Skip to content

Commit

Permalink
update changelog
Browse files Browse the repository at this point in the history
  • Loading branch information
@raydouglass
raydouglass committed Apr 21, 2021
1 parent b4ed0ad commit 46174b7
Showing 1 changed file with 120 additions and 3 deletions.
123 changes: 120 additions & 3 deletions CHANGELOG.md
View file
Original file line number Diff line number Diff line change
@@ -1,6 +1,123 @@
# cuML 0.19.0 (Date TBD)

Please see https://github.com/rapidsai/cuml/releases/tag/v0.19.0a for the latest changes to this development branch.
# cuML 0.19.0 (21 Apr 2021)

## 🚨 Breaking Changes

- Use the new RF backend by default for classification ([#3686](https://github.com//rapidsai/cuml/pull/3686)) [@hcho3](https://github.com/hcho3)
- Deprecating quantile-per-tree and removing three previously deprecated Random Forest parameters ([#3667](https://github.com//rapidsai/cuml/pull/3667)) [@vinaydes](https://github.com/vinaydes)
- Update predict() / predict_proba() of RF to match sklearn ([#3609](https://github.com//rapidsai/cuml/pull/3609)) [@hcho3](https://github.com/hcho3)
- Upgrade FAISS to 1.7.x ([#3509](https://github.com//rapidsai/cuml/pull/3509)) [@viclafargue](https://github.com/viclafargue)
- cuML's estimator Base class for preprocessing models ([#3270](https://github.com//rapidsai/cuml/pull/3270)) [@viclafargue](https://github.com/viclafargue)

## 🐛 Bug Fixes

- Fix brute force KNN distance metric issue ([#3755](https://github.com//rapidsai/cuml/pull/3755)) [@viclafargue](https://github.com/viclafargue)
- Fix min_max_axis ([#3735](https://github.com//rapidsai/cuml/pull/3735)) [@viclafargue](https://github.com/viclafargue)
- Fix NaN errors observed with ARIMA in CUDA 11.2 builds ([#3730](https://github.com//rapidsai/cuml/pull/3730)) [@Nyrio](https://github.com/Nyrio)
- Fix random state generator ([#3716](https://github.com//rapidsai/cuml/pull/3716)) [@viclafargue](https://github.com/viclafargue)
- Fixes the out of memory access issue for computeSplit kernels ([#3715](https://github.com//rapidsai/cuml/pull/3715)) [@vinaydes](https://github.com/vinaydes)
- Fixing umap gtest failure under cuda 11.2. ([#3696](https://github.com//rapidsai/cuml/pull/3696)) [@cjnolet](https://github.com/cjnolet)
- Fix irreproducibility issue in RF classification ([#3693](https://github.com//rapidsai/cuml/pull/3693)) [@vinaydes](https://github.com/vinaydes)
- BUG fix BatchedLevelAlgo DtClsTest & DtRegTest failing tests ([#3690](https://github.com//rapidsai/cuml/pull/3690)) [@venkywonka](https://github.com/venkywonka)
- Restore the functionality of RF score() ([#3685](https://github.com//rapidsai/cuml/pull/3685)) [@hcho3](https://github.com/hcho3)
- Use main build.sh to build docs in docs CI ([#3681](https://github.com//rapidsai/cuml/pull/3681)) [@dantegd](https://github.com/dantegd)
- Revert "Update conda recipes pinning of repo dependencies" ([#3680](https://github.com//rapidsai/cuml/pull/3680)) [@raydouglass](https://github.com/raydouglass)
- Skip tests that fail on CUDA 11.2 ([#3679](https://github.com//rapidsai/cuml/pull/3679)) [@dantegd](https://github.com/dantegd)
- Dask KNN Cl&Re 1D labels ([#3668](https://github.com//rapidsai/cuml/pull/3668)) [@viclafargue](https://github.com/viclafargue)
- Update conda recipes pinning of repo dependencies ([#3666](https://github.com//rapidsai/cuml/pull/3666)) [@mike-wendt](https://github.com/mike-wendt)
- OOB access in GLM SoftMax ([#3642](https://github.com//rapidsai/cuml/pull/3642)) [@divyegala](https://github.com/divyegala)
- SilhouetteScore C++ tests seed ([#3640](https://github.com//rapidsai/cuml/pull/3640)) [@divyegala](https://github.com/divyegala)
- SimpleImputer fix ([#3624](https://github.com//rapidsai/cuml/pull/3624)) [@viclafargue](https://github.com/viclafargue)
- Silhouette Score `make_monotonic` for non-monotonic label set ([#3619](https://github.com//rapidsai/cuml/pull/3619)) [@divyegala](https://github.com/divyegala)
- Fixing support for empty rows in sparse Jaccard / Cosine ([#3612](https://github.com//rapidsai/cuml/pull/3612)) [@cjnolet](https://github.com/cjnolet)
- Fix train_test_split with stratify option ([#3611](https://github.com//rapidsai/cuml/pull/3611)) [@Nanthini10](https://github.com/Nanthini10)
- Update predict() / predict_proba() of RF to match sklearn ([#3609](https://github.com//rapidsai/cuml/pull/3609)) [@hcho3](https://github.com/hcho3)
- Change dask and distributed branch to main ([#3593](https://github.com//rapidsai/cuml/pull/3593)) [@dantegd](https://github.com/dantegd)
- Fixes memory allocation for experimental backend and improves quantile computations ([#3586](https://github.com//rapidsai/cuml/pull/3586)) [@vinaydes](https://github.com/vinaydes)
- Add ucx-proc package back that got lost during an auto merge conflict ([#3550](https://github.com//rapidsai/cuml/pull/3550)) [@dantegd](https://github.com/dantegd)
- Fix failing Hellinger gtest ([#3549](https://github.com//rapidsai/cuml/pull/3549)) [@cjnolet](https://github.com/cjnolet)
- Directly invoke make for non-CMake docs target ([#3534](https://github.com//rapidsai/cuml/pull/3534)) [@wphicks](https://github.com/wphicks)
- Fix Codecov.io Coverage Upload for Branch Builds ([#3524](https://github.com//rapidsai/cuml/pull/3524)) [@mdemoret-nv](https://github.com/mdemoret-nv)
- Ensure global_output_type is thread-safe ([#3497](https://github.com//rapidsai/cuml/pull/3497)) [@wphicks](https://github.com/wphicks)
- List as input for SimpleImputer ([#3489](https://github.com//rapidsai/cuml/pull/3489)) [@viclafargue](https://github.com/viclafargue)

