10000 Merging main · scikit-learn/scikit-learn@7352549 · GitHub
[go: up one dir, main page]
Skip to content

Commit 7352549

Browse files
adam2392adam2392
committed
Merging main
Signed-off-by: Adam Li <adam2392@gmail.com>
2 parents bee1876 + e7af195 commit 7352549

40 files changed

+1318
-1632
lines changed

‎.codecov.yml

Lines changed: 0 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -30,5 +30,4 @@ ignore:
3030
- "sklearn/_build_utils"
3131
- "sklearn/__check_build"
3232
- "sklearn/_min_dependencies.py"
33-
- "**/setup.py"
3433
- "**/conftest.py"

‎.coveragerc

Lines changed: 0 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -6,4 +6,3 @@ omit =
66
*/sklearn/externals/*
77
*/sklearn/_build_utils/*
88
*/benchmarks/*
9-
**/setup.py

‎.github/workflows/wheels.yml

Lines changed: 0 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -164,7 +164,6 @@ jobs:
164164
CIBW_PRERELEASE_PYTHONS: ${{ matrix.prerelease_pythons }}
165165
CIBW_FREE_THREADED_SUPPORT: ${{ matrix.free_threaded_support }}
166166
CIBW_ENVIRONMENT: SKLEARN_SKIP_NETWORK_TESTS=1
167-
SKLEARN_BUILD_PARALLEL=3
168167
CIBW_BUILD: cp${{ matrix.python }}-${{ matrix.platform_id }}
169168
CIBW_ARCHS: all
170169
CIBW_MANYLINUX_X86_64_IMAGE: ${{ matrix.manylinux_image }}
@@ -216,8 +215,6 @@ jobs:
216215

217216
- name: Build source distribution
218217
run: bash build_tools/github/build_source.sh
219-
env:
220-
SKLEARN_BUILD_PARALLEL: 3
221218

222219
- name: Test source distribution
223220
run: bash build_tools/github/test_source.sh

‎Makefile

Lines changed: 3 additions & 12 deletions
Original file line numberDiff line numberDiff line change
@@ -7,9 +7,7 @@ all:
77
@echo "Please use 'make <target>' where <target> is one of"
88
@echo " dev build scikit-learn with Meson"
99
@echo " clean clean scikit-learn Meson build. Very rarely needed,"
10-
@echo " one use case is when switching back to setuptools"
11-
@echo " dev-setuptools build scikit-learn with setuptools (deprecated)"
12-
@echo " clean-setuptools clean scikit-learn setuptools build (deprecated)"
10+
@echo " since meson-python recompiles on import."
1311

1412
.PHONY: all
1513

@@ -23,14 +21,7 @@ clean: clean-meson
2321
clean-meson:
2422
pip uninstall -y scikit-learn
2523
# It seems in some cases removing the folder avoids weird compilation
26-
# errors (e.g. when switching from numpy>=2 to numpy<2). For some
24+
# errors (e.g. when switching from numpy>=2 to numpy<2). For some
2725
# reason ninja clean -C $(DEFAULT_MESON_BUILD_DIR) is not
28-
# enough
26+
# enough.
2927
rm -rf $(DEFAULT_MESON_BUILD_DIR)
30-
31-
dev-setuptools:
32-
$(PYTHON) setup.py build_ext -i
33-
34-
clean-setuptools:
35-
$(PYTHON) setup.py clean
36-
rm -rf dist

‎azure-pipelines.yml

Lines changed: 1 addition & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -208,7 +208,6 @@ jobs:
208208
SKLEARN_ENABLE_DEBUG_CYTHON_DIRECTIVES: '1'
209209
SKLEARN_RUN_FLOAT32_TESTS: '1'
210210
SKLEARN_TESTS_GLOBAL_RANDOM_SEED: '2' # non-default seed
211-
BUILD_WITH_SETUPTOOLS: 'true'
212211
# Linux environment to test the latest available dependencies.
213212
# It runs tests requiring lightgbm, pandas and PyAMG.
214213
pylatest_pip_openblas_pandas:
@@ -236,7 +235,7 @@ jobs:
236235
)
237236
matrix:
238237
debian_atlas_32bit:
239-
DOCKER_CONTAINER: 'i386/debian:11.2'
238+
DOCKER_CONTAINER: 'i386/debian:12'
240239
DISTRIB: 'debian-32'
241240
COVERAGE: "true"
242241
LOCK_FILE: './build_tools/azure/debian_atlas_32bit_lock.txt'

