summaryrefslogtreecommitdiff
path: root/ot/bregman.py
AgeCommit message (Collapse)Author
2023-03-09[MRG] Semi-relaxed (fused) gromov-wasserstein divergence and improvements of ↵Cédric Vincent-Cuaz
gromov-wasserstein solvers (#431) * maj gw/ srgw/ generic cg solver * correct pep8 on current state * fix bug previous tests * fix pep8 * fix bug srGW constC in loss and gradient * fix doc html * fix doc html * start updating test_optim.py * update tests gromov and optim - plus fix gromov dependencies * add symmetry feature to entropic gw * add symmetry feature to entropic gw * add exemple for sr(F)GW matchings * small stuff * remove (reg,M) from line-search/ complete srgw tests with backend * remove backend repetitions / rename fG to costG/ fix innerlog to True * fix pep8 * take comments into account / new nx parameters still to test * factor (f)gw2 + test new backend parameters in ot.gromov + harmonize stopping criterions * split gromov.py in ot/gromov/ + update test_gromov with helper_backend functions * manual documentaion gromov * remove circular autosummary * trying stuff * debug documentation * alphabetic ordering of module * merge into branch * add note in entropic gw solvers --------- Co-authored-by: Rémi Flamary <remi.flamary@gmail.com>
2023-02-23[MRG] Added argument for warmstart of dual vectors in Sinkhorn-based methods ↵Huy Tran
in `ot.bregman` (#437) * Allow warmstart in sinkhorn and sinkhorn_log * Added argument for warmstart of dual vectors in Sinkhorn-based methods in * Add the number of the PR * [WIP] CO-Optimal Transport * Revert "[WIP] CO-Optimal Transport" This reverts commit f3d36b2705013409ac69b346585e311bc25fcfb7. * reformat with PEP8 * Fix W291 trailing whitespace error in pep8 test * Rearange position of warmstart argument and edit its description --------- Co-authored-by: Rémi Flamary <remi.flamary@gmail.com>
2022-12-23[MRG] Fix warning bug in sinkhorn2 (#417)arincbulgur
* Pass warn argument downstream in sinkhorn2 method. * releases.md * Fix unittest. Co-authored-by: Rémi Flamary <remi.flamary@gmail.com>
2022-12-21[MRG] Fix#421 pass stopThr to the sinkhorn function in ↵Théo Lacombe
empirical_sinkhorn_divergence (#422) * fix stopThr hardcoded in some places * added fix documentation in RELEASES.Md Co-authored-by: Rémi Flamary <remi.flamary@gmail.com>
2022-12-20[MRG] Change Doc empirical_sinkhorn_div to reflect what is computed (#423)Théo Lacombe
* Changes empirical_sinkhorn_divergence doc to mention that the entroipc term is not taken into account. * change typo regularization->contribution Co-authored-by: Rémi Flamary <remi.flamary@gmail.com>
2022-08-17debug sinkhorn divergence gradientsRémi Flamary
2022-07-27[MRG] Free support Sinkhorn barycenters (#387)Eduardo Fernandes Montesuma
* Adding function for computing Sinkhorn Free Support barycenters * Adding exampel on Free Support Sinkhorn Barycenter * Fixing typo on free support sinkhorn barycenter example * Adding info on new Free Support Barycenter solver * Removing extra line so that code follows pep8 * Fixing issues with pep8 in example * Correcting issues with pep8 standards * Adding tests for free support sinkhorn barycenter * Adding section on Sinkhorn barycenter to the example * Changing distributions for the Sinkhorn barycenter example * Removing file that should not be on the last commit * Adding PR number to REALEASES.md * Adding new contributors * Update CONTRIBUTORS.md Co-authored-by: Rémi Flamary <remi.flamary@gmail.com>
2022-05-09[MRG] Fix barycenter mass (#375)Hicham Janati
* fix transpose in sinkhorn barycenters * add test for assymetric cost barycenters * fix pep8 Co-authored-by: Hicham Janati <hicham.janati@inria.fr>
2022-03-24[MRG] Domain adaptation and unbalanced solvers with backend support (#343)Nathan Cassereau
