rapidsai/cudf v26.08.00 on GitHub

What's Changed

🚨 Breaking Changes

Enforce pandas 3.0 deprecations and subsequent related cuDF deprecations part 1 by @mroeschke in #20823
Enforce Series.interpolate deprecations by @mroeschke in #20905
Remove FutureWarning checks from series/test_pct_change by @mroeschke in #20904
Enforce Series.replace argument deprecations by @mroeschke in #20903
Enforce pandas 3 DataFrame/Series.first/last removal by @mroeschke in #20909
Enforce pandas 3 fillna deprecation by @mroeschke in #20912
Enforce pandas 3 timedelta frequency string deprecation by @mroeschke in #20915
Enforce pandas 3 concat with empties deprecation by @mroeschke in #20916
Enforce pandas 3 to_numeric(errors='ignore') deperecation by @mroeschke in #20920
Enforce pandas 3 date_range(periods=) deprecation by @mroeschke in #20921
Enforce pandas 3 resample deprecations by @mroeschke in #20923
Enforce pandas 3 TimedeltaIndex/DatetimeIndex deprecations by @mroeschke in #20924
Enforce pandas 3 future_stack argument switch by @mroeschke in #20926
Remove pandas 3.0 Series.getitem int positional deprecation by @mroeschke in #20930
Test pandas 3.0.0rc1 by @mroeschke in #20941
Change to_datetime to match pandas-3 by @galipremsagar in #20913
More misc cuDF classic test fixes for pandas 3 by @mroeschke in #21049
Switch default string dtype by @galipremsagar in #20890
Change CUDF_STRING_DTYPE to pandas.StringDtype for pandas 3.0 by @mroeschke in #21062
Adjust cuDF grouby unit tests for existing pandas object types in pandas 3 by @mroeschke in #21081
Fix StringDtype dask serialization in pandas 3 by @mroeschke in #21103
Avoid object dtype in ColumnAccessor.to_pandas_index for pandas 3 by @mroeschke in #21110
Move CategoricalDtype higher in cudf.dtype elif chain for pandas 3 by @mroeschke in #21129
Match groupby idxmin/max skipna=True pandas 3.0 behavior with all null groups by @mroeschke in #21189
Map numpy dtypes to new default string dtype in np_dtypes_to_pandas_dtypes for pandas 3 by @mroeschke in #21201
Fix test_string_char_types for pandas 3 by @mroeschke in #21210
Adjust more misc tests for pandas 3 by @mroeschke in #21228
Preserve StringDtype correctly in StringColumn.to_pandas for pandas 3 by @mroeschke in #21206
Update pandas pinnings in conda-forge recipes for pandas 3 by @mroeschke in #21302
Update cudf_pandas proxies for pandas 3 by @mroeschke in #21606
Change date_range time unit inferencing logic by @galipremsagar in #21694
Make copy-on-write the default for pandas3 by @vyasr in #21801
Fix multiple repr issues in cudf with pandas-3.0 by @galipremsagar in #21778
Fix replace API & pytests by @galipremsagar in #21938
Fix cudf.pandas pytest failures by @galipremsagar in #22137
Fix setitem dtype validation, NaT column inference, and boolean-dtype join by @galipremsagar in #22255
Reject pd.NA string-to-object casts in pandas-compatible mode by @galipremsagar in #22295
Align groupby apply/rolling defaults with pandas 3 by @galipremsagar in #22448
Align cudf find_common_type with pandas for empty/bool/datetime+timedelta inputs by @galipremsagar in #22466
Add validations in various code-paths by @galipremsagar in #22336
Merge pandas3 feature branch into main by @galipremsagar in #22518
Forward-merge release/26.06 into main by @madsbk in #22585
Remove deprecated .values_host property in cudf by @mroeschke in #22649
Remove deprecated nvtext::edit_distance_matrix by @davidwendt in #22644
Fix capture group handling in cudf::strings::findall by @davidwendt in #22408
Removed deprecated multi-pattern overload of replace_re by @davidwendt in #22639
Skip sum updates once SUM_WITH_OVERFLOW flag is set by @PointKernel in #22696
Remove deprecated strings::like string_scalar API by @davidwendt in #22678
Fix Categorical astype for nulls in pandas-compatible mode by @galipremsagar in #22652
[cudf_streaming] Using rapidsmpf spill unspill partition by @nirandaperera in #22810
Change behavior for regex class pattern [a-m-z] by @davidwendt in #22601
