Hendrik/xgboost
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
6,211 Commits 32 Branches 59 Tags
Commit Graph

6211 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Philip Hyunsu Cho
37c75aac41
[CI] Add external Docker build cache (#4331) 2019-04-04 13:36:39 -07:00
Jiaming Yuan
82dca3c108
Don't store DMatrix handle until it's initialized. (#4317) 
* Use a temporary variable to store the handle.
* Decode c++ error message.
* Simple note about saved binary.
 2019-04-01 18:29:28 +08:00
sriramch
2f7087eba1 Improve HostDeviceVector exception safety (#4301) 
* make the assignments of HostDeviceVector exception safe.
* storing a dummy GPUDistribution instance in HDV for CPU based code.
* change testxgboost binary location to build directory.
 2019-03-31 22:48:58 +08:00
Hajime Morrita
680a1b36f3 Get rid of a few trivial compiler warnings. (#4312) 2019-03-31 00:02:29 +08:00
Nan Zhu
ad4de0d718
[jvm-packages] handle NaN as missing value explicitly (#4309) 
* handle nan

* handle nan explicitly

* make code better and handle sparse vector in spark

* Update XGBoostGeneralSuite.scala
 2019-03-30 19:34:26 +08:00
Rong Ou
7ea5b772fb do not filter shared library files (#4303) 2019-03-28 19:40:54 +08:00
Philip Hyunsu Cho
7aed8f3d48
[CI] Upgrade to GCC 5.3.1, CMake 3.6.0 (#4306) 
* Upgrade to GCC 5.3.1, CMake 3.6.0

* <regex> is now okay
 2019-03-28 00:21:21 -07:00
Rong Ou
8c8021dfa7 use all cores to build on linux (#4304) 2019-03-27 19:51:08 -07:00
Rory Mitchell
3f312e30db
Retire DVec class in favour of c++20 style span for device memory. (#4293) 2019-03-28 13:59:58 +13:00
Jiaming Yuan
c85181dd8a
Remove remaining silent and debug_verbose. (#4299) 2019-03-28 03:30:46 +08:00
Rory Mitchell
6d5b34d824
Further optimisations for gpu_hist. (#4283) 
- Fuse final update position functions into a single more efficient kernel

- Refactor gpu_hist with a more explicit ellpack  matrix representation
 2019-03-24 17:17:22 +13:00
Rong Ou
5aa42b5f11 jenkins build for cuda 10.0 (#4281) 
* jenkins build for cuda 10.0

* yum install nccl2 for cuda 10.0
 2019-03-22 22:35:18 -07:00
Chen Qin
ecd4bf7aae [rabit harden] replace hardcopy dmlc-core headers with submodule links (#86) 
* backport dmlc header changes to rabit

* use gitmodule to reference latest dmlc header files

* include ref to dmlc-core
fix cmake

* update cmake file, add cmake build traivs task

* try force using g++-4.8

* per feedback, update cmake
 2019-03-23 13:11:29 +08:00
Philip Hyunsu Cho
263e2038e9
Bump Python version number (#4285) 2019-03-21 14:40:44 -07:00
Harry Braviner
b374e0a7ab [jvm-packages] Allow supression of Rabit output in Booster::train in xgboost4j (#4262) 
* Make train in xgboost4j respect print params

Previously no setting in params argument of Booster::train would prevent
the Rabit.trackerPrint call. This can fill up a lot of screen space in
the case that many folds are being trained.
* Setting "silent" in this map to "true", "True", a non-zero integer, or
  a string that can be parsed to such an int will prevent printing.
* Setting "verbose_eval" to "False" or "false" will prevent printing.
* Setting "verbose_eval" to an int (or a String parseable to an int) n
  will result in printing every n steps, or no printing is n is zero.

This is to match the python behaviour described here:
https://www.kaggle.com/c/rossmann-store-sales/discussion/17499

* Fixed 'slient' typo in xgboost4j test

* private access on two methods
 2019-03-21 18:25:12 +08:00
Nan Zhu
45c89a6792
[jvm-packages] logging version number (#4271) 
* print version number

* add property file
 2019-03-21 18:24:29 +08:00
Rory Mitchell
8eab966998
Allow unique prediction vector for each input matrix (#4275) 2019-03-21 11:38:16 +13:00
Jiaming Yuan
09bd9e68cf
Use Monitor in quantile hist. (#4273) 2019-03-20 09:26:22 +08:00
Rory Mitchell
00465d243d
Optimisations for gpu_hist. (#4248) 
* Optimisations for gpu_hist.

* Use streams to overlap operations.

* ColumnSampler now uses HostDeviceVector to prevent repeatedly copying feature vectors to the device.
 2019-03-20 13:30:06 +13:00
Rory Mitchell
7814183199
Fix travis R tests (#4277) 2019-03-20 12:56:04 +13:00
Nan Zhu
359ed9c5bc
[jvm-packages] add configuration flag to control whether to cache transformed training set (#4268) 
* control whether to cache data

