Hendrik/xgboost
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
3,475 Commits 32 Branches 59 Tags
Commit Graph

3475 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Juzer Shakir
1db28b8718 Typo fixed (#3784) 
The word 'make' was been repeated twice, fixed to single.
 2018-10-10 10:23:27 -07:00
KOLANICH
5480e05173 Added some instructions on using MinGW-built XGBoost with python. (#3774) 
* Added some instructions on using MinGW-built XGBoost with python.

* Changes according to the discussion and some additions

* Fixed wording and removed redundancy.

* Even more fixes

* Fixed links. Removed redundancy.

* Some fixes according to the discussion

* fixes

* Some fixes

* fixes
 2018-10-09 09:07:00 -07:00
weitian
9504f411c1 [jvm-packages] For training data with group, empty RDD partition threw exception (#3749) (#3750) 2018-10-09 09:03:22 -07:00
Philip Hyunsu Cho
ca33bf6476
Document gblinear parameters: feature_selector and top_k (#3780) 2018-10-08 22:41:54 -07:00
Philip Hyunsu Cho
133b8d94df
Fix Jenkins syntax (#3777) 2018-10-08 14:56:42 -07:00
Philip Hyunsu Cho
11eaf3eed1
Retry Jenkins CI tests up to 3 times to improve reliability (redux) (#3776) 2018-10-08 11:39:00 -07:00
Philip Hyunsu Cho
6d42e56c85
Retry Jenkins CI tests up to 3 times to improve reliability (redux) (#3775) 2018-10-08 11:24:01 -07:00
Philip Hyunsu Cho
7a7269e983
Retry Jenkins CI tests up to 3 times to improve reliability (#3769) 2018-10-08 09:55:39 -07:00
Philip Hyunsu Cho
ea99b53d8e
Document behavior of get_fscore() for zero-importance features (#3763) 2018-10-08 01:52:25 -07:00
Philip Hyunsu Cho
10cd7c8447
Fix #3714: preserve feature names when slicing DMatrix (#3766) 
* Fix #3714: preserve feature names when slicing DMatrix

* Add test
 2018-10-08 01:04:33 -07:00
Philip Hyunsu Cho
813d2436d3
Produce xgboost.so for XGBoost-R on Mac OSX, so that make install works (#3767) 
* Produce xgboost.so for XGBoost-R on Mac OSX, so that `make install` works

* Modernize R build instructions

* Fix crossref
 2018-10-07 14:09:54 -07:00
Philip Hyunsu Cho
c23783a0d1
Add notes to doc (#3765) 2018-10-07 14:09:09 -07:00
Philip Hyunsu Cho
91903ac5d4
Fix broken doc build due to Matplotlib 3.0 release (#3764) 2018-10-07 13:34:37 -07:00
Philip Hyunsu Cho
ae7e58b96e
Test wheel compatibility on CPU containers, for all pull requests (#3762) 
* Test wheel compatibility on CPU containers, for all pull requests

* Run wheel test only when multi-GPU flag is not set
 2018-10-06 20:18:58 -07:00
Saumya Bhatnagar
e0fd60f4e5 [doc] Fix link in rank demo README.md . (#3759) 2018-10-06 12:12:54 -07:00
trivialfis
4b892c2b30 Remove obsoleted QuantileHistMaker. (#3761) 
Fix #3755.
 2018-10-06 11:22:15 -07:00
Nan Zhu
785094db53
[jvm-packages] fix issue when spark job execution thread cannot return before we execute first() (#3758) 
* add back train method but mark as deprecated

* add back train method but mark as deprecated

* add back train method but mark as deprecated

* add back train method but mark as deprecated

* fix scalastyle error

* fix scalastyle error

* fix scalastyle error

* fix scalastyle error

* sparjJobThread

* update

* fix issue when spark job execution thread cannot return before we execute first()
 2018-10-05 22:20:50 -07:00
zengxy
9e73087324 [jvm-packages] support specified feature names when getModelDump and getFeatureScore (#3733) 
* [jvm-packages] support specified feature names for jvm when get ModelDump and get FeatureScore (#3725)

* typo and style fix
 2018-10-04 09:05:42 -07:00
Rory Mitchell
34522d56f0
Allow plug-ins to be built by cmake (#3752) 
* Remove references to AVX code.