## 📖 Documentation

- Add sparse docstring comments ([#3712](https://github.com//rapidsai/cuml/pull/3712)) [@JohnZed](https://github.com/JohnZed)
- FIL and Dask demo ([#3698](https://github.com//rapidsai/cuml/pull/3698)) [@miroenev](https://github.com/miroenev)
- Deprecating quantile-per-tree and removing three previously deprecated Random Forest parameters ([#3667](https://github.com//rapidsai/cuml/pull/3667)) [@vinaydes](https://github.com/vinaydes)
- Fixing Indentation for Docstring Generators ([#3650](https://github.com//rapidsai/cuml/pull/3650)) [@mdemoret-nv](https://github.com/mdemoret-nv)
- Update doc to indicate ExtraTree support ([#3635](https://github.com//rapidsai/cuml/pull/3635)) [@hcho3](https://github.com/hcho3)
- Update doc, now that FIL supports multi-class classification ([#3634](https://github.com//rapidsai/cuml/pull/3634)) [@hcho3](https://github.com/hcho3)
- Document model_type='xgboost_json' in FIL ([#3633](https://github.com//rapidsai/cuml/pull/3633)) [@hcho3](https://github.com/hcho3)
- Including log loss metric to the documentation website ([#3617](https://github.com//rapidsai/cuml/pull/3617)) [@lowener](https://github.com/lowener)
- Update the build doc regarding the use of GCC 7.5 ([#3605](https://github.com//rapidsai/cuml/pull/3605)) [@hcho3](https://github.com/hcho3)
- Update One-Hot Encoder doc ([#3600](https://github.com//rapidsai/cuml/pull/3600)) [@lowener](https://github.com/lowener)
- Fix documentation of KMeans ([#3595](https://github.com//rapidsai/cuml/pull/3595)) [@lowener](https://github.com/lowener)