‎build_tools/azure/install.sh

Lines changed: 1 addition & 13 deletions
Original file line numberDiff line numberDiff line change
@@ -107,10 +107,6 @@ scikit_learn_install() {
107107
setup_ccache
108108
show_installed_libraries
109109

110-
# Set parallelism to 3 to overlap IO bound tasks with CPU bound tasks on CI
111-
# workers with 2 cores when building the compiled extensions of scikit-learn.
112-
export SKLEARN_BUILD_PARALLEL=3
113-
114110
if [[ "$UNAMESTR" == "Darwin" && "$SKLEARN_TEST_NO_OPENMP" == "true" ]]; then
115111
# Without openmp, we use the system clang. Here we use /usr/bin/ar
116112
# instead because llvm-ar errors
@@ -129,9 +125,7 @@ scikit_learn_install() {
129125
export LDFLAGS="$LDFLAGS -Wl,--sysroot=/"
130126
fi
131127

132-
if [[ "$BUILD_WITH_SETUPTOOLS" == "true" ]]; then
133-
python setup.py develop
134-
elif [[ "$PIP_BUILD_ISOLATION" == "true" ]]; then
128+
if [[ "$PIP_BUILD_ISOLATION" == "true" ]]; then
135129
# Check that pip can automatically build scikit-learn with the build
136130
# dependencies specified in pyproject.toml using an isolated build
137131
# environment:
@@ -143,12 +137,6 @@ scikit_learn_install() {
143137
# toolchain
144138
ADDITIONAL_PIP_OPTIONS='-Csetup-args=--vsenv'
145139
fi
146-
# TODO Always add --check-build-dependencies when all CI builds have
147-
# pip >= 22.1.1. At the time of writing, two CI builds (debian32_atlas and
148-
# ubuntu_atlas) have an older pip
149-
if pip install --help | grep check-build-dependencies; then
150-
ADDITIONAL_PIP_OPTIONS="$ADDITIONAL_PIP_OPTIONS --check-build-dependencies"
151-
fi
152140
# Use the pre-installed build dependencies and build directly in the
153141
# current environment.
154142
pip install --verbose --no-build-isolation --editable . $ADDITIONAL_PIP_OPTIONS

‎build_tools/azure/pylatest_pip_openblas_pandas_environment.yml

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -4,7 +4,7 @@
44
channels:
55
- defaults
66
dependencies:
7-
- python=3.9
7+
- python=3.11
88
- ccache
99
- pip
1010
- pip:

‎build_tools/azure/pylatest_pip_openblas_pandas_linux-64_conda.lock

Lines changed: 22 additions & 25 deletions
Large diffs are not rendered by default.

‎build_tools/azure/pymin_conda_defaults_openblas_environment.yml

Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -15,9 +15,11 @@ dependencies:
1515
- pytest
1616
- pytest-xdist
1717
- pillow
18+
- ninja
1819
- pytest-cov
1920
- coverage
2021
- ccache
2122
- pip
2223
- pip:
2324
- threadpoolctl==3.1.0 # min
25+
- meson-python==0.16.0 # min

