Hendrik/xgboost
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
5,287 Commits 32 Branches 59 Tags
Commit Graph

5287 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Jiaming Yuan
522b8977c2
Make 1.4.2 patch release. (#6962) v1.4.2 2021-05-13 16:17:14 +08:00
Jiaming Yuan
8147d78b12
[back port] Fix multiclass auc with empty dataset. (#6947) (#6960) 2021-05-13 12:31:52 +08:00
Jiaming Yuan
651c4ac03b
[CI] Fix CI/CD pipeline broken by latest auditwheel (4.0.0) (#6951) (#6952) 
Co-authored-by: Philip Hyunsu Cho <chohyu01@cs.washington.edu>
 2021-05-11 20:45:04 +08:00
Philip Hyunsu Cho
1fb75d2460
Make it easier to release GPU/CPU code artifacts to Maven Central (#6907) 2021-05-04 12:43:42 -07:00
Jiaming Yuan
6609211517
[backport] Enforce input data is not object. (#6927) (#6938) 
* Check for object data type.
* Allow strided arrays with greater underlying buffer size.
 2021-05-04 16:10:16 +08:00
Jiaming Yuan
b78ad1e623
[backport] Ensure predict leaf output 1-dim vector where there's only 1 tree. (#6889) (#6937) 2021-05-04 16:09:49 +08:00
Jiaming Yuan
96f8843694
[backport] CI fixes (#6933) 
* Relax shotgun test. (#6900)

It's non-deterministic algorithm, the test is flaky.

* Disable pylint error. (#6911)

* [CI] Skip external memory gtest on osx. (#6901)

* [CI] Fix custom metric test with empty dataset. (#6917)

* Reduce Travis environment setup time. (#6912)

* Remove unused r from travis.
* Don't update homebrew.
* Don't install indirect/unused dependencies like libgit2, wget, openssl.
* Move graphviz installation to conda.

* Relax shotgun test. (#6918)

* Relax test for decision stump in distributed environment. (#6919)

* Backport cupy fix.
 2021-05-03 21:30:26 +08:00
Jiaming Yuan
a6d1fbf8d1
Fix warning on Windows. (#6883) 2021-04-22 02:57:23 +08:00
Jiaming Yuan
a347ef7c66
Release 1.4.1 (#6874) v1.4.1 2021-04-20 08:31:38 +08:00
Jiaming Yuan
d05c47dcf8
[back port] Copy output data for argsort. (#6866) (#6870) 
Fix GPU AUC.
 2021-04-17 01:33:22 +08:00
Jiaming Yuan
9f5e2c52ce
[back port] Update dmlc-core. (#6862) (#6864) 
* Install pandoc, pandoc-citeproc on CI.
 2021-04-16 03:49:41 +08:00
Hyunsu Cho
c42387162d Upload Doxygen to correct destination 2021-04-13 15:07:29 -07:00
Hyunsu Cho
1220024442 Release 1.4.0 v1.4.0 2021-04-10 17:42:00 -07:00
Philip Hyunsu Cho
964ee6b605
[CI] Pack R package tarball with pre-built xgboost.so (with GPU support) (#6827) (#6836) 
* Add scripts for packaging R package with GPU-enabled libxgboost.so

* [CI] Automatically build R package tarball

* Add comments

* Don't build tarball for pull requests

* Update the installation doc
 2021-04-07 22:47:54 -07:00
Jiaming Yuan
04fedefd4d
[back port] Use batched copy if. (#6826) (#6834) 2021-04-07 04:50:52 +08:00
Jiaming Yuan
f814d4027a
[back port] Remove unnecessary calls to iota. (#6797) (#6833) 2021-04-07 04:47:29 +08:00
Jiaming Yuan
2cc37370e2
[back port] Fix approximated predict contribution. (#6811) (#6832) 2021-04-07 04:47:07 +08:00
Jiaming Yuan
c6a0bdbb5a
[back port] More general predict proba. (#6817) (#6831) 
* Use `output_margin` for `softmax`.
* Add test for dask binary cls.

Co-authored-by: Philip Hyunsu Cho <chohyu01@cs.washington.edu>
 2021-04-07 04:46:11 +08:00
Jiaming Yuan
357a78b3de
[back port] Optimize dart inplace predict perf. (#6804) (#6829) 2021-04-07 00:21:12 +08:00
Jiaming Yuan
d231e7c35f
[back port] Don't estimate sketch batch size when rmm is used. (#6807) (#6830) 2021-04-07 00:16:39 +08:00
Jiaming Yuan
604ae01b7a
[back port] Use CPU input for test_boost_from_prediction. (#6818) (#6824) 2021-04-05 18:32:04 +08:00
Hyunsu Cho
43f52ed33c Release 1.4.0 RC1 2021-03-28 01:10:20 +00:00
Jiaming Yuan
a59c7323b4
Fix inplace predict missing value. (#6787) 2021-03-27 05:36:10 +08:00
Jiaming Yuan
5c87c2bba8
Update demo for prediction. (#6789) 
* Remove use of deprecated ntree_limit.
* Add sklearn demo.
 2021-03-27 03:09:25 +08:00
ShvetsKS
8825670c9c
Memory consumption fix for row-major adapters (#6779) 
Co-authored-by: Kirill Shvets <kirill.shvets@intel.com>
Co-authored-by: fis <jm.yuan@outlook.com>
 2021-03-26 08:44:30 +08:00
Philip Hyunsu Cho
744c46995c
[CI] Upload xgboost4j.dll to S3 (#6781) 2021-03-25 11:34:34 -07:00
Jiaming Yuan
a7083d3c13
Fix dart inplace prediction with GPU input. (#6777) 
* Fix dart inplace predict with data on GPU, which might trigger a fatal check
for device access right.
* Avoid copying data whenever possible.
 2021-03-25 12:00:32 +08:00
Jiaming Yuan
1d90577800
Verify strictly positive labels for gamma regression. (#6778) 
Co-authored-by: Philip Hyunsu Cho <chohyu01@cs.washington.edu>
 2021-03-25 11:46:52 +08:00
Jiaming Yuan
794fd6a46b
Support v3 cuda array interface. (#6776) 2021-03-25 09:58:09 +08:00
Jiaming Yuan
bcc0277338
Re-implement ROC-AUC. (#6747) 
* Re-implement ROC-AUC.