Fix groupby any/all on null-containing string columns by @galipremsagar in #22926
Align cudf indexing/setitem validation and alignment with pandas by @galipremsagar in #22912
refactor(streaming): flatten namespaces and rename to snake_case by @vyasr in #22909
[cudf_streaming] rapidsmpf backrefmixin API changes by @nirandaperera in #22911
Refactor packed metadata to use an explicit table header by @madsbk in #22951
Support skipna in groupby reductions (first/last, sum/prod/mean/median/min/max, idxmin/idxmax) by @galipremsagar in #22925
Make pylibcudf.Table.columns() return a tuple by @madsbk in #23040
Align DataFrame.reindex dtype and validation behavior with pandas by @galipremsagar in #22914
Match pandas suffix handling in merge (None/non-string suffixes, invalid-container TypeError) by @galipremsagar in #23059
Rapidsmpf logger API changes by @nirandaperera in #22972
Refactor OrderScheme to track multiple Ordering descriptions by @rjzamora in #22896
Return null for non-participating regex capture groups in strings::extract by @galipremsagar in #23123
wheels: build CUDA 13 wheels with latest CTK (13.3.0) by @jameslamb in #23174
Support explicit row counts for zero-column tables by @madsbk in #22765
Fix MultiIndex indexing/.loc to match pandas by @galipremsagar in #22966
Match pandas merge/join dtype, index and error semantics by @galipremsagar in #23060
Rapidsmpf backref API changes for Host and Pinned MR by @nirandaperera in #23128
Rewrite mixed inner/left/full join with post-filtering by @PointKernel in #23012
Fix DataFrame.stack level resolution, ordering, and dtype preservation by @galipremsagar in #23370
Rework approach to cudf-streaming bloom filter sizing by @wence- in #23067
Match pandas ordering, dtypes, and validation in unstack/_pivot by @galipremsagar in #23368
JNI bridge for the retention filter support for deletion vectors by @jihoonson in #23423

🐛 Bug Fixes

Add back Index.fillna by @mroeschke in #21002
Account of pandas.NamedAgg subclass change in pandas 3 by @mroeschke in #21023
Account for non-fixed pandas.offsets.Day in date_range for pandas 3 by @mroeschke in #21036
Fix columns validation logic of DataFrame.from_records by @mroeschke in #21034
Unify _is_null_host_scalar into is_na_like by @galipremsagar in #21079
Fix datetime binops issues in pandas-3.0 by @galipremsagar in #21172
Fix convert_dtypes with new StringDtype in pandas 3 by @mroeschke in #21200
Preserve StringDtype variants from pandas inputs for pandas 3 by @mroeschke in #21208
Re-enable dtype=category in pandas-3.0 branch by @galipremsagar in #21604
Fix all string type bugs by @galipremsagar in #21609
Fix astype CategoricalDtype cases. by @galipremsagar in #21627
Fix column type for empty columns by @galipremsagar in #21638
Fix interval_range type inference bugs by @galipremsagar in #21686
Fix Index.intersection for categorical types by @galipremsagar in #21723
Fix string udf failures by @galipremsagar in #21727
Fix join bug that shows up in Index.difference by @galipremsagar in #21733
Fix all binops failures by @galipremsagar in #21738
Use reduction_axis for dataframes by @vyasr in #21743
Fix astype(category) for empty column scenarios by @galipremsagar in #21741
Fix all explode failures by @galipremsagar in #21766
Fix orc pytest failures by @galipremsagar in #21847
Fix failures in rolling pytests by @galipremsagar in #21846
Fix all Series.astype bugs by @galipremsagar in #21860
Fix numpy ufunc return type by @galipremsagar in #21876
Fix dtype issues in test_between.py by @galipremsagar in #21887
Fix isin test failures by @galipremsagar in #21886
Fix string.to_pandas for nullable & arrow_type cases by @galipremsagar in #21890
Fix concat & union bugs by @galipremsagar in #21889
Fix all test_iloc.py failures by @galipremsagar in #21872
Fix all failures in test_constructors.py by @galipremsagar in #21905
Fix all string related failures by @galipremsagar in #21939
Fix Parquet pytest failures by @galipremsagar in #21967