‎build_tools/azure/pymin_conda_defaults_openblas_linux-64_conda.lock

Lines changed: 7 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,6 @@
11
# Generated by conda-lock.
22
# platform: linux-64
3-
# input_hash: 244b6a064d3785ea62baaf9436848821d153846b455c2976f5e811182e848c83
3+
# input_hash: e4db53ad2240ff5f57679dd93701c30b6712ac3a43ec04e18b74132f2948b4cd
44
@EXPLICIT
55
https://repo.anaconda.com/pkgs/main/linux-64/_libgcc_mutex-0.1-main.conda#c3473ff8bdb3d124ed5ff11ec380d6f9
66
https://repo.anaconda.com/pkgs/main/linux-64/blas-1.0-openblas.conda#9ddfcaef10d79366c90128f5dc444be8
@@ -28,6 +28,7 @@ https://repo.anaconda.com/pkgs/main/linux-64/libwebp-base-1.3.2-h5eee18b_0.conda
2828
https://repo.anaconda.com/pkgs/main/linux-64/libxcb-1.15-h7f8727e_0.conda#ada518dcadd6aaee9aae47ba9a671553
2929
https://repo.anaconda.com/pkgs/main/linux-64/lz4-c-1.9.4-h6a678d5_1.conda#2ee58861f2b92b868ce761abb831819d
3030
https://repo.anaconda.com/pkgs/main/linux-64/ncurses-6.4-h6a678d5_0.conda#5558eec6e2191741a92f832ea826251c
31+
https://repo.anaconda.com/pkgs/main/linux-64/ninja-base-1.10.2-hd09550d_5.conda#09dcbad622d58caaeefe46cd399f0a76
3132
https://repo.anaconda.com/pkgs/main/linux-64/openssl-3.0.14-h5eee18b_0.conda#37b6dad6aa49000a4230a9f0cad172f6
3233
https://repo.anaconda.com/pkgs/main/linux-64/xz-5.4.6-h5eee18b_1.conda#1562802f843297ee776a50b9329597ed
3334
https://repo.anaconda.com/pkgs/main/linux-64/zlib-1.2.13-h5eee18b_1.conda#92e42d8310108b0a440fb2e60b2b2a25
@@ -56,7 +57,7 @@ https://repo.anaconda.com/pkgs/main/linux-64/libclang-14.0.6-default_hc6dbbc7_1.
5657
https://repo.anaconda.com/pkgs/main/linux-64/libpq-12.17-hdbd6064_0.conda#6bed363e25859faff66bf546a11c10e8
5758
https://repo.anaconda.com/pkgs/main/linux-64/openjpeg-2.4.0-h9ca470c_1.conda#dfd4b36eb8ddaffeca0ab412de63c3e2
5859
https://repo.anaconda.com/pkgs/main/linux-64/python-3.9.19-h955ad1f_1.conda#4b453281859c293c9d577271f3b18a0d
59-
https://repo.anaconda.com/pkgs/main/linux-64/certifi-2024.6.2-py39h06a4308_0.conda#738daf43271605d7291ecae0e8cac41c
60+
https://repo.anaconda.com/pkgs/main/linux-64/certifi-2024.7.4-py39h06a4308_0.conda#add87fa3b69a43e4e9ea1e619b267c4b
6061
https://repo.anaconda.com/pkgs/main/noarch/cycler-0.11.0-pyhd3eb1b0_0.conda#f5e365d2cdb66d547eb8c3ab93843aab
6162
https://repo.anaconda.com/pkgs/main/linux-64/cython-3.0.10-py39h5eee18b_0.conda#1419a658ed2b4d5c3ac1964f33143b64
6263
https://repo.anaconda.com/pkgs/main/linux-64/exceptiongroup-1.2.0-py39h06a4308_0.conda#960e2cb83ac5134df8e593a130aa11af
@@ -66,6 +67,7 @@ https://repo.anaconda.com/pkgs/main/noarch/iniconfig-1.1.1-pyhd3eb1b0_0.tar.bz2#
6667
https://repo.anaconda.com/pkgs/main/linux-64/joblib-1.2.0-py39h06a4308_0.conda#ac1f5687d70aa1128cbecb26bc9e559d
6768
https://repo.anaconda.com/pkgs/main/linux-64/kiwisolver-1.4.4-py39h6a678d5_0.conda#3d57aedbfbd054ce57fb3c1e4448828c
6869
https://repo.anaconda.com/pkgs/main/linux-64/mysql-5.7.24-h721c034_2.conda#dfc19ca2466d275c4c1f73b62c57f37b
70+
https://repo.anaconda.com/pkgs/main/linux-64/ninja-1.10.2-h06a4308_5.conda#6fc219bbc4c8dbb9060b5b7fe31ae83d
6971
https://repo.anaconda.com/pkgs/main/linux-64/numpy-base-1.21.6-py39h375b286_1.conda#0061d9193658774ab79fc85d143a94fc
7072
https://repo.anaconda.com/pkgs/main/linux-64/packaging-24.1-py39h06a4308_0.conda#e80d41ffc9450162ef10cbbb9b4ec7e9
7173
https://repo.anaconda.com/pkgs/main/linux-64/pillow-10.3.0-py39h5eee18b_0.conda#b346d6c71267c1553b6c18d3db5fdf6d
@@ -96,4 +98,7 @@ https://repo.anaconda.com/pkgs/main/linux-64/pyamg-4.2.3-py39h79cecc1_0.conda#af
9698
https://repo.anaconda.com/pkgs/main/linux-64/qt-main-5.15.2-h53bd1ea_10.conda#bd0c79e82df6323f638bdcb871891b61
9799
https://repo.anaconda.com/pkgs/main/linux-64/pyqt-5.15.10-py39h6a678d5_0.conda#52da5ff9b1144b078d2f41bab0b213f2
98100
https://repo.anaconda.com/pkgs/main/linux-64/matplotlib-3.3.4-py39h06a4308_0.conda#384fc5e01ebfcf30e7161119d3029b5a
101+
# pip meson @ https://files.pythonhosted.org/packages/44/b2/d4433391a7c5e94a39b50ca7295a8ceba736e7c72c455752a60122f52453/meson-1.4.1-py3-none-any.whl#sha256=d5acc3abae2dad3c70ddcbd10acac92b78b144d34d43f40f5b8ac31dfd8a826a
99102
# pip threadpoolctl @ https://files.pythonhosted.org/packages/61/cf/6e354304bcb9c6413c4e02a747b600061c21d38ba51e7e544ac7bc66aecc/threadpoolctl-3.1.0-py3-none-any.whl#sha256=8b99adda265feb6773280df41eece7b2e6561b772d21ffd52e372f999024907b
103+
# pip pyproject-metadata @ https://files.pythonhosted.org/packages/aa/5f/bb5970d3d04173b46c9037109f7f05fc8904ff5be073ee49bb6ff00301bc/pyproject_metadata-0.8.0-py3-none-any.whl#sha256=ad858d448e1d3a1fb408ac5bac9ea7743e7a8bbb472f2693aaa334d2db42f526
104+
# pip meson-python @ https://files.pythonhosted.org/packages/91/c0/104cb6244c83fe6bc3886f144cc433db0c0c78efac5dc00e409a5a08c87d/meson_python-0.16.0-py3-none-any.whl#sha256=842dc9f5dc29e55fc769ff1b6fe328412fe6c870220fc321060a1d2d395e69e8