* First draft * Add matrix inverse and square root to backend * Eigen decomposition for older versions of pytorch (1.8.1 and older) * Corrected eigen decomposition for pytorch 1.8.1 and older * Spectral theorem is a thing * Optimization * small optimization * More functions converted * pep8 * remove a warning and prepare torch meshgrid for future torch release (which will change default indexing) * dots and pep8 * Meshgrid corrected for older version and prepared for future versions changes * New backend functions * Base transport * LinearTransport * All transport classes + pep8 * PR added to release file * Jcpot barycenter test * unbalanced with backend * pep8 * bug solve * test of domain adaptation with backends * solve bug for tic toc & macos * solving scipy deprecation warning * solving scipy deprecation warning attempt2 * solving scipy deprecation warning attempt3 * A warning is triggered when a float->int conversion is detected * bug solve * docs * release file updated * Better handling of float->int conversion in EMD * Corrected test for is_floating_point * docs * release file updated * cupy does not allow implicit cast * fromnumpy * added test * test da tf jax * test unbalanced with no provided histogram * using type_as argument in unif function correctly * pep8 * transport plan cast in emd changed behaviour, now trying to cast as histogram's dtype, defaulting to cost matrix Co-authored-by: Rémi Flamary <remi.flamary@gmail.com>
2021-12-09[MRG] Tensorflow backend & Benchmarker & Myst_parser (#316)Nathan Cassereau
* First batch of tf methods (to be continued) * Second batch of method (yet to debug) * tensorflow for cpu * add tf requirement * pep8 + bug * small changes * attempt to solve pymanopt bug with tf2 * attempt #2 * attempt #3 * attempt 4 * docstring * correct pep8 violation introduced in merge conflicts resolution * attempt 5 * attempt 6 * just a random try * Revert "just a random try" This reverts commit 8223e768bfe33635549fb66cca2267514a60ebbf. * GPU tests for tensorflow * pep8 * attempt to solve issue with m2r2 * Remove transpose backend method * first draft of benchmarker (need to correct time measurement) * prettier bench table * Bitsize and prettier device methods * prettified table bench * Bug corrected (results were mixed up in the final table) * Better perf counter (for GPU support) * pep8 * EMD bench * solve bug if no GPU available * pep8 * warning about tensorflow numpy api being required in the backend.py docstring * Bug solve in backend docstring * not covering code which requires a GPU * Tensorflow gradients manipulation tested * Number of warmup runs is now customizable * typo * Remove some warnings while building docs * Change prettier_device to device_type in backend * Correct JAX mistakes preventing to see the CPU if a GPU is present * Attempt to solve JAX bug in case no GPU is found * Reworked benchmarks order and results storage & clear GPU after usage by benchmark * Add bench to backend docstring * better benchs * remove useless stuff * Better device_type * Now using MYST_PARSER and solving links issue in the README.md / online docs
2021-11-03[MRG] Docs updates (#298)ncassereau-idris
* bregman docs * sliced docs * docs partial * unbalanced docs * stochastic docs * plot docs * datasets docs * utils docs * dr docs * dr docs corrected * smooth docs * docs da * pep8 * docs gromov * more space after min and argmin * docs lp * bregman docs * bregman docs mistake corrected * pep8 Co-authored-by: Rémi Flamary <remi.flamary@gmail.com>
2021-11-03[WIP] Add debiased barycenter (Sinkhorn + convolutional sinkhorn) (#291)Hicham Janati
* add debiased sinkhorn barycenter + make loops pythonic * add debiased arg in tests * add 1d and 2d examples of debiased barycenters * fix doctest * fix flake8 * pep8 + make func private + add convergence warnings * remove rel paths + add rng + pylab to pyplot * fix stopping criterion debiased * pass alex * change params with new API * add logdomain barycenters + separate debiased API * test new API * fix jax read-only ? * raise error for jax * test catch jax error * fix pytest catch error * fix relative path * fix flake8 * add warn arg everywhere * fix ref number * catch warnings in tests * add contrib to readme + change ref number * fix convolution example + gallery thumbnails * increase coverage * fix flake Co-authored-by: Hicham Janati <hicham.janati@inria.fr> Co-authored-by: Rémi Flamary <remi.flamary@gmail.com> Co-authored-by: Alexandre Gramfort <alexandre.gramfort@m4x.org>