Switch default time resolutions for DatetimeIndex & TimedeltaIndex by @galipremsagar in #21997
Fix add_prefix & add_suffix pytests by @galipremsagar in #22000
Fix all json test failures by @galipremsagar in #21998
Fix runtime warnings in apply pytests by @galipremsagar in #22017
Fix string column to_arrow to generate large strings when needed by @galipremsagar in #21891
Fix type inconsistencies in pytests by @galipremsagar in #22016
Fix replace pytest by @galipremsagar in #22019
Fix parquet metadata failures in pytests by @galipremsagar in #22020
Fix copy-on-write failures related to astype by @galipremsagar in #22022
Fix str reductions to match pandas3 by @galipremsagar in #22026
Fix right-join column ordering by @galipremsagar in #22002
[pandas3] Fix test_categorical_series_with_nan_repr for pandas 3 nan_as_null change by @galipremsagar in #22066
[pandas3] Fix test_diff_many_dtypes for boolean Series diff null representation by @galipremsagar in #22065
[pandas3] Fix test_timedelta_astype_unicode_dtype_pandas_compat for pandas 3 StringDtype by @galipremsagar in #22064
[pandas3] Fix test_categorical_empty: update categories dtype assertion by @galipremsagar in #22062
[pandas3] Fix test_rename_shallow_copy: use np.shares_memory instead of identity check by @galipremsagar in #22060
[pandas3] Fix test_groupby_pct_change_empty_columns for pandas 3 dtype inference by @galipremsagar in #22059
[pandas3] Fix two regressions in Series.describe() by @galipremsagar in #22058
[pandas3] Fix test_alltypes_plain_avro timestamp conversion for pandas 3 by @galipremsagar in #22061
[pandas3] Add _CachedAccessor to replace removed pandas.core.accessor.CachedAccessor by @galipremsagar in #22056
Update memory usage test by @galipremsagar in #22080
[pandas3] Fix float-category NaN handling in CategoricalColumn.to_pandas() by @galipremsagar in #22057
Skip pandas failure cases in ewm by @galipremsagar in #22082
Fix sort_values when return_indexer=True by @galipremsagar in #22098
Fix pandas 3 compatibility bugs in index, groupby, and dask_cudf tests by @galipremsagar in #22101
Make cudf.pandas pandas tests able to run by @galipremsagar in #22140
Fix 12k pandas test suite failures by @galipremsagar in #22158
Allow SeriesApplyKernel to support pd.StringDtype in pandas 3 by @mroeschke in #22177
Fix datetime & timedelta failures in pandas test suite by @galipremsagar in #22174
Fix return types for rank by @galipremsagar in #22180
Fix regression in eval frame fetching by @galipremsagar in #22221
Propagate writes to internal or private properties to slow objects only by @galipremsagar in #22231
Fix cudf.pandas view semantics across fast/slow conversions and add array to datetimelike ExtensionArray proxies by @galipremsagar in #22259
Fix cudf.pandas pickle round-trips for proxy classes, DatetimeTZ columns, and NaT by @galipremsagar in #22297
Add proxy types for missing offsets by @galipremsagar in #22314
Fix failures in masked array pandas unit tests under cudf.pandas by @galipremsagar in #22296
Fix StataReader proxy type by @galipremsagar in #22322
Fix MultiIndex getitem by @galipremsagar in #22326
Define attributes that are not discoverable by dir(cls) by @galipremsagar in #22330
Drop invalidated frequency in DatetimeIndex by @galipremsagar in #22328
Fix to_period parameter by @galipremsagar in #22329
Expose from_arrow on masked extension dtype proxies by @galipremsagar in #22373
Fix datetime resolution for empty data inputs by @galipremsagar in #22363
Expose additional proxy attributes for IntervalArray, Styler, and _MethodProxy by @galipremsagar in #22374
Fix tz aware binops by @galipremsagar in #22243
Implement min_count for groupby reductions by @galipremsagar in #22372
Fix datetime tz + categorical creation by @galipremsagar in #22278
Expose DateOffset instance attributes on cudf.pandas proxy by @galipremsagar in #22447
Fix np.ufunc(Index, Series) dispatch and np.matmul scalar return type by @galipremsagar in #22444
Fix fillna in binop by @galipremsagar in #22442
