main
Move project directory from scikits.learn to sklearn
DOC: Correct many a/an usage mistakes (#32996)
FIX Explicitly disable array API dispatching for reference computation (#34378)
TST use global_random_seed in sklearn/utils/tests/test_arrayfuncs.py (#34158)
MNT move bitset from HGBT to utils (#33957)
API Use grid_values instead of pdp_values in partial_dependence (#25732)
MAINT cleanup utils.__init__: move chunking tools into dedicated submodule (#28516)
Fix linear svc handling sample weights under class_weight="balanced" (#30057)
Fix BLAS_Order.RowMajor import and similar in test_cython_blas with Cython 3.1 (#31301)
MNT replace is_pandas_df by narwhals (#34089)
MNT Use pyrefly instead of mypy for type-checking (#34527)
MNT: Simplify unknown handling in encoders/`_encode.py` (#34452)
TST Add common test for mixed namespace/device array inputs to estimators (#33525)
MNT upgrade black version (#28802)
Enable config setting `sparse_interface` to control sparray and spmatrix creation (#31177)
MNT add isort to ruff's rules (#26649)
MAINT Clean-up utils.__init__: move tests into corresponding test files (#28842)
MNT _convert_container "dataframe" -> "pandas" (#33956)
Revert "API make murmurhash3_32 private (#32103)" (#32131)
ENH add Array API to newton-cg in LogisticRegression (#34412)
CI Add Python 3.14 free-threaded wheels (#32079)
MNT Fix typos found by typos (#33916)
FIX Set default transform_input of Pipeline to ["X_val"] (#34263)
MNT Update internal sparse code to support both sparray and spmatrix (#30858)
FIX add missing random states and cloning in `test_response.py` (#33221)
Make the test suite itself thread-safe to be able to detect thread-safety problems with or without free-threading (#30041)
MNT remove get_columns/callable in set_output (#34079)
ENH Include Narwhals in `show_versions` (#34138)
FIX: Use intro sort for `simultaneous_sort` (2 way partitioning) (#33252)
CI Fix scipy-dev remaining failures due to scipy sparse DeprecationWarning (#34632)
MNT remove unnecessary cython.import_array (#28656)
PERF speedup classification_report by attaching unique values to dtype.metadata (#29738)
MAINT cleanup utils.__init__: move _print_elapsed_time into dedicated submodule (#28662)
FIX indexable with dataframes (#34605)