‎build_tools/circle/build_doc.sh

Lines changed: 0 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -176,9 +176,6 @@ conda activate $CONDA_ENV_NAME
176176

177177
show_installed_libraries
178178

179-
# Set parallelism to 3 to overlap IO bound tasks with CPU bound tasks on CI
180-
# workers with 2 cores when building the compiled extensions of scikit-learn.
181-
export SKLEARN_BUILD_PARALLEL=3
182179
pip install -e . --no-build-isolation
183180

184181
echo "ccache build summary:"

‎build_tools/cirrus/arm_wheel.yml

Lines changed: 0 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -8,7 +8,6 @@ linux_arm64_wheel_task:
88
memory: 4G
99
env:
1010
CIBW_ENVIRONMENT: SKLEARN_SKIP_NETWORK_TESTS=1
11-
SKLEARN_BUILD_PARALLEL=5
1211
CIBW_TEST_COMMAND: bash {project}/build_tools/wheels/test_wheels.sh
1312
CIBW_TEST_REQUIRES: pytest pandas threadpoolctl pytest-xdist
1413
CIBW_BUILD_VERBOSITY: 1

‎build_tools/cirrus/build_test_arm.sh

Lines changed: 0 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -37,10 +37,6 @@ setup_ccache
3737

3838
python --version
3939

40-
# Set parallelism to $N_CORES + 1 to overlap IO bound tasks with CPU bound tasks on CI
41-
# workers with $N_CORES cores when building the compiled extensions of scikit-learn.
42-
export SKLEARN_BUILD_PARALLEL=$(($N_CORES + 1))
43-
4440
# Disable the build isolation and build in the tree so that the same folder can be
4541
# cached between CI runs.
4642
pip install --verbose --no-build-isolation .

‎build_tools/update_environments_and_lock_files.py

Lines changed: 7 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -179,7 +179,7 @@ def remove_from(alist, to_remove):
179179
"channels": ["defaults"],
180180
"conda_dependencies": remove_from(
181181
common_dependencies,
182-
["pandas", "threadpoolctl", "pip", "ninja", "meson-python"],
182+
["pandas", "threadpoolctl", "pip", "meson-python"],
183183
)
184184
+ ["ccache"],
185185
"package_constraints": {
@@ -191,10 +191,11 @@ def remove_from(alist, to_remove):
191191
"cython": "min",
192192
"joblib": "min",
193193
"threadpoolctl": "min",
194+
"meson-python": "min",
194195
},
195196
# TODO: put pip dependencies back to conda dependencies when required
196197
# version is available on the defaults channel.
197-
"pip_dependencies": ["threadpoolctl"],
198+
"pip_dependencies": ["threadpoolctl", "meson-python"],
198199
},
199200
{
200201
"name": "pymin_conda_forge_openblas_ubuntu_2204",
@@ -230,9 +231,10 @@ def remove_from(alist, to_remove):
230231
+ ["array-api-compat", "array-api-strict"]
231232
),
232233
"package_constraints": {
233-
# XXX: we would like to use the latest version of Python but this makes
234-
# the CI much slower. We need to investigate why.
235-
"python": "3.9",
234+
# XXX: we would like to use the latest Python version, but for now using
235+
# Python 3.12 makes the CI much slower so we use Python 3.11. See
236+
# https://github.com/scikit-learn/scikit-learn/pull/29444#issuecomment-2219550662.
237+
"python": "3.11",
236238
},
237239
},
238240
{