Fix is_numeric_dtype for Index by @galipremsagar in #22438
Preserve StringDtype storage and na_value in get_dtype_of_same_kind by @galipremsagar in #22289
Raise TypeError for numeric column arithmetic with pd.NaT by @galipremsagar in #22451
Make StringArray proxy inherit from NumpyExtensionArray by @galipremsagar in #22457
Fix return types of quantile by @galipremsagar in #22433
Preserve extension dtypes in groupby reductions and exclude grouping-key columns by identity by @galipremsagar in #22369
Fix several to_datetime edge cases for pandas 3 compatibility by @galipremsagar in #22446
Fix arrow types related failures in cudf.pandas by @galipremsagar in #22194
Implement groupby all/any via bool-coercion + min/max by @galipremsagar in #22371
Fix AssertionError: DataFrame.columns are different failures in cudf.pandas by @galipremsagar in #22351
Fix factorize for pandas 3 compatibility by @galipremsagar in #22445
Fix empty series dtype by @galipremsagar in #22443
Implement groupby sum on StringDtype columns as per-group concatenation by @galipremsagar in #22370
Preserve pandas-nullable dtypes through ufuncs and empty value_counts by @galipremsagar in #22461
Fix all datetime type inconsistencies with extension types by @galipremsagar in #22409
Serialize engine config in new pdsh benchmark CLI by @TomAugspurger in #22365
Fix unsafe, missing close() of HostColumnVectors by @mythrocks in #22432
Skip flaky test_make_spill_function on pinned-memory allocation failure by @Matt711 in #22551
Fix JSON token count estimate to account for the delimiter in recovery mode by @shrshi in #22559
Fix AST expression evaluator handling of intermediate values for fixed-point types by @davidwendt in #22512
Remove arrow_array from NumPy ndarray cudf.pandas proxy by @galipremsagar in #22617
Expose per-column had_schema_mismatch diagnostic from JSON reader by @wjxiz1992 in #22450
Fix Arrow device string view variadic buffers by @0ax1 in #22620
[BUG] Parquet reader: treat LIST-annotated REPEATED group as the list, not a stub by @wjxiz1992 in #22597
Fix invalid use of data_type constructor in row_ir logic by @davidwendt in #22676
Preserve single-field unannotated repeated groups in parquet schema sanitization by @wjxiz1992 in #22567
Expose __array_ufunc__ on NumpyExtensionArray cudf.pandas proxy by @galipremsagar in #22607
Expose from_arrow on DatetimeTZDtype cudf.pandas proxy by @galipremsagar in #22608
Fix DataFrame mask Series condition alignment by @aryansri05 in #22701
fix(cudf.pandas): fix Excel/openpyxl test failures and add xlsxwriter dependency by @vyasr in #22722
Fix to_numpy object null handling by @aryansri05 in #22621
Match pandas dtype semantics in groupby quantile by @galipremsagar in #22750
Make IntervalDtype construction pandas-compatible by @galipremsagar in #22731
Remove cudf-polars CI timeouts by @TomAugspurger in #22768
Fix scalar as column view for null scalars by @mhaseeb123 in #22773
Increase timeouts for slower cudf-polars tests by @TomAugspurger in #22780
Enable chunked row masks in hybrid scan reader by @mhaseeb123 in #22716
Fix Parquet page header decoder to allow empty dict pages by @mhaseeb123 in #22777
declare nvidia-nvjitlink dependency in 'libcudf' wheels by @jameslamb in #22785
Fix replace_with_backrefs copy remainder logic for empty pattern match by @davidwendt in #22711
Fix ArrowDtype integer arithmetic to match pandas pyarrow semantics by @galipremsagar in #22633
Fix zero-count regex quantifier expansion by @fallintoplace in #22800
Build cudf_streaming wheel against this CI run's pylibcudf by @TomAugspurger in #22819
Fix groupby.apply columns axis name for row-like UDF results by @galipremsagar in #22809
Use (un)spill_partitions from RapidsMPF by @pentschev in #22837
Fix cudf.pandas DateOffset equality and empty-frame column dtype preservation by @galipremsagar in #22806
Fix MultiIndex.join to join on all common levels by @galipremsagar in #22827
Let NVRTC choose automatic PCH directory by @bdice in #22812
