GitLab

* bring consistency and clearness into early_stopping_rounds desc, metric desc and implementation

* hotfix

* hotfix

* used NDCG as default metric for lambdarank task

* fixed missed methods at ReadTheDocs and changed default eval_metric

* leaved only unique metrics

* fixed comment

* added clang test

* hotfix

* downgrade required AppleCLang version

* fixed CMakeLists

* prevent case when CMake is downgraded by libomp

* fixed Clang test

* moved MSVC version check to the better place

* compile with Clang at Travis on Linux

* clear .travis.yml

* fixed indents

* added support of MPI test at Travis with macOS

* added FAQ entry about OpenMP library error

* updated installation guide

* moved APPLE_OUTPUT_DYLIB to other options

* fix check-docs test: ignore old url from error message

* added example of symlink command in FAQ

* fixed markdown

* capitalized heading

* updated Python installation guide

* changed mode in message command

* added notes about MinGW efficiency problems

* updated R-package installation guide

* use quotes for gcc version

* removed excess 'Note' phrase

* added new aliases for params

* run helper/parameter_generator.py

* removed useless test

* Fix lgb.Dataset construct issue with 1:n sequence in R v 3.5

* Add tests for feature penalties

* added missed description of plot_example in python_guide folder and fixed consistency for packages naming

* more reliable OS detection

* fixed grammar

* made pylint happy

2.1.0 -> 2.1.1

The document of `early_stopping_rounds` says it will check all of
eval_set. But, this is not true. It doesn't check the dataset
specified as the training data.

This change appends an extra phrase "except the training data" to all
of the sentences "If there's more than one, will check all of them" in
documents.

* added a note about g++ versions for macOS users

* hotfix excess dot in FAQ

* added notes about gcc version in installation guides

* Bump R version

* Bump CLI/Py version

* Roll back to 2.1.1 in R

* Fix Microsoft/LightGBM#1280

* Update lgb.Booster.R

* Update lgb.Booster.R

* Add authors Microsoft/LightGBM#1260

* Add line break at the end of file

* Fixed misc CRAN issues

* Added additional details to R-package DESCRIPTION

* added code to expose c_api_pred_contrib in the R package

* removed Rprintf

* reverted to previous version of install.libs.R

objective funciton in reset_parameter()

Fixes error algb.params2str to lgb.params2str

* fix the bug

* fix R

* Singlethread make to lower RAM usage significantly

* Update install.libs.R

* fix early stopping edge case

* fix message.

* fix tests

* fix GPU tests.

* updated travis

* renamed OSX->macOS; updated gcc installation on macOS

* updated travis macos installation

 Please enter the commit message for your changes. Lines starting

* updated installation guide for macos

* fixed spelling

* renamed appveyor file

* updated Travis

* use int64 for string

* [R] Fatal when CSC exceed int32.max