4990 Commits

Author SHA1 Message Date
Chen Qin
ddcc2d85da Clean up cmake script and code includes (#106)
* Clean up CMake scripts and related include paths.
* Add unittests.
2019-09-26 02:29:04 -04:00
Vibhu Jawa
2fa8b359e0 Add support for cudf.Series (#4891) 2019-09-25 23:52:28 -04:00
Liangcai Li
82ee2317e8 Add case for LongParam. (#4885)
To support specifying long parameter as String, the same as other basic
type, such as Int, Double ...
2019-09-25 05:41:53 -07:00
Jiaming Yuan
b8433c455a
Rewrite Dask interface. (#4819) 2019-09-25 01:30:14 -04:00
Rong Ou
562bb0ae31 remove device shards (#4867) 2019-09-25 13:15:46 +08:00
Jiaming Yuan
0b89cd1dfa
Support gamma in GPU_Hist. (#4874)
* Just prevent building the tree instead of using an explicit pruner.
2019-09-24 10:16:08 +08:00
Jiaming Yuan
a40b72d127
Workaround isnan across different environments. (#4883) 2019-09-23 21:34:27 -04:00
Jiaming Yuan
c7416002e9
Fix DMatrix doc. (#4884) 2019-09-23 01:55:04 -04:00
Nan Zhu
fc8c9b0521
[jvm-packages] enable deterministic repartitioning when checkpoint is enabled (#4807)
* do reparititoning in DataUtil

* keep previous behavior of partitioning without checkpoint

* deterministic repartitioning

* change
2019-09-19 15:21:05 -07:00
Xu Xiao
277e25797b [jvm-packages] refine numAliveCores method of SparkParallelismTracker (#4858)
* refine numAliveCores

* refine XGBoostToMLlibParams

* fix waitForCondition

* resolve conflicts

* Update SparkParallelismTracker.scala
2019-09-19 15:18:29 -07:00
Honza Sterba
22209b7b95 [jvm-packages] Add BigDenseMatrix (#4383)
* Add BigDenseMatrix

* ability to create DMatrix with bigger than Integer.MAX_VALUE size arrays
* uses sun.misc.Unsafe

* make DMatrix test work from a jar as well
2019-09-18 20:46:14 -07:00
Xu Xiao
e92641887b remove unreached code of AllreduceRobust::CheckAndRecover (#108) 2019-09-18 23:06:59 -04:00
Jiaming Yuan
d4ce6807c7
Don't use _builtin_FUNCTION. (#107) 2019-09-18 12:05:23 -04:00
Jiaming Yuan
57106a3459
Fix parsing empty json object. (#4868)
* Fix parsing empty json object.

* Better error message.
2019-09-18 03:31:46 -04:00
Rong Ou
006eb80578 ignore vscode and clion files (#4866)
* ignore vscode and clion files

* ignore all .idea directories
2019-09-17 21:27:40 -04:00
Jiaming Yuan
d669ea1eaa
Deprecate set group (#4864)
* Convert jvm package and R package.

* Restore for compatibility.
2019-09-17 21:26:54 -04:00
Philip Hyunsu Cho
0e0955a6d8
Add link to Ruby XGBoost gem (#4856) 2019-09-17 10:40:44 -07:00
Jiaming Yuan
5374f52531
Complete cudf support. (#4850)
* Handles missing value.
* Accept all floating point and integer types.
* Move to cudf 9.0 API.
* Remove requirement on `null_count`.
* Arbitrary column types support.
2019-09-16 23:52:00 -04:00
Rong Ou
125bcec62e Move ellpack page construction into DMatrix (#4833) 2019-09-16 23:50:55 -04:00
Chen Qin
512f037e55 [rabit_bootstrap_cache ] failed xgb worker recover from other workers (#4808)
* Better recovery support.  Restarting only the failed workers.
2019-09-16 23:31:52 -04:00
Xu Xiao
c89bcc4de5 [blocking] fix parallel eval_split of hist updater (#4851)
* Don't call rabit functions inside parallel loop.
2019-09-13 09:35:03 -04:00
Chen Qin
9a7ac85d7e remove is_bootstrap parameter (#102)
* apply openmp simd

* clean __buildin detection, moving windows build check from xgboost project, add openmp support for vectorize reduce

* apply openmp only to rabit

* orgnize rabit signature

* remove is_bootstrap, use load_checkpoint as implict flag

* visual studio don't support latest openmp

* orgnize omp declarations

* replace memory copy with vector cast

* Revert "replace memory copy with vector cast"

This reverts commit 28de4792dcdff40d83d458510d23b7ef0b191d79.

* Revert "orgnize omp declarations"

This reverts commit 31341233d31ce93ccf34d700262b1f3f6690bbfe.