[cudf-polars] Sort ORDER BY keys by column, not alias by @madsbk in #22781
Fix cudf_polars benchmarking summarizing mean times by @mroeschke in #22845
Fix masked dtype equality in cudf.pandas so real_dtype == proxy_dtype holds by @galipremsagar in #22842
Reduce size of input dataframe in test_parallel_scan by @TomAugspurger in #22835
Fix findall logic handling for empty match regex patterns by @davidwendt in #22714
Reduce setup runtime of multi_orderby_range rolling benchmarks by @davidwendt in #22869
Skip polars tests relying on newer sqlite features by @TomAugspurger in #22870
Fix ARROW_FLAG_NULLABLE flag usage in cudf interop by @davidwendt in #22828
Pin pytest-cases version to fix nightly build errors by @davidwendt in #22888
Drop dynamic predicates from logical plan by @Matt711 in #22820
Fix group-key columns in SeriesGroupBy list-agg and any/all with as_index=False by @galipremsagar in #22813
Preserve object dtype for string columns when constructing cudf DataFrame from pandas by @galipremsagar in #22831
Reduce concurrency of cudf-polars tests by @TomAugspurger in #22894
Fix groupby and single-column to_numpy pandas-compatibility bugs by @galipremsagar in #22783
Preserve categorical column index in DataFrame.describe and handle pure iterators in cudf.pandas fallback by @galipremsagar in #22821
[BUG] Fix conditional linking of RTCX embed dependencies & Allow null fragment names by @lamarrr in #22898
Allow IR fallback to handle an ErrorNode by @mroeschke in #22855
Fix type annotations in pylibucdf hybrid scan by @TomAugspurger in #22118
Destroy RTCX cache before teardown by @fallintoplace in #22843
create_string_column bug fix by @abigalekim in #22942
Fix categorical replace no-op remapping by @anthonyguerrera in #22436
Enable cudf.pandas test_numpy_ufuncs_basic via tolerant Index comparison by @galipremsagar in #22924
Make to_string match pandas (drop repr-style truncation/footer) by @galipremsagar in #22929
Fix cudf.to_numeric to match pandas semantics by @galipremsagar in #22934
Protect against integer overflow in groupby_actor by @rjzamora in #22950
Skip code coverage where dynamic predicates are dropped by @Matt711 in #22973
Work around pola-rs/polars#23214 in streaming dataframe scan by @wence- in #23007
Fix cudf.pandas --line-profile clobbering __file__ by @galipremsagar in #23017
Pin pandas to exclude 3.0.4 (segfault constructing pd.Timedelta on Python 3.14) by @galipremsagar in #23022
Fix stream ordering bug in join of expression-based keys by @wence- in #23024
Fix DataFrame and GroupBy.value_counts to match pandas ordering, dropna, and categorical handling by @galipremsagar in #22903
Fix Distinct expression lowering by @rjzamora in #22970
Add bounds checking for AST column_reference out-of-range indices by @utkarshparekh in #23021
Implement DataFrame.count(axis=1) on the GPU by @galipremsagar in #23016
Fix ptr null check in packed_data_from_cudf_packed_columns by @Matt711 in #23038
Fix cudf-java native_jstring to emit standard UTF-8 for supplementary characters and embedded NULs by @igorpeshansky in #23018
Fix cudf.pandas Excel reader failures (empty-column dtype, string offset width, proxy monkeypatch) by @galipremsagar in #22927
Keep Nanoarrow ArrowArrayStream chunks alive during cudaMemcpyBatchAsync by @mroeschke in #23009
Support host buffers (BytesIO) in cudf.read_text by @nethum529 in #23032
Fully initialize PageInfo before sort_pages copies it by @pramodsatya in #22957
Declare return type on device lambdas queried in host code by @PointKernel in #23046
Add exception for unsupported decimal128 AST operations by @davidwendt in #22982
Fix explain() failing on UnaryFunction with multiple children by @Matt711 in #23050
Raise parquet partitions_info validation errors by @fallintoplace in #22801
Re-derive cudf.pandas intermediate proxies when their parents change by @galipremsagar in #22998
Add Numba pin to cuDF wheel build requirements by @bdice in #23084
Return np.intp from argsort to match numpy/pandas by @galipremsagar in #23061