* Allow plugins to be built by cmake
 2018-10-04 22:03:52 +13:00
trivialfis
c6b5df67f6 Catch dmlc::Error. (#3751) 
Fix #3643.
 2018-10-04 16:51:38 +13:00
weitian
efc4f85505 [jvm-packages] Fix #3489: Spark repartitionForData can potentially shuffle all data and lose ordering required for ranking objectives (#3654) 2018-10-03 08:43:55 -07:00
trivialfis
d594b11f35 Implement transform to reduce CPU/GPU code duplication. (#3643) 
* Implement Transform class.
* Add tests for softmax.
* Use Transform in regression, softmax and hinge objectives, except for Cox.
* Mark old gpu objective functions deprecated.
* static_assert for softmax.
* Split up multi-gpu tests.
 2018-10-02 15:06:21 +13:00
Sergei Lebedev
87aca8c244 [jvm-packages] Fixed the distributed updater check (#3739) 
The updater used in distributed training is grow_histmaker and not 
grow_colmaker as the error message stated prior to this commit.
 2018-10-01 11:22:01 -07:00
Rory Mitchell
70d208d68c
Dmatrix refactor stage 2 (#3395) 
* DMatrix refactor 2

* Remove buffered rowset usage where possible

* Transition to c++11 style iterators for row access

* Transition column iterators to C++ 11
 2018-10-01 01:29:03 +13:00
Philip Hyunsu Cho
b50bc2c1d4
Add multi-GPU unit test environment (#3741) 
* Add multi-GPU unit test environment

* Better assertion message

* Temporarily disable failing test

* Distinguish between multi-GPU and single-GPU CPP tests

* Consolidate Python tests. Use attributes to distinguish multi-GPU Python tests from single-CPU counterparts
 2018-09-29 11:20:58 -07:00
Philip Hyunsu Cho
baef5741df
Separate out restricted and unrestricted tasks (#3736) 2018-09-27 23:06:14 -07:00
trivialfis
5a7f7e7d49 Implement devices to devices reshard. (#3721) 
* Force clearing device memory before Reshard.
* Remove calculating row_segments for gpu_hist and gpu_sketch.
* Guard against changing device.
 2018-09-28 17:40:23 +12:00
Tong He
0b7fd74138 fix R check warning (#3728) 2018-09-27 17:53:49 -07:00
Philip Hyunsu Cho
51478a39c9
Fix #3730: scikit-learn 0.20 compatibility fix (#3731) 
* Fix #3730: scikit-learn 0.20 compatibility fix

sklearn.cross_validation has been removed from scikit-learn 0.20,
so replace it with sklearn.model_selection

* Display test names for Python tests for clarity
 2018-09-27 15:03:05 -07:00
Philip Hyunsu Cho
fbe9d41dd0 Disable flaky tests in R-package/tests/testthat/test_update.R (#3723) 2018-09-26 14:21:41 -07:00
Nan Zhu
79d854c695
[jvm-packages] fix errors in example (#3719) 
* add back train method but mark as deprecated

* fix scalastyle error

* add back train method but mark as deprecated

* add back train method but mark as deprecated

* add back train method but mark as deprecated

* fix scalastyle error

* fix scalastyle error

* fix scalastyle error

* instrumentation

* use log console

* better measurement

* fix erros in example

* update histmaker
 2018-09-22 16:39:38 -07:00
BruceZhao
3b5a1f389a [R] add a demo of multi-class classification R version (#3695) 
* add a demo of multi-class classification R version