## 🚀 New Features

- Reduce the size of the cuml libraries ([#3702](https://github.com//rapidsai/cuml/pull/3702)) [@robertmaynard](https://github.com/robertmaynard)
- Use ninja as default CMake generator ([#3664](https://github.com//rapidsai/cuml/pull/3664)) [@wphicks](https://github.com/wphicks)
- Single-Linkage Hierarchical Clustering Python Wrapper ([#3631](https://github.com//rapidsai/cuml/pull/3631)) [@cjnolet](https://github.com/cjnolet)
- Support for precomputed distance matrix in DBSCAN ([#3585](https://github.com//rapidsai/cuml/pull/3585)) [@Nyrio](https://github.com/Nyrio)
- Adding haversine to brute force knn ([#3579](https://github.com//rapidsai/cuml/pull/3579)) [@cjnolet](https://github.com/cjnolet)
- Support for sample_weight parameter in LogisticRegression ([#3572](https://github.com//rapidsai/cuml/pull/3572)) [@viclafargue](https://github.com/viclafargue)
- Provide "--ccache" flag for build.sh ([#3566](https://github.com//rapidsai/cuml/pull/3566)) [@wphicks](https://github.com/wphicks)
- Eliminate unnecessary includes discovered by cppclean ([#3564](https://github.com//rapidsai/cuml/pull/3564)) [@wphicks](https://github.com/wphicks)
- Single-linkage Hierarchical Clustering C++ ([#3545](https://github.com//rapidsai/cuml/pull/3545)) [@cjnolet](https://github.com/cjnolet)
- Expose sparse distances via semiring to Python API ([#3516](https://github.com//rapidsai/cuml/pull/3516)) [@lowener](https://github.com/lowener)
- Use cmake --build in build.sh to facilitate switching build tools ([#3487](https://github.com//rapidsai/cuml/pull/3487)) [@wphicks](https://github.com/wphicks)
- Add cython hinge_loss ([#3409](https://github.com//rapidsai/cuml/pull/3409)) [@Nanthini10](https://github.com/Nanthini10)
- Adding CodeCov Info for Dask Tests ([#3338](https://github.com//rapidsai/cuml/pull/3338)) [@mdemoret-nv](https://github.com/mdemoret-nv)
- Add predict_proba() to XGBoost-style models in FIL C++ ([#2894](https://github.com//rapidsai/cuml/pull/2894)) [@levsnv](https://github.com/levsnv)

