite 0.6.1

What's Changed

• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci in #280
• Templating power iteration algorithm by matrix storage order by @quentinll in #279

New Contributors

• @quentinll made their first contribution in #279

Full Changelog: v0.6.0...v0.6.1

ProxSuite 0.6.0

News

We add the implementation of QPLayer. QPLayer enables to use a QP as a layer within standard learning architectures. QPLayer allows for parallelized calculus over CPUs, and is interfaced with PyTorch. QPLayer can also differentiate over LPs.

What's Changed

• QPLayer: efficient differentiation of convex quadratic optimization by @fabinsch, @Bambade and @quentinll in #264

Full Changelog: v0.5.1...v0.6.0

ProxSuite 0.5.1

What's Changed

• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci in #265
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci in #268
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci in #269
• Check model is_valid up to eps by @fabinsch in #272

Full Changelog: v0.5.0...v0.5.1