|
1 | 1 | # POT Releases |
2 | 2 |
|
3 | 3 |
|
| 4 | +## 0.6 Year 3 |
| 5 | +*July 2019* |
| 6 | + |
| 7 | +This is the first official stable release of POT and this means a jump to 0.6! |
| 8 | +The library has been used in |
| 9 | +the wild for a while now and we have reached a state where a lot of fundamental |
| 10 | +OT solvers are available and tested. It has been quite stable in the last months |
| 11 | +but kept the beta flag in its Pypi classifiers until now. |
| 12 | + |
| 13 | +Note that this release will be the last one supporting officially Python 2.7 (See |
| 14 | +https://python3statement.org/ for more reasons). For next release we will keep |
| 15 | +the travis tests for Python 2 but will make them non necessary for merge in 2020. |
| 16 | + |
| 17 | +The features are never complete in a toolbox designed for solving mathematical |
| 18 | +problems and research but with the new contributions we now implement algorithms and solvers |
| 19 | +from 24 scientific papers (listed in the README.md file). New features include a |
| 20 | +direct implementation of the [empirical Sinkhorn divergence](https://pot.readthedocs.io/en/latest/all.html#ot.bregman.empirical_sinkhorn_divergence) |
| 21 | +, a new efficient (Cython implementation) solver for [EMD in 1D](https://pot.readthedocs.io/en/latest/all.html#ot.lp.emd_1d) |
| 22 | +and corresponding [Wasserstein |
| 23 | +1D](https://pot.readthedocs.io/en/latest/all.html#ot.lp.wasserstein_1d). We now also |
| 24 | +have implementations for [Unbalanced OT](https://github.com/rflamary/POT/blob/master/notebooks/plot_UOT_1D.ipynb) |
| 25 | +and a solver for [Unbalanced OT barycenters](https://github.com/rflamary/POT/blob/master/notebooks/plot_UOT_barycenter_1D.ipynb). |
| 26 | +A new variant of Gromov-Wasserstein divergence called [Fused |
| 27 | +Gromov-Wasserstein](https://pot.readthedocs.io/en/latest/all.html?highlight=fused_#ot.gromov.fused_gromov_wasserstein) |
| 28 | + has been also contributed with exemples of use on [structured data](https://github.com/rflamary/POT/blob/master/notebooks/plot_fgw.ipynb) |
| 29 | +and computing [barycenters of labeld graphs](https://github.com/rflamary/POT/blob/master/notebooks/plot_barycenter_fgw.ipynb). |
| 30 | + |
| 31 | + |
| 32 | +A lot of work has been done on the documentation with several new |
| 33 | +examples corresponding to the new features and a lot of corrections for the |
| 34 | +docstrings. But the most visible change is a new |
| 35 | +[quick start guide](https://pot.readthedocs.io/en/latest/quickstart.html) for |
| 36 | +POT that gives several pointers about which function or classes allow to solve which |
| 37 | +specific OT problem. When possible a link is provided to relevant examples. |
| 38 | + |
| 39 | +We will also provide with this release some pre-compiled Python wheels for Linux |
| 40 | +64bit on |
| 41 | +github and pip. This will simplify the install process that before required a C |
| 42 | +compiler and numpy/cython already installed. |
| 43 | + |
| 44 | +Finally we would like to acknowledge and thank the numerous contributors of POT |
| 45 | +that has helped in the past build the foundation and are still contributing to |
| 46 | +bring new features and solvers to the library. |
| 47 | + |
| 48 | + |
| 49 | + |
| 50 | +#### Features |
| 51 | + |
| 52 | +* Add compiled manylinux 64bits wheels to pip releases (PR #91) |
| 53 | +* Add quick start guide (PR #88) |
| 54 | +* Make doctest work on travis (PR #90) |
| 55 | +* Update documentation (PR #79, PR #84) |
| 56 | +* Solver for EMD in 1D (PR #89) |
| 57 | +* Solvers for regularized unbalanced OT (PR #87, PR#99) |
| 58 | +* Solver for Fused Gromov-Wasserstein (PR #86) |
| 59 | +* Add empirical Sinkhorn and empirical Sinkhorn divergences (PR #80) |
| 60 | + |
| 61 | + |
| 62 | +#### Closed issues |
| 63 | + |
| 64 | +- Issue #59 fail when using "pip install POT" (new details in doc+ hopefully |
| 65 | + wheels) |
| 66 | +- Issue #85 Cannot run gpu modules |
| 67 | +- Issue #75 Greenkhorn do not return log (solved in PR #76) |
| 68 | +- Issue #82 Gromov-Wasserstein fails when the cost matrices are slightly different |
| 69 | +- Issue #72 Macosx build problem |
| 70 | + |
| 71 | + |
4 | 72 | ## 0.5.0 Year 2 |
5 | 73 | *Sep 2018* |
6 | 74 |
|
|
0 commit comments