## 🛠️ Improvements

- Updating docs, readme, and umap param tests for 0.19 ([#3731](https://github.com//rapidsai/cuml/pull/3731)) [@cjnolet](https://github.com/cjnolet)
- Locking RAFT hash for 0.19 ([#3721](https://github.com//rapidsai/cuml/pull/3721)) [@cjnolet](https://github.com/cjnolet)
- Upgrade to Treelite 1.1.0 ([#3708](https://github.com//rapidsai/cuml/pull/3708)) [@hcho3](https://github.com/hcho3)
- Update to XGBoost 1.4.0rc1 ([#3699](https://github.com//rapidsai/cuml/pull/3699)) [@hcho3](https://github.com/hcho3)
- Use the new RF backend by default for classification ([#3686](https://github.com//rapidsai/cuml/pull/3686)) [@hcho3](https://github.com/hcho3)
- Update LogisticRegression documentation ([#3677](https://github.com//rapidsai/cuml/pull/3677)) [@viclafargue](https://github.com/viclafargue)
- Preprocessing out of experimental ([#3676](https://github.com//rapidsai/cuml/pull/3676)) [@viclafargue](https://github.com/viclafargue)
- ENH Decision Tree new backend `computeSplit*Kernel` histogram calculation optimization ([#3674](https://github.com//rapidsai/cuml/pull/3674)) [@venkywonka](https://github.com/venkywonka)
- Remove `check_cupy8` ([#3669](https://github.com//rapidsai/cuml/pull/3669)) [@viclafargue](https://github.com/viclafargue)
- Use custom conda build directory for ccache integration ([#3658](https://github.com//rapidsai/cuml/pull/3658)) [@dillon-cullinan](https://github.com/dillon-cullinan)
- Disable three flaky tests ([#3657](https://github.com//rapidsai/cuml/pull/3657)) [@hcho3](https://github.com/hcho3)
- CUDA 11.2 developer environment ([#3648](https://github.com//rapidsai/cuml/pull/3648)) [@dantegd](https://github.com/dantegd)
- Store data frequencies in tree nodes of RF ([#3647](https://github.com//rapidsai/cuml/pull/3647)) [@hcho3](https://github.com/hcho3)
- Row major Gram matrices ([#3639](https://github.com//rapidsai/cuml/pull/3639)) [@tfeher](https://github.com/tfeher)
- Converting all Estimator Constructors to Keyword Arguments ([#3636](https://github.com//rapidsai/cuml/pull/3636)) [@mdemoret-nv](https://github.com/mdemoret-nv)
- Adding make_pipeline + test score with pipeline ([#3632](https://github.com//rapidsai/cuml/pull/3632)) [@viclafargue](https://github.com/viclafargue)
- ENH Decision Tree new backend `computeSplitClassificationKernel` histogram calculation and occupancy optimization ([#3616](https://github.com//rapidsai/cuml/pull/3616)) [@venkywonka](https://github.com/venkywonka)
- Revert "ENH Fix stale GHA and prevent duplicates " ([#3614](https://github.com//rapidsai/cuml/pull/3614)) [@mike-wendt](https://github.com/mike-wendt)
- ENH Fix stale GHA and prevent duplicates ([#3613](https://github.com//rapidsai/cuml/pull/3613)) [@mike-wendt](https://github.com/mike-wendt)
- KNN from RAFT ([#3603](https://github.com//rapidsai/cuml/pull/3603)) [@viclafargue](https://github.com/viclafargue)
- Update Changelog Link ([#3601](https://github.com//rapidsai/cuml/pull/3601)) [@ajschmidt8](https://github.com/ajschmidt8)
- Move SHAP explainers out of experimental ([#3596](https://github.com//rapidsai/cuml/pull/3596)) [@dantegd](https://github.com/dantegd)
- Fixing compatibility issue with CUDA array interface ([#3594](https://github.com//rapidsai/cuml/pull/3594)) [@lowener](https://github.com/lowener)
- Remove cutlass usage in row major input for euclidean exp/unexp, cosine and L1 distance matrix ([#3589](https://github.com//rapidsai/cuml/pull/3589)) [@mdoijade](https://github.com/mdoijade)
- Test FIL probabilities with absolute error thresholds in python ([#3582](https://github.com//rapidsai/cuml/pull/3582)) [@levsnv](https://github.com/levsnv)
- Removing sparse prims and fused l2 nn prim from cuml ([#3578](https://github.com//rapidsai/cuml/pull/3578)) [@cjnolet](https://github.com/cjnolet)
- Prepare Changelog for Automation ([#3570](https://github.com//rapidsai/cuml/pull/3570)) [@ajschmidt8](https://github.com/ajschmidt8)
- Print debug message if SVM convergence is poor ([#3562](https://github.com//rapidsai/cuml/pull/3562)) [@tfeher](https://github.com/tfeher)
- Fix merge conflicts in 3552 ([#3557](https://github.com//rapidsai/cuml/pull/3557)) [@ajschmidt8](https://github.com/ajschmidt8)
- Additional distance metrics for ANN ([#3533](https://github.com//rapidsai/cuml/pull/3533)) [@viclafargue](https://github.com/viclafargue)
- Improve warning message when QN solver reaches max_iter ([#3515](https://github.com//rapidsai/cuml/pull/3515)) [@tfeher](https://github.com/tfeher)
- Fix merge conflicts in 3502 ([#3513](https://github.com//rapidsai/cuml/pull/3513)) [@ajschmidt8](https://github.com/ajschmidt8)
- Upgrade FAISS to 1.7.x ([#3509](https://github.com//rapidsai/cuml/pull/3509)) [@viclafargue](https://github.com/viclafargue)
- ENH Pass ccache variables to conda recipe & use Ninja in CI ([#3508](https://github.com//rapidsai/cuml/pull/3508)) [@Ethyling](https://github.com/Ethyling)
- Fix forward-merger conflicts in #3502 ([#3506](https://github.com//rapidsai/cuml/pull/3506)) [@dantegd](https://github.com/dantegd)
- Sklearn meta-estimators into namespace ([#3493](https://github.com//rapidsai/cuml/pull/3493)) [@viclafargue](https://github.com/viclafargue)
- Add flexibility to copyright checker ([#3466](https://github.com//rapidsai/cuml/pull/3466)) [@lowener](https://github.com/lowener)
- Update sparse KNN to use rmm device buffer ([#3460](https://github.com//rapidsai/cuml/pull/3460)) [@lowener](https://github.com/lowener)
- Fix forward-merger conflicts in #3444 ([#3455](https://github.com//rapidsai/cuml/pull/3455)) [@ajschmidt8](https://github.com/ajschmidt8)
- Replace ML::MetricType with raft::distance::DistanceType ([#3389](https://github.com//rapidsai/cuml/pull/3389)) [@lowener](https://github.com/lowener)
- RF param initialization cython and C++ layer cleanup ([#3358](https://github.com//rapidsai/cuml/pull/3358)) [@venkywonka](https://github.com/venkywonka)
- MNMG RF broadcast feature ([#3349](https://github.com//rapidsai/cuml/pull/3349)) [@viclafargue](https://github.com/viclafargue)
- cuML's estimator Base class for preprocessing models ([#3270](https://github.com//rapidsai/cuml/pull/3270)) [@viclafargue](https://github.com/viclafargue)
- Make `_get_tags` a class/static method ([#3257](https://github.com//rapidsai/cuml/pull/3257)) [@dantegd](https://github.com/dantegd)
- NVTX Markers for RF and RF-backend ([#3014](https://github.com//rapidsai/cuml/pull/3014)) [@venkywonka](https://github.com/venkywonka)

# cuML 0.18.0 (24 Feb 2021)

Expand Down

0 comments on commit 46174b7

Please sign in to comment.