* uncache
 2019-03-18 10:13:28 +08:00
Jiaming Yuan
29a1356669
Deprecate reg:linear' in favor of reg:squarederror'. (#4267) 
* Deprecate `reg:linear' in favor of `reg:squarederror'.
* Replace the use of `reg:linear'.
* Replace the use of `silent`.
 2019-03-17 17:55:04 +08:00
Jiaming Yuan
cf8d5b9b76
Mark CUDA 10.1 as unsupported. (#4265) 2019-03-17 16:59:15 +08:00
Jiaming Yuan
fdcae024e7
Remove deprecated C APIs. (#4266) 2019-03-17 16:42:44 +08:00
Jiaming Yuan
7b1b11390a
Mark Scikit-Learn RF interface as experimental in doc. (#4258) 
* Mark Scikit-Learn RF interface as experimental in doc.
 2019-03-16 00:45:32 +08:00
Chen Qin
785d7e54d3 [mpi] add engine_mpi travis build (#83) 2019-03-15 22:58:47 +08:00
Rory Mitchell
5465b73e7c
Fix multi-GPU test failures (#4259) 2019-03-15 14:40:43 +13:00
Chen Qin
ed06e0c6af [rabit harden] fix rabit tests (#81) 
* enable model recovery tests
* force use gcc4.8 in Travis
 2019-03-15 07:16:45 +08:00
Andy Adinets
4352fcdb15 Brought the silent parameter for the SKLearn-like API back, marked it deprecated. (#4255) 
* Brought the silent parameter for the SKLearn-like API back, marked it deprecated.

- added deprecation notice and warning
- removed silent from the tests for the SKLearn-like API
 2019-03-14 09:45:08 +13:00
Andy Adinets
b833b642ec Improved multi-node multi-GPU random forests. (#4238) 
* Improved multi-node multi-GPU random forests.

- removed rabit::Broadcast() from each invocation of column sampling
- instead, syncing the PRNG seed when a ColumnSampler() object is constructed
- this makes non-trivial column sampling significantly faster in the distributed case
- refactored distributed GPU tests
- added distributed random forests tests
 2019-03-13 12:36:28 +13:00
Philip Hyunsu Cho
99a714be64 Simplify README page (#4254) 2019-03-12 11:58:08 -07:00
Jiaming Yuan
7b9043cf71
Fix clang-tidy warnings. (#4149) 
* Upgrade gtest for clang-tidy.
* Use CMake to install GTest instead of mv.
* Don't enforce clang-tidy to return 0 due to errors in thrust.
* Add a small test for tidy itself.

* Reformat.
 2019-03-13 02:25:51 +08:00
Tong He
259fb809e9
fix R-devel errors (#4251) 2019-03-12 10:06:54 -07:00
Andy Adinets
a36c3ed4f4 Added SKLearn-like random forest Python API. (#4148) 
* Added SKLearn-like random forest Python API.

- added XGBRFClassifier and XGBRFRegressor classes to SKL-like xgboost API
- also added n_gpus and gpu_id parameters to SKL classes
- added documentation describing how to use xgboost for random forests,
  as well as existing caveats
 2019-03-12 22:28:19 +08:00
jess
6fb4c5efef Activating Open Collective (#4244) 
* Added backers and sponsors on the README

* Re-arrange sections

* Resize AWS logo
 2019-03-11 15:36:29 -07:00
Rory Mitchell
4eeeded7d1
Remove various synchronisations from cuda API calls, instrument monitor (#4205) 
* Remove various synchronisations from cuda API calls, instrument monitor
with nvtx profiler ranges.
 2019-03-10 15:01:23 +13:00
Philip Hyunsu Cho
f83e62dca5
Address #4042: Prevent out-of-range access in column matrix (#4231) 2019-03-08 17:11:42 -08:00
Philip Hyunsu Cho
331cd3e4f7
Document limitation of one-split-at-a-time Greedy tree learning heuristic (#4233) 2019-03-08 10:05:39 -08:00
Jiaming Yuan
617f572c0f
Update R contribute link. (#4236) 2019-03-09 01:50:07 +08:00
Philip Hyunsu Cho
20845e8ccf
Broken link for NCCL: cannot use CUDA 10.1 (#4232) 2019-03-08 09:10:29 -08:00
Shaochen Shi
224786f67f [xgboost4j-spark] Allow set the parameter "maxLeaves". (#4226) 
* Allow set the parameter "maxLeaves".

* Add "setMaxLeaves" to XGBoostRegressor.
 2019-03-07 18:36:47 -08:00
Rong Ou
9837b09b20 support cuda 10.1 (#4223) 
* support cuda 10.1

* add cuda 10.1 to jenkins build matrix
 2019-03-08 12:22:12 +13:00
Rong Ou
0944360416 minor fix: log InitDataOnce() only when it is actually called (#4206) 2019-03-08 10:53:09 +13:00
Christopher Suchanek
ac3d03089b [jvm-packages] remove shutdown of handler shutdown (#4224) 2019-03-06 19:32:43 -08:00
Philip Hyunsu Cho
28bd6cde22 Add sponsors (#4222) 2019-03-06 13:11:06 -08:00
Jonas
00ea7b83c9 Fix docs for num_parallel_tree (#4221) 
Minor formatting correction for `num_parallel_tree`.
 2019-03-06 23:47:51 +08:00
Philip Hyunsu Cho
67c38805a1
Update build doc: PyPI wheel now support multi-GPU (#4219) 2019-03-05 13:25:31 -08:00
Nan Zhu
5f34078fba
[jvm-packages] bump version for master (#4209) 
* update version

* bump version
 2019-03-04 23:12:24 -08:00
Philip Hyunsu Cho
3f83dcd502
Release 0.82 (#4201) v0.82 2019-03-04 18:14:36 -08:00
Adam November
0c1d5f1120 Fix snapshot artifact name in docs. (#4196) 2019-03-03 13:27:50 -08:00