GitLab

* [python] add type hints in docs/conf.py

* more specific hint for sphinx app

* Apply suggestions from code review
Co-authored-by: Nikita Titov <nekit94-08@mail.ru>
Co-authored-by: Nikita Titov <nekit94-08@mail.ru>

* [R-package] fix OpenMP checking on macOS

* adding clang flags back

* fix detection on Mac

* remove CC since it is now unused

* use flags specific to C++11

* regenerate configure

Co-authored-by: Nikita Titov <nekit94-12@hotmail.com>

* [R-package] pass compiler flags to checks in configure

* leave OpenMP check alone

* prefer C++11 flags

* install Matrix in valgrind tests

* one less AC_LANG() call
Co-authored-by: Nikita Titov <nekit94-12@hotmail.com>

Co-authored-by: Nikita Titov <nekit94-12@hotmail.com>

* [ci] prefer 'apt update' to 'apt-get update' in some R builds

* use --allow-releaseinfo-change

* [R-package] use C++ compiler for pre-compile checks on Windows

* install Matrix in valgrind test

* Add {Matrix} in more places in CI and docs

* use CXX11

* use flags specific to C++11

* missing backtick
Co-authored-by: Nikita Titov <nekit94-12@hotmail.com>

* [docs] [R-package] use CRAN-style builds when building pkgdown site

* install with --with-keep.source

* empty commit

* set new_proccess = FALSE to get a better traceback

* copy pkgdown config

* [ci] move Solaris and valgrind test steps into scripts

* Update .github/workflows/r_solaris.yml
Co-authored-by: Nikita Titov <nekit94-08@mail.ru>

* relatiev paths
Co-authored-by: Nikita Titov <nekit94-08@mail.ru>

* reduce number of collisions tests

* measure tests execution time

* measure tests execution time in bdist task

* remove durations in bdist task

* find all needed ports in each worker at once

* lint

* better naming

* use _HostWorkers in test

* Update c_api LGBM_SampleIndices() comment.

rand.Sample() now returns exactly given number of samples, thus the
comment should be fixed.

* Update include/LightGBM/c_api.h
Co-authored-by: Nikita Titov <nekit94-08@mail.ru>
Co-authored-by: Nikita Titov <nekit94-08@mail.ru>

* [python] support Dataset.get_data for Sequence input.

* Tweaks according to review comments.

* Apply suggestions from code review
Co-authored-by: Nikita Titov <nekit94-08@mail.ru>

* Add test cases.

* fix import order in test_basic.py
Co-authored-by: Nikita Titov <nekit94-08@mail.ru>

* remove R docs

* bump deps

* Update parameter_generator.py

* Update parameter_generator.py

* Update parameter_generator.py

* Update parameter_generator.py

* Update parameter_generator.py

* Update parameter_generator.py

* Update helpers/parameter_generator.py
Co-authored-by: James Lamb <jaylamb20@gmail.com>

* Update helpers/parameter_generator.py
Co-authored-by: James Lamb <jaylamb20@gmail.com>

* Update helpers/parameter_generator.py
Co-authored-by: James Lamb <jaylamb20@gmail.com>

* Update helpers/parameter_generator.py
Co-authored-by: James Lamb <jaylamb20@gmail.com>

* Update helpers/parameter_generator.py
Co-authored-by: James Lamb <jaylamb20@gmail.com>

* Update helpers/parameter_generator.py
Co-authored-by: James Lamb <jaylamb20@gmail.com>

* Update helpers/parameter_generator.py
Co-authored-by: James Lamb <jaylamb20@gmail.com>

* Update helpers/parameter_generator.py
Co-authored-by: James Lamb <jaylamb20@gmail.com>

* Update helpers/parameter_generator.py
Co-authored-by: James Lamb <jaylamb20@gmail.com>

* Update helpers/parameter_generator.py
Co-authored-by: James Lamb <jaylamb20@gmail.com>

* additional hinting fixes

* Update parameter_generator.py
Co-authored-by: James Lamb <jaylamb20@gmail.com>

This removes code duplication for creating sample indices.

* Fix undefined behavior with NaN input in CategoricalDecision()

* Always associate the right child with NaN inputs

* remove preinstalled possibly conflicting software from PATH in CI jobs

* preserve pandoc

* [ci] add CI job running rchk

* try commenting out more stuff

* ignore R internal error

* pipes

* remove PROTECT()

* try removing testthat

* revert temporary testing changes

* use absolute() on paths first

* migrate to pathlib in setup.py

* call predict on one row of data to determine output shape

* make DaskLGBMRanker predict method equal to the others

* remove extra drop_axis

* allow to pass some params as pathlib.Path objects

* fix lint

* improve indentation

[dask] Make output of feature contribution predictions for sparse matrices match those from sklearn estimators (fixes #3881) (#4378)

* test_classifier working

* adding tests

* docs

* tests

* revert unnecessary changes in tests

* test output type

* linting

* linting

* use from_delayed() instead

* docstring pycodestyle is happy with

* isort

* put pytest skips back

* respect sparse return type

* fix doc

* remove unnecessary dask_array_concatenate()

* Apply suggestions from code review
Co-authored-by: Nikita Titov <nekit94-08@mail.ru>

* Apply suggestions from code review
Co-authored-by: Nikita Titov <nekit94-08@mail.ru>

* update predict_proba() docstring

* remove unnecessary np.array()

* Update python-package/lightgbm/dask.py
Co-authored-by: Nikita Titov <nekit94-08@mail.ru>

* fix assertion

* fix test use of len()

* restore np.array() in tests

* use np.asarray() instead

* use toarray()

* remove empty functions in compat
Co-authored-by: Nikita Titov <nekit94-08@mail.ru>