‎doc/metadata_routing.rst

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -301,6 +301,7 @@ Meta-estimators and functions supporting metadata routing:
301301
- :class:`sklearn.model_selection.HalvingGridSearchCV`
302302
- :class:`sklearn.model_selection.HalvingRandomSearchCV`
303303
- :class:`sklearn.model_selection.RandomizedSearchCV`
304+
- :class:`sklearn.model_selection.permutation_test_score`
304305
- :func:`sklearn.model_selection.cross_validate`
305306
- :func:`sklearn.model_selection.cross_val_score`
306307
- :func:`sklearn.model_selection.cross_val_predict`
@@ -324,4 +325,3 @@ Meta-estimators and tools not supporting metadata routing yet:
324325
- :class:`sklearn.feature_selection.RFE`
325326
- :class:`sklearn.feature_selection.RFECV`
326327
- :class:`sklearn.feature_selection.SequentialFeatureSelector`
327-
- :class:`sklearn.model_selection.permutation_test_score`

‎doc/modules/array_api.rst

Lines changed: 18 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -123,7 +123,9 @@ Metrics
123123
- :func:`sklearn.metrics.pairwise.additive_chi2_kernel`
124124
- :func:`sklearn.metrics.pairwise.chi2_kernel`
125125
- :func:`sklearn.metrics.pairwise.cosine_similarity`
126+
- :func:`sklearn.metrics.pairwise.euclidean_distances` (see :ref:`device_support_for_float64`)
126127
- :func:`sklearn.metrics.pairwise.paired_cosine_distances`
128+
- :func:`sklearn.metrics.pairwise.rbf_kernel` (see :ref:`device_support_for_float64`)
127129
- :func:`sklearn.metrics.r2_score`
128130
- :func:`sklearn.metrics.zero_one_loss`
129131

@@ -172,6 +174,8 @@ automatically skipped. Therefore it's important to run the tests with the
172174
pip install array-api-compat # and other libraries as needed
173175
pytest -k "array_api" -v
174176

177+
.. _mps_support:
178+
175179
Note on MPS device support
176180
--------------------------
177181

@@ -191,3 +195,17 @@ To enable the MPS support in PyTorch, set the environment variable
191195

192196
At the time of writing all scikit-learn tests should pass, however, the
193197
computational speed is not necessarily better than with the CPU device.
198+
199+
.. _device_support_for_float64:
200+
201+
Note on device support for ``float64``
202+
--------------------------------------
203+
204+
Certain operations within scikit-learn will automatically perform operations
205+
on floating-point values with `float64` precision to prevent overflows and ensure
206+
correctness (e.g., :func:`metrics.pairwise.euclidean_distances`). However,
207+
certain combinations of array namespaces and devices, such as `PyTorch on MPS`
208+
(see :ref:`mps_support`) do not support the `float64` data type. In these cases,
209+
scikit-learn will revert to using the `float32` data type instead. This can result in
210+
different behavior (typically numerically unstable results) compared to not using array
211+
API dispatching or using a device with `float64` support.

‎doc/whats_new/v1.6.rst

Lines changed: 36 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -43,7 +43,9 @@ See :ref:`array_api` for more details.
4343
- :func:`sklearn.metrics.pairwise.additive_chi2_kernel` :pr:`29144` by :user:`Yaroslav Korobko <Tialo>`;
4444
- :func:`sklearn.metrics.pairwise.chi2_kernel` :pr:`29267` by :user:`Yaroslav Korobko <Tialo>`;
4545
- :func:`sklearn.metrics.pairwise.cosine_similarity` :pr:`29014` by :user:`Edoardo Abati <EdAbati>`;
46-
- :func:`sklearn.metrics.pairwise.paired_cosine_distances` :pr:`29112` by :user:`Edoardo Abati <EdAbati>`.
46+
- :func:`sklearn.metrics.pairwise.euclidean_distances` :pr:`29433` by :user:`Omar Salman <OmarManzoor>`;
47+
- :func:`sklearn.metrics.pairwise.paired_cosine_distances` :pr:`29112` by :user:`Edoardo Abati <EdAbati>`;
48+
- :func:`sklearn.metrics.pairwise.rbf_kernel` :pr:`29433` by :user:`Omar Salman <OmarManzoor>`.
4749