* remove openmp settings, merge into a upcoming pr

* mis

* per feedback, update comments
2019-09-10 11:45:50 -07:00
Jiaming Yuan
6a5e805886
Add n_jobs as an alias of nthread. (#4842) 2019-09-09 19:57:12 -04:00
Stephanie Yang
0fc7dcfe6c Add public group getter for java and scala (#4838)
* Add public group getter for java and scala

* Remove unnecessary param from javadoc

* Fix typo

* Fix another typo

* Add semicolon

* Fix javadoc return statement

* Fix missing return statement

* Add a unit test
2019-09-09 10:07:48 -07:00
Jiaming Yuan
f90e7f9aa8
Some comments for row partitioner. (#4832) 2019-09-06 03:01:42 -04:00
Jiaming Yuan
a5f232feb8
Fix calling GPU predictor (#4836)
* Fix calling GPU predictor
2019-09-05 19:09:38 -04:00
Jiaming Yuan
52d44e07fe
monitor for distributed envorinment. (#4829)
* Collect statistics from other ranks in monitor.

* Workaround old GCC bug.
2019-09-05 13:18:09 +08:00
Jiaming Yuan
c0fbeff0ab
Restrict access to cfg_ in gbm. (#4801)
* Restrict access to `cfg_` in gbm.

* Verify having correct updaters.

* Remove `grow_global_histmaker`

This updater is the same as `grow_histmaker`.  The former is not in our
document so we just remove it.
2019-09-02 00:43:19 -04:00
TinkleG
2aed0ae230 Fix auc error in distributed mode (#4798)
Need more work for a complete fix.  See #4663 .
2019-09-01 02:54:14 -04:00
Rong Ou
733ed24dd9 further cleanup of single process multi-GPU code (#4810)
* use subspan in gpu predictor instead of copying
* Revise `HostDeviceVector`
2019-08-30 05:27:23 -04:00
Nan Zhu
0184eb5d02
[jvm-packages] Refactor XGBoost.scala to put all params processing in one place (#4815)
* cleaning checkpoint file after a successful file

* address comments

* refactor xgboost.scala to avoid multiple changes when adding params

* consolidate params

* fix compilation issue

* fix failed test

* fix wrong name

* tyep conversion
2019-08-28 22:41:05 -07:00
Cyprien Ricque
830e73901d eval_metrics print fixed (#4803) 2019-08-28 23:52:18 -04:00
Chen Qin
5797dcb64e support bootstrap allreduce/broadcast (#98)
* support run rabit tests as xgboost subproject using xgboost/dmlc-core

* support tracker config set/get

* remove redudant printf

* remove redudant printf

* add c++0x declaration

* log allreduce/broadcast caller, engine should track caller stack for
investigation

* tracker support binary config format

* Revert "tracker support binary config format"

This reverts commit 2a28e5e2b55c200cb621af8d19f17ab1bc62503b.

* remove caller, prototype fetch allreduce/broadcast results from resbuf

* store cached allreduce/broadcast seq_no to tracker

* allow restore all caches from other nodes

* try new rabit collective cache, todo: recv_link seems down

* link up cache restore with main recovery

* cleanup load cache state

* update cache api

* pass test.mk

* have a working tests

* try to unify check into actionsummary

* more logging to debug distributed hist three method issue

* update rabit interface to support caller signature matching

* splite seq_counter from cur_cache_seq to different variables

* still see issue with inf loop

* support debug print caller as well as allreduce op

* cleanup

* remove get/set cache from model_recover, adding recover in
loadcheckpoint

* clarify rabit cache strategy, cache is set only by successful collective
call involving all nodes with unique cache key. if all nodes call
getcache at same time, we keep rabit run collective call. If some nodes
call getcache while others not, we backfill cache from those nodes with
most entries

* revert caller logs

* fix lint error

* fix engine mpi signature

* support getcache by ref

* allow result buffer presiet to filestream

* add loging

* try fix checkpoint failure recovery case

* use int64_t to avoid overflow caused seq fault

* try avoid int overflow

* try fix checkpoint failure recovery case

* try avoid seqno overflow to negative by offseting specifial flag value
adding cache seq no to checkpoint/load checkpoint/check point ack to avoid
confusion from cache recovery

* fix cache seq assert error

* remove loging, handle edge case

* add extensive log to checkpoint state  with different seq no

* fix lint errors

* clean up comments before merge back to master

* add logs to allreduce/broadcast/checkpoint

* use unsinged int 32 and give seq no larger range

* address remove allreduce dropseq code segment

* using caller signature to filter bootstrapallreduces

* remove get/set cache from empty

* apply signature to reducer

* apply signature to broadcast

* add key to broadcat log

* fix broadcast signature

* fix default _line value for non linux system

* adding comments, remove sleep(1)

* fix osx build issue

* try fix mpi

* fix doc

* fix engine_empty api

* logging, adding more logs, restore immutable assertion

* print unsinged int with ud

* fix lint

* rename seqtype to kSeq and KCache indicating it's usage
apply kDiffSeq check to load_cache routine

* comment allreduce/broadcast log

* allow tests run on arm

* enable flag to turn on / off cache

* add log info alert if user choose to enable rabit bootstrap cache

* add rabit_debug setting so user can use config to turn on

* log flags when user turn on rabit_debug

* force rabit restart if tracker assign -1 rank

* use OPENMP to vecotrize reducer

* address comment

* Revert "address comment"

This reverts commit 1dc61f33e7357dad8fa65528abeb81db92c5f9ed.