* add a demo of multi-class classification result

* add intro to the demo readme

* Delete train.md

* Update README.md
 2018-09-21 23:06:40 -07:00
Takahiro Kojima
2405c59352 remove extra of (#3713) 2018-09-21 11:55:39 -07:00
Philip Hyunsu Cho
73140ce84c
Fix #3702: do not round up integer thresholds for integer features in JSON dump (#3717) 2018-09-21 01:11:21 -07:00
Nan Zhu
aa53e9fc8d
[jvm-packages] bump spark version (#3709) 2018-09-19 11:18:01 -07:00
trivialfis
9119f9e369 Fix gpu devices. (#3693) 
* Fix gpu_set normalized and unnormalized.
* Fix DeviceSpan.
 2018-09-19 17:39:42 +12:00
Andy Adinets
0f99cdfe0e Fixed an uninitialized pointer. (#3703) 2018-09-16 18:02:31 +12:00
Michael Mui
20a9e716bd [jvm-packages] Fix "obj_type" error to enable custom objectives and evaluations (#3646) 
credits to @mmui
 2018-09-14 12:06:33 -07:00
Dmitriy Rybalko
7bbb44182a update eval_metric doc (#3687) 2018-09-14 08:47:05 -07:00
Jerry Lin
9acd549dc7 [jvm-packages] Add rank:ndcg and rank:map to Spark supported objectives (#3697) 2018-09-13 09:51:24 -07:00
Chen Qin
42b108136f [jvm-packages] bump flink version number (#3686) 
* bump flink version number

* bump flink version number

* add missing hadoop dependency
 2018-09-13 09:33:09 -07:00
Philip Hyunsu Cho
bd41bd6605
Better error message for failed library loading (#3690) 
* Better error message for failed lib loading

* Address review comment + fix lint
 2018-09-12 22:37:26 -07:00
Philip Hyunsu Cho
3209b42b07
Include full text of Apache 2.0 license (#3698) 2018-09-12 20:46:55 -07:00
jakehoare
7707982a85 Amend xgb.createFolds to handle classes of a single element. (#3630) 
* Amend xgb.createFolds to handle classes of a single element.

* Fix variable name
 2018-09-12 09:23:05 -05:00
Vadim Khotilovich
ad3a0bbab8
Add the missing max_delta_step (#3668) 
* add max_delta_step to SplitEvaluator

* test for max_delta_step

* missing x2 factor for L1 term

* remove gamma from ElasticNet
 2018-09-12 08:43:41 -05:00
Nan Zhu
d1e75d615e
[jvm-packages] Remove copy paste error in test suite (#3692) 
* add back train method but mark as deprecated

* add back train method but mark as deprecated

* fix scalastyle error

* fix scalastyle error

* remove copy paste error
 2018-09-11 13:08:36 -07:00
Joseph Bradley
14a8b96476 [jvm-packages] xgboost-spark warning when Spark encryption is turned on (#3667) 
* added test, commented out right now

* reinstated test

* added fix for checking encryption settings

* fix by using RDD conf

* fix compilation

* renamed conf

* use SparkSession if available

* fix message

* nop

* code review fixes
 2018-09-10 14:21:01 -07:00
Philip Hyunsu Cho
3564b68b98
Fix #3397: early_stop callback does not maximize metric of form NDCG@n- (#3685) 
* Fix #3397: early_stop callback does not maximize metric of form NDCG@n-

Early stopping callback makes splits with '-' letter, which interferes
with metrics of form NDCG@n-. As a result, XGBoost tries to minimize
NDCG@n-, where it should be maximized instead.

Fix. Specify maxsplit=1.

* Python 2.x compatibility fix
 2018-09-08 19:46:25 -07:00
Andy Adinets
f606cb8ef4 Fixed the performance regression within EvaluateSplits(). (#3680) 
- it turns out creating an std::vector on every call is faster
  than cudaMallocHost()/cudaFreeHost()
 2018-09-08 14:48:45 +12:00
Matthew Tovbin
beab6e08dd Remove println in jsonDecode (#3665) 
Following issue  #3578
 2018-09-07 15:47:26 -07:00