4850
**Classes:**
4951

@@ -71,7 +73,7 @@ more details.
7173
:class:`ensemble.StackingRegressor` now support metadata routing and pass
7274
``**fit_params`` to the underlying estimators via their `fit` methods.
7375
:pr:`28701` by :user:`Stefanie Senger <StefanieSenger>`.
74-
76+
7577
- |Feature| :class:`compose.TransformedTargetRegressor` now supports metadata
7678
routing in its `fit` and `predict` methods and routes the corresponding
7779
params to the underlying regressor.
@@ -89,6 +91,19 @@ more details.
8991
passed to the underlying estimators via their respective methods.
9092
:pr:`28494` by :user:`Adam Li <adam2392>`.
9193

94+
- |Feature| :func:`model_selection.permutation_test_score` now supports metadata routing
95+
for the `fit` method of its estimator and for its underlying CV splitter and scorer.
96+
:pr:`29266` by :user:`Adam Li <adam2392>`.
97+
98+
Dropping support for building with setuptools
99+
---------------------------------------------
100+
101+
From scikit-learn 1.6 onwards, support for building with setuptools has been
102+
removed. Meson is the only supported way to build scikit-learn, see
103+
:ref:`Building from source <install_bleeding_edge>` for more details.
104+
105+
:pr:`29400` by :user:`Loïc Estève <lesteve>`
106+
92107
Dropping official support for PyPy
93108
----------------------------------
94109

@@ -126,6 +141,15 @@ Changelog
126141
on the input data.
127142
:pr:`29124` by :user:`Yao Xiao <Charlie-XIAO>`.
128143

144+
145+
:mod:`sklearn.datasets`
146+
.......................
147+
148+
- |Feature| :func:`datasets.fetch_file` allows downloading arbitrary data-file
149+
from the web. It handles local caching, integrity checks with SHA256 digests
150+
and automatic retries in case of HTTP errors. :pr:`29354` by :user:`Olivier
151+
Grisel <ogrisel>`.
152+
129153
:mod:`sklearn.discriminant_analysis`
130154
....................................
131155

@@ -142,6 +166,12 @@ Changelog
142166
by parallelizing the initial search for bin thresholds
143167
:pr:`28064` by :user:`Christian Lorentzen <lorentzenchr>`.
144168

169+
- |Enhancement| The verbosity of :class:`ensemble.HistGradientBoostingClassifier`
170+
and :class:`ensemble.HistGradientBoostingRegressor` got a more granular control. Now,
171+
`verbose = 1` prints only summary messages, `verbose >= 2` prints the full
172+
information as before.
173+
:pr:`28179` by :user:`Christian Lorentzen <lorentzenchr>`.
174+
145175
- |Efficiency| :class:`ensemble.IsolationForest` now runs parallel jobs
146176
during :term:`predict` offering a speedup of up to 2-4x on sample sizes
147177
larger than 2000 using `joblib`.
@@ -150,7 +180,7 @@ Changelog
150180

151181
- |Feature| :class:`ensemble.ExtraTreesClassifier` and :class:`ensemble.ExtraTreesRegressor` now support
152182
missing-values in the data matrix `X`. Missing-values are handled by randomly moving all of
153-
the samples to the left, or right child node as the tree is traversed.
183+
the samples to the left, or right child node as the tree is traversed.
154184
:pr:`28268` by :user:`Adam Li <adam2392>`.
155185

156186
:mod:`sklearn.impute`
@@ -195,6 +225,9 @@ Changelog
195225
estimator without re-fitting it.
196226
:pr:`29067` by :user:`Guillaume Lemaitre <glemaitre>`.
197227

228+
- |Fix| Improve error message when :func:`model_selection.RepeatedStratifiedKFold.split` is called without a `y` argument
229+
:pr:`29402` by :user:`Anurag Varma <Anurag-Varma>`.
230+
198231
:mod:`sklearn.neighbors`
199232
........................
200233

0 commit comments

Comments
 (0)
0