* fix checkpoint size print 0

* per feedback, remove DISABLEOPEMP, address race condition

* - remove openmp from this pr
- update name from cache to boostrapcache

* add default value of signature macros

* remove openmp from cmake file

* Update src/allreduce_robust.cc

Co-Authored-By: Philip Hyunsu Cho <chohyu01@cs.washington.edu>

* Update src/allreduce_robust.cc

Co-Authored-By: Philip Hyunsu Cho <chohyu01@cs.washington.edu>

* run test with cmake

* remove openmp

* fix cmake based tests

* use cmake test fix darwin .dylib issue

* move around rabit_signature definition due to windows build

* misc, add c++ check in CMakeFile

* per feedback

* resolve CMake file

* update rabit version
2019-08-27 18:12:33 -07:00
Oleksandr Pryimak
516955564b Update xgboost-spark doc (#4804) 2019-08-27 10:51:00 -07:00
Rong Ou
38ab79f889 Make HostDeviceVector single gpu only (#4773)
* Make HostDeviceVector single gpu only
2019-08-26 09:51:13 +12:00
Michaël Benesty
41227d1933 fix #4764 (#4800)
check all classes in xgb.get.handle
2019-08-21 21:42:37 -04:00
Jiaming Yuan
6e6216ad67
Skip related tests when sklearn is not installed. (#4791) 2019-08-21 00:32:52 -04:00
Jiaming Yuan
fba298fecb
Prevent copying data to host. (#4795) 2019-08-20 23:06:27 -04:00
Jiaming Yuan
3fa2ceb193
Add self. (#4794) 2019-08-20 14:41:30 +08:00
Jiaming Yuan
9700776597 Cudf support. (#4745)
* Initial support for cudf integration.

* Add two C APIs for consuming data and metainfo.

* Add CopyFrom for SimpleCSRSource as a generic function to consume the data.

* Add FromDeviceColumnar for consuming device data.

* Add new MetaInfo::SetInfo for consuming label, weight etc.
2019-08-19 16:51:40 +12:00
Jiaming Yuan
ab357dd41c
Remove plugin, cuda related code in automake & autoconf files (#4789)
* Build plugin example with CMake.

* Remove plugin, cuda related code in automake & autoconf files.

* Fix typo in GPU doc.
2019-08-18 16:54:34 -04:00
Jiaming Yuan
c358d95c44
Remove initializing stringstream reference. (#4788) 2019-08-18 09:59:47 -04:00
Jiaming Yuan
c81238b5c4
Clean up after removing gpu_exact. (#4777)
* Removed unused functions.
* Removed unused parameters.
* Move ValueConstraints into constraints.cuh since it's now only used in GPU_Hist.
2019-08-17 01:05:57 -04:00
Jiaming Yuan
b9b57f2289 Use long key id. (#4783) 2019-08-16 11:19:22 -07:00
Evan Kepner
53d4272c2a add os.PathLike support for file paths to DMatrix and Booster Python classes (#4757) 2019-08-15 04:46:25 -04:00
Nan Zhu
7b5cbcc846
[jvm-packages] cleaning checkpoint file after a successful training (#4754)
* cleaning checkpoint file after a successful file

* address comments
2019-08-14 10:57:47 -07:00
Xu Xiao
ef9af33a00 [HOTFIX] distributed training with hist method (#4716)
* add parallel test for hist.EvalualiteSplit

* update test_openmp.py

* update test_openmp.py

* update test_openmp.py

* update test_openmp.py

* update test_openmp.py

* fix OMP schedule policy

* fix clang-tidy

* add logging: total_num_bins

* fix

* fix

* test

* replace guided OPENMP policy with static in updater_quantile_hist.cc
2019-08-13 11:27:29 -07:00
Jiaming Yuan
c0ffe65f5c
Mimic cuda assert output in span check. (#4762) 2019-08-13 01:44:54 -04:00
Rong Ou
c5b229632d [BREAKING] prevent multi-gpu usage (#4749)
* prevent multi-gpu usage

* fix distributed test

* combine gpu predictor tests

* set upper bound on n_gpus
2019-08-13 09:11:35 +12:00
sriramch
198f3a6c4a Enable natural copies of the batch iterators without the need of the clone method (#4748)
- the synthesized copy constructor should do the appropriate job
2019-08-09 11:47:35 -04:00