* Binary
* MultiClass
* LTR
* Add documents.

This PR resolves a few issues:
  - Define a value when the dataset is invalid, which can happen if there's an
  empty dataset, or when the dataset contains only positive or negative values.
  - Define ROC-AUC for multi-class classification.
  - Define weighted average value for distributed setting.
  - A correct implementation for learning to rank task.  Previous
  implementation is just binary classification with averaging across groups,
  which doesn't measure ordered learning to rank.
 2021-03-20 16:52:40 +08:00
Jiaming Yuan
4ee8340e79
Support column major array. (#6765) 2021-03-20 05:19:46 +08:00
Jiaming Yuan
f6fe15d11f
Improve parameter validation (#6769) 
* Add quotes to unused parameters.
* Check for whitespace.
 2021-03-20 01:56:55 +08:00
Jiaming Yuan
23b4165a6b
Fix gamma deviance (#6761) 2021-03-20 01:56:17 +08:00
ReeceGoding
c2b6b80600
R documentation: Make construction of DMatrix consistent. 
* Fix inconsistency of construction of DMatrix.
* Fix missing parameters.
 2021-03-20 01:55:13 +08:00
Qingyun Wu
642336add7
[doc] Add FLAML as a fast tuning tool for XGBoost (#6770) 
Co-authored-by: Qingyun Wu <qiw@microsoft.com>
 2021-03-20 01:47:39 +08:00
Philip Hyunsu Cho
4230dcb614
Re-introduce double buffer in UpdatePosition, to fix perf regression in gpu_hist (#6757) 
* Revert "gpu_hist performance tweaks (#5707)"

This reverts commit f779980f7ea7f6f07e86229b8e78144e8a74e6b3.

* Address reviewer's comment

* Fix build error
 2021-03-18 13:56:10 -07:00
Jiaming Yuan
e2d8a99413
Add document for tests directory. [skip ci] (#6760) 2021-03-18 15:15:50 +08:00
ReeceGoding
4e00737c60
Fix R documentation for xgb.train. (#6764) 
The [general documentation](https://xgboost.readthedocs.io/en/latest/parameter.html#parameters-for-tree-booster) clearly has alpha and lambda under its "Parameters for Tree Booster" heading. Furthermore, the R package clearly uses alpha and lambda when told to use the tree booster. This update adds those two parameters to the documentation for the R package.


Closed issue #6763.
 2021-03-18 15:04:00 +08:00
Jiaming Yuan
4f75f514ce
Fix GPU RF (#6755) 
* Fix sampling.
 2021-03-17 06:23:35 +08:00
Jiaming Yuan
1a73a28511
Add device argsort. (#6749) 
This is part of https://github.com/dmlc/xgboost/pull/6747 .
 2021-03-16 16:05:22 +08:00
Jiaming Yuan
325bc93e16
[dask] Use distributed.MultiLock (#6743) 
* [dask] Use `distributed.MultiLock`

This enables training multiple models in parallel.

* Conditionally import `MultiLock`.
* Use async train directly in scikit learn interface.
* Use `worker_client` when available.
 2021-03-16 14:19:41 +08:00
Igor Rukhovich
19a2c54265
Prediction by indices (subsample < 1) (#6683) 
* Another implementation of predicting by indices

* Fixed omp parallel_for variable type

* Removed SparsePageView from Updater
 2021-03-16 15:08:20 +13:00
Philip Hyunsu Cho
366f3cb9d8
Add use_rmm flag to global configuration (#6656) 
* Ensure RMM is 0.18 or later

* Add use_rmm flag to global configuration

* Modify XGBCachingDeviceAllocatorImpl to skip CUB when use_rmm=True

* Update the demo

* [CI] Pin NumPy to 1.19.4, since NumPy 1.19.5 doesn't work with latest Shap
 2021-03-09 14:53:05 -08:00
Philip Hyunsu Cho
e4894111ba
Update dmlc-core submodule (#6745) 2021-03-07 00:30:26 -08:00
Bobby Wang
49c22c23b4
[jvm-packages] fix early stopping doesn't work even without custom_eval setting (#6738) 
* [jvm-packages] fix early stopping doesn't work even without custom_eval setting

* remove debug info

* resolve comment
 2021-03-06 20:19:40 -08:00
Philip Hyunsu Cho
5ae7f9944b
[CI] Clear R package cache (#6746) 2021-03-06 08:37:16 -08:00
Jiaming Yuan
f20074e826
Check for invalid data. (#6742) 2021-03-04 14:37:20 +08:00
Jiaming Yuan
a9b4a95225
Fix learning rate scheduler with cv. (#6720) 
* Expose more methods in cvpack and packed booster.
* Fix cv context in deprecated callbacks.
* Fix document.
 2021-02-28 13:57:42 +08:00
kangsheng89
9c8523432a
fix relocatable include in CMakeList (#6734) (#6737) 2021-02-27 19:17:29 +08:00
Roffild
1fa6793a4e
Tests for regression metrics with weights. (#6729) 2021-02-25 22:08:14 +08:00