GitLab

* Add parameter max_bin_by_feature.

* Fix minor bug.

* Fix minor bug.

* Fix calculation of header size for writing binary file.

* Fix style issues.

* Fix python style issue.

* Fix test and python style issue.

* add balanced bagging

* refine code

* fix format

* clarify usage only for binary application

* Implement top-k multiclass error metric. Add new parameter top_k_threshold.

* Add test for multiclass metrics

* Make test less sensitive to avoid floating-point issues.

* Change tabs to spaces.

* Fix problem with test in Python 2. Refactor to use np.testing. Decrease number of training rounds so loss is larger and easier to compare.

* Move multiclass tests into test_engine.py

* Change parameter name from top_k_threshold to multi_error_top_k.

* Fix top-k error metric to handle case where scores are equal. Update tests and docs.

* Change name of top-k metric to multi_error@k.

* Change tabs to spaces.

* Fix formatting.

* Fix minor issues in docs.

* first metric only in earlystopping for cli

* code clean

* added note about CLI only usage

* removed note about CLI only usage

* PR #1879

* Update docs with parameter_generator.py

* Update wrapper doc for sklearn

* fix travis badge

* updated GitHub Microsoft URL

* Update doc string for pred_contrib

See comments at the end of #1969

* Update basic.py

* Update basic.py

* update doc strings

* update equals sign in doc string

* strip whitespace and gen rst

* strip whitespace

* Add configuration parameters for CEGB.

* Add skeleton CEGB tree learner

Like the original CEGB version, this inherits from SerialTreeLearner.
Currently, it changes nothing from the original.

* Track features used in CEGB tree learner.

* Pull CEGB tradeoff and coupled feature penalty from config.

* Implement finding best splits for CEGB

This is heavily based on the serial version, but just adds using the coupled penalties.

* Set proper defaults for cegb parameters.

* Ensure sanity checks don't switch off CEGB.

* Implement per-data-point feature penalties in CEGB.

* Implement split penalty and remove unused parameters.

* Merge changes from CEGB tree learner into serial tree learner

* Represent features_used_in_data by a bitset, to reduce the memory overhead of CEGB, and add sanity checks for the lengths of the penalty vectors.

* Fix bug where CEGB would incorrectly penalise a previously used feature

The tree learner did not update the gains of previously computed leaf splits when splitting a leaf elsewhere in the tree.
This caused it to prefer new features due to incorrectly penalising splitting on previously used features.

* Document CEGB parameters and add them to the appropriate section.

* Remove leftover reference to cegb tree learner.

* Remove outdated diff.

* Fix warnings

* Fix minor issues identified by @StrikerRUS.

* Add docs section on CEGB, including citation.

* Fix link.

* Fix CI failure.

* Add some unit tests

* Fix pylint issues.

* Fix remaining pylint issue

[docs] added notes about params usage when data is provided via path and removed unused param (#2024)

* added notes about params usage when data is provided via path

* fixed init score and valid init score params note

* fixed binary params description

* renamed helper folder to helpers

* added library dependencies check

* quick fix for better understanding

* update document for forced split

* typo fix

* made NOTE bold

* made Note bold

* Update GPU-Targets.rst

Fix some inaccurate information in docs

* fix travisCI warning

* fix typos

* update config.h

* warning on categorical feature with sparse values

* [docs] categorical features note

* pass params to _InnerPredictor in train and cv

* fixed verbosity param description

* treat silent param as Fatal log level

* create Dataset in refit method silently

* do not overwrite verbose param by silent argument

* 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

* add support of refit-decay

* add refit into c_api

* add test

* update document

* Update basic.py

* Update test_engine.py

* Update basic.py

* Update test_engine.py

* fix comments

* update test

* fix the comments

* Update test_engine.py

* broadcast info about negative values in categorical features to python package

* update link to categorical_feature parameter

* updated docs of num_iterations parameter

* updated num_iterations param description for R

* fix custom metric for multiclass

* fix alias

* fix bug

* fix indent

* added new aliases for params

* run helper/parameter_generator.py

* removed useless test

* fixed typo in parameter list Parameters.rst

fixed typo from contributs to contributes

* Update config.h

* added permalink to each parameter

* run parameter_generator.py

* concretized links

* hotfix

* run parameter_generator.py

* added space between param name and permalink

* added info about initscore_filename in Continued Training section and edited links

* replaced aliases for weight and group with original names

* disable emoji

* add per-feature-penalites

* fix comment

* removed excess whitespaces

* don't use built-in name for variable

* simplified line parsing

* chanched link to related

* run parameter_generator.py

* removed old targets

* use tuples instead of list where possible

* hotfix for descriptions were erased and only last one was kept

* run parameter_generator.py

* separated checks from aliases section

* generate parameters description from config file

* made pylint happy

* added checks description

* fixed links handling in desc and descl2 fields

* refine Core Parameters section

* removed excess quotes from default values

* fixed parameter type retrieve

* added note about auto-generated parameters

* A nitpicky grammer edit with minor clarifications added.

* fix link

* strike s

* try a different optimal-split link, clarify experimental details

* smoothing the FAQ

* edit Features.rst

* several minor edits throughout docs

* historgram-based

* Note the relationship between `binary` and `xentropy` in the docs and provide an example that compares them

* Pass pylint and fix docs formatting. "pylint: disable = no-name-in-module" because pylint does not believe scipy.special contains expit

* pass pycodestyle

* pass pycodestyle with correct flags; document example in readme