2021-11-02[MRG] Backend for gromov (#294)ncassereau-idris
* bregman: small correction * gromov backend first draft * Removing decorators * Reworked casting method * Bug solve * Removing casting * Bug solve * toarray renamed todense ; expand_dims removed * Warning (jax not supporting sparse matrix) moved * Mistake corrected * test backend * Sparsity test for older versions of pytorch * Trying pytorch/1.10 * Attempt to correct torch sparse bug * Backend version of gromov tests * Random state introduced for remaining gromov functions * review changes * code coverage * Docs (first draft, to be continued) * Gromov docs * Prettified docs * mistake corrected in the docs * little change Co-authored-by: Rémi Flamary <remi.flamary@gmail.com>
2021-10-27[WIP] Sinkhorn in log space (#290)Rémi Flamary
* adda sinkhorn log and working sinkhorn2 function * more tests pass * more tests pass * it works but not by default yet * remove warningd * update circleci doc * update circleci doc * new sinkhorn implemeted but not by default * better * doctest pass * test doctest * new test utils * remove pep8 errors * remove pep8 errors * doc new implementtaion with log * test sinkhorn 2 * doc for log implementation
2021-10-25[MRG] Bregman backend (#280)ncassereau-idris
* Bregman * Resolve conflicts * Bug solve * Bregman updated for JAX compatibility * Tests coherence between backend improved * No longer enforcing 64 bits operations on Jax except for tests * Now using mixtures, to make backend dependent tests with less code * Better test skipping code * Pep8 + test optimizations * redundancy removed * Docs * Typo corrected * Typo * Typo * Docs * Docs * pep8 * Backend docs * Prettier docs * Mistake corrected * small changes * Better wording Co-authored-by: Rémi Flamary <remi.flamary@gmail.com>
2021-09-30[MRG] GPU bugs solve (#288)ncassereau-idris
* gpus tests now passing * pep8 compliance * GPU tests succeeding even if b has rank higher than 1 Co-authored-by: Rémi Flamary <remi.flamary@gmail.com>
2021-06-14[MRG] Batch/Lazy Log Sinkhorn Knopp on samples (#259)Huy Tran
* Add batch implementation of Sinkhorn * Reformat to pep8 and modify parameter * Fix error in batch size * Code review and add test * Fix accidental typo in test_empirical_sinkhorn * Remove whitespace * Edit config.yml
2021-06-01[MRG] POT numpy/torch/jax backends (#249)Rémi Flamary
* add numpy and torch backends * stat sets on functions * proper import * install recent torch on windows * install recent torch on windows * now testing all functions in backedn * add jax backedn * clenaup windowds * proper convert for jax backedn * pep8 * try again windows tests * test jax conversion * try proper widows tests * emd fuction ses backedn * better test partial OT * proper tests to_numpy and teplate Backend * pep8 * pep8 x2 * feaking sinkhorn works with torch * sinkhorn2 compatible * working ot.emd2 * important detach * it should work * jax autodiff emd * pep8 * no tast same for jax * new independat tests per backedn * freaking pep8 * add tests for gradients * deprecate ot.gpu * worging dist function * working dist * dist done in backedn * not in * remove indexing * change accuacy for jax * first pull backend * projection simplex * projection simplex * projection simplex * projection simplex no ci * projection simplex no ci * projection simplex no ci * pep8 * add backedn discusion to quickstart guide * projection simplex no ci * projection simplex no ci * projection simplex no ci * pep8 + better doc * proper links * corect doctest * big debug documentation * doctest again * doctest again bis * doctest again ter (last one or i kill myself) * backend test + doc proj simplex * correction test_utils * correction test_utils * correction cumsum * correction flip * correction flip v2 * more debug * more debug * more debug + pep8 * pep8 * argh * proj_simplex * backedn works for sort * proj simplex * jax sucks * update doc * Update test/test_utils.py Co-authored-by: Alexandre Gramfort <alexandre.gramfort@m4x.org> * Update docs/source/quickstart.rst Co-authored-by: Alexandre Gramfort <alexandre.gramfort@m4x.org> * Update docs/source/quickstart.rst Co-authored-by: Alexandre Gramfort <alexandre.gramfort@m4x.org> * Update docs/source/quickstart.rst Co-authored-by: Alexandre Gramfort <alexandre.gramfort@m4x.org> * Update docs/source/readme.rst Co-authored-by: Alexandre Gramfort <alexandre.gramfort@m4x.org> * Update test/test_utils.py Co-authored-by: Alexandre Gramfort <alexandre.gramfort@m4x.org> * Update ot/utils.py Co-authored-by: Alexandre Gramfort <alexandre.gramfort@m4x.org> * Update docs/source/readme.rst Co-authored-by: Alexandre Gramfort <alexandre.gramfort@m4x.org> * Update ot/lp/__init__.py Co-authored-by: Alexandre Gramfort <alexandre.gramfort@m4x.org> * begin comment alex * comment alex part 2 * optimize test gromov * proj_simplex on vectors * add awesome gradient decsnt example on the weights * pep98 of course * proof read example by alex * pep8 again * encoding oos in translation * correct legend Co-authored-by: Nicolas Courty <ncourty@irisa.fr> Co-authored-by: Alexandre Gramfort <alexandre.gramfort@m4x.org>
2021-04-19BUG/DOC FIX - Sinkhorn divergence used the wrong weights, and sinkhorn2 ↵AdrienCorenflos
didn't support epsilon_scaling method. (#235) * FIX: 1. Documentation of loss specific functions 2. Sinkhorn divergence weights handling 3. Sinkhorn2 does not support epsilon scaling, so I removed it (it *should* arguably support it, but this would require a refactoring of the sinkhorn iterates pretty much everywhere, maybe should be done in torch first?) * Had some PEP8 issues Co-authored-by: Rémi Flamary <remi.flamary@gmail.com>
2020-12-22[WIP] Update documentation "Why OT" section (#220)Rémi Flamary
* add some text + discussion sinkhorn * stating wrk on why POT * fix sphinx warnings + make html-noplot * discussion when not to use POT * add discussion which sinkhorn * edits on quickstart * more * remove warnings :any: * more * done * remove ref Co-authored-by: Alexandre Gramfort <alexandre.gramfort@m4x.org>
2020-06-22Revert "[WIP] Update bregman.py : barycenter_sinkhorn function (#195)" (#196)Rémi Flamary
This reverts commit b2eae39e8a422b18ecc3fadc08bc909ee1dae55f.
2020-06-22[WIP] Update bregman.py : barycenter_sinkhorn function (#195)Rémi Flamary
* Update bregman.py * correct call to function Co-authored-by: FerdinandGns <56926826+FerdinandGns@users.noreply.github.com>
2020-04-24better doc per moduleRémi Flamary
2020-04-24better coverage options verbose and logRémi Flamary
2020-04-15fix soft labels, remove gammas from jcpotievred
2020-04-08added test barycenter + modif targetievred
2020-04-08remove laplace from jcpotievred
2020-04-07updievred
2020-04-07pep bregmanievred
2020-04-07pep bregmanievred
2020-04-02laplace v1ievred
2020-04-01fix test example add M to logievred
2020-03-31readme move to bregmanievred
2020-03-06Fix convolutional_barycenter kernel for non-symmetric imagesAlex Tong
Add authorship
2020-01-24add a warning for non installed Botteleneck moduleMokhtar Z. Alaya
2020-01-18cleanMokhtar Z. Alaya
2020-01-18cleanMokhtar Z. Alaya
2020-01-10improve docMokhtar Z. Alaya
2020-01-10improve docMokhtar Z. Alaya
2020-01-10update screenkhorn exampleMokhtar Z. Alaya
2020-01-10improve documentationMokhtar Z. Alaya
2020-01-08fix binary indexingMokhtar Z. Alaya
2020-01-08using binary indexing for definition the active setsMokhtar Z. Alaya
2020-01-07fix typos in documentationMokhtar Z. Alaya
2020-01-07fix definitions complementary active sets Ic, JcMokhtar Z. Alaya
2020-01-07replace reshape by numpy slicing in returnMokhtar Z. Alaya
2020-01-07fix default values for the budget argumentsMokhtar Z. Alaya
2020-01-07set default param. for LBFGS in the function's prototypeMokhtar Z. Alaya
2020-01-07replace @ operator by np.dotMokhtar Z. Alaya
2020-01-07delete "ArrayBox"-type test of dist. matrix MMokhtar Z. Alaya
generated by cgit v1.2.3 (git 2.39.1) at 2024-06-04 19:54:21 +0000