Commits · 3586673ab28616ab1cd47f0e021c8be89bc06cca · tianlh / LightGBM-DCU

"git@developer.sourcefind.cn:tianlh/lightgbm-dcu.git" did not exist on "3ed0027bffe820ba6c81f32b82f02ef8302c6aad"

15 Nov, 2016 19 commits
- clean code · 3586673a
  Guolin Ke authored Nov 14, 2016
  
  3586673a
- simple exception handle and error message. Need RAII in the future. · 2a5d7abe
  Guolin Ke authored Nov 14, 2016
  
  2a5d7abe
- add internal_value_ in tree and output to txt model, so we can use for... · 45cbcb05
  allen authored Nov 13, 2016
```
add internal_value_ in tree and output to txt model, so we can use for calculating feature importance when predicting (#80)
```
  45cbcb05
- fix bugs in bin finder · 952099d6
  Guolin Ke authored Nov 13, 2016
  
  952099d6
- change dropping_seed to drop_seed · 9f04f276
  Guolin Ke authored Nov 12, 2016
  
  9f04f276
- change macOS library suffix to `.so` (#77) · da9baaf7
  wxchan authored Nov 12, 2016
```
change macOS library suffix to `.so` (#77)
```
  da9baaf7
- add DART: Dropouts meet Multiple Additive Regression Trees (#67) · 9390dc21
  wxchan authored Nov 11, 2016
```
add DART: Dropouts meet Multiple Additive Regression Trees (#67)
```
  9390dc21
- warning fixed · 5a483a8d
  Guolin Ke authored Nov 11, 2016
  
  5a483a8d
- change default prediction type · bbb528d7
  Guolin Ke authored Nov 11, 2016
  
  bbb528d7
- fix some bugs in bin construction. · d1e0bab5
  Guolin Ke authored Nov 11, 2016
  
  d1e0bab5
- change GetTrainnigScore to non-const. · 0f4ea846
  Guolin Ke authored Nov 11, 2016
  
  0f4ea846
- Update CMakeLists.txt · e096ea27
  Guolin Ke authored Nov 10, 2016
  
  e096ea27
- update travis, clean code · a44bc1ca
  Guolin Ke authored Nov 09, 2016
  
  a44bc1ca
- fix array index cross when met unknown features indices. · 63b122ca
  Guolin Ke authored Nov 09, 2016
  
  63b122ca
- support predict from file. add more tests · 4a451b89
  Guolin Ke authored Nov 09, 2016
  
  4a451b89
- reduce memory cost at sample phase · c6512e01
  Guolin Ke authored Nov 09, 2016
  
  c6512e01
- clean code · 2e5e9134
  Guolin Ke authored Nov 09, 2016
  
  2e5e9134
- Added complexity regularization parameters (L1, L2, min_gain_to_split) (#69) · 63d974bc
  Allard van Mossel authored Nov 08, 2016
  
  63d974bc
- add a small c_api test · 652d1618
  Guolin Ke authored Nov 08, 2016
  
  652d1618
08 Nov, 2016 4 commits
- update cmake to build library · 55cd6096
  Guolin Ke authored Nov 08, 2016
  
  55cd6096
- support int64_t for CSR/CSR's indptr · 044f79aa
  Guolin Ke authored Nov 08, 2016
  
  044f79aa
- some bugs fixed · 31b19afc
  Guolin Ke authored Nov 08, 2016
  
  31b19afc
- update VS project, fixed some bugs in dataset interface · 4a6bdcbe
  Guolin Ke authored Nov 08, 2016
  
  4a6bdcbe
07 Nov, 2016 7 commits
- simplify the logic of load model from string · fa643b3d
  Guolin Ke authored Nov 07, 2016
  
  fa643b3d
- clean code · 01e10529
  Guolin Ke authored Nov 07, 2016
  
  01e10529
- Add Interface of Booster · e161a746
  Guolin Ke authored Nov 07, 2016
  
  e161a746
- fix bug when splitting sparse bin with passed num_data=0 (#68) · ba2217ae
  Guolin Ke authored Nov 07, 2016
  
  ba2217ae
- clean code · 8c235f67
  Guolin Ke authored Nov 07, 2016
  
  8c235f67
- some refactor for predict and metric logic · 7b8bb4f2
  Guolin Ke authored Nov 07, 2016
  
  7b8bb4f2
- clean code · 33344088
  Guolin Ke authored Nov 07, 2016
  
  33344088
06 Nov, 2016 1 commit
- fix log · 7d25c253
  Qiwei Ye authored Nov 06, 2016
  
  7d25c253
05 Nov, 2016 4 commits
- add suport for csc format · c9d7ec24
  Guolin Ke authored Nov 05, 2016
  
  c9d7ec24
- finish the expose interface of dataset · 494fef34
  Guolin Ke authored Nov 05, 2016
  
  494fef34
- add load data from mat · d41c78f9
  Guolin Ke authored Nov 05, 2016
  
  d41c78f9
- fix sumup too small when set unbalance=true · 0241d323
  Guolin Ke authored Nov 05, 2016
  
  0241d323
04 Nov, 2016 1 commit
- use dataset_loader to load data · 1c08e71e
  Guolin Ke authored Nov 04, 2016
  
  1c08e71e
03 Nov, 2016 4 commits
- Improved consistency and wording of user-facing logs and documentation · 8497af62
  Allardvm authored Nov 03, 2016
```
Packages that parse LightGBM’s logs will require minor changes to
parsing logic to work correctly.
```
  8497af62
- move num_used_model out of predict function · 4e291459
  Guolin Ke authored Nov 03, 2016
  
  4e291459
- support init_score for multiclass classification (#62) · 01ed04df
  wxchan authored Nov 03, 2016
```
support init_score for multiclass classification (#62)
```
  01ed04df
- change the type in set_query_at · 665c9dba
  Guolin Ke authored Nov 03, 2016
  
  665c9dba