Fix pinned-host spill reservations for packed table alignment by @pentschev in #23088
Enable cudf_streaming distributed (MPI/UCXX) tests in CI by @nirandaperera in #22954
Fix DataFrame.rename for an unnamed MultiIndex level by @galipremsagar in #23062
Fix rtcx embed target-level dependencies under the Makefiles generator by @galipremsagar in #23106
fix(pylibcudf): accept inf/-inf as valid FLOAT32 scalar values by @nethum529 in #23099
Match pandas semantics in groupby.rolling and groupby.apply result construction by @galipremsagar in #23122
Hybrid scan reports correct number of input RGs by @mhaseeb123 in #23052
Fix uninitialized read for empty segments in segmented top-k by @ttnghia in #23109
Aggregate category values instead of codes in window operations by @galipremsagar in #23120
Sum interleave benchmark byte counts across all columns by @fallintoplace in #23103
Fix type annotation for CachedParquetInfo.size by @TomAugspurger in #23142
Raise TypeError for NaT in where() on masked dtypes by @galipremsagar in #23107
Fix Scan projection with pandas index by @rjzamora in #23135
Match pandas semantics in DataFrame.select_dtypes by @galipremsagar in #23116
Fix to_datetime nanosecond inference and DatetimeIndex tz wall-time interpretation by @galipremsagar in #23141
Return value_counts groups in order of first appearance by @galipremsagar in #23140
Implement factorize for MultiIndex by @galipremsagar in #23139
Fix cudf.pandas datetimelike plotting failures (canonical freq offsets + Period/offset pickling) by @galipremsagar in #23054
Use rapids_test_add() again by @KyleFromNVIDIA in #23181
Add error checking to format inference by @vyasr in #23172
Set RUN_SERIAL property on streaming tests by @KyleFromNVIDIA in #23188
Handle Arrow bool masks in DataFrame getitem by @dttutty in #23195
Supply stream and mr to column synthesizers by @mhaseeb123 in #23209
Fix out-of-bounds access in segmented top-k with partial offsets by @ttnghia in #23105
Remove duplicate return in ndsh apply_inner_join by @fallintoplace in #23104
Fix cuco::extent type for comparator_helper key_set in distinct_count by @davidwendt in #23236
Fix string reductions and isin across string dtype flavors by @galipremsagar in #23214
fix(pre-commit): normalize SPDX handling by @bdice in #23228
Fix cudf.pandas arithmetic test failures (pow null identities, RangeIndex range-preserving binops, Index proxy array priority) by @galipremsagar in #23176
Fix benchmark generate_input create_ascii_string_column function signature by @davidwendt in #23239
Raise OutOfBoundsDatetime for datetime conversions that exceed the target unit's range by @galipremsagar in #23216
Fix cudf.pandas series accessor failures across numpy/nullable/arrow dtypes by @galipremsagar in #23056
Fix Series.isin dtype and value semantics for masked (nullable) dtypes by @galipremsagar in #23057
Replace thrust::reduce_by_key with cub::DeviceReduce::ReduceByKey in tdigest generation by @PointKernel in #23204
Change cuco::extent type usage from size_type to size_t by @davidwendt in #23259
Fix DataFrame.quantile and Series.quantile to match pandas by @galipremsagar in #23058
Fix data_type ctor calls in the row_ir get_return_type function by @davidwendt in #23237
Fix groupby size dtype for masked series and apply group order for sort=False by @galipremsagar in #23260
Rewrite GroupBy.nth as a pandas-compatible positional row filter by @galipremsagar in #23257
Keep the dtype family in GroupBy.rank results by @galipremsagar in #23258
Fix stale num_rows argument breaking mypy in cudf-polars duplicated-output path by @galipremsagar in #23303
Pin pyarrow<24 in the cudf and pylibcudf conda recipes by @galipremsagar in #23319
Fix cudf.pandas isinstance checks for unproxied pandas subclasses and wrap groupby resamplers by @galipremsagar in #23273
Fix GroupBy.apply result assembly, UDF closure side effects, and empty-frame dtypes by @galipremsagar in #23272
Avoid leaving multi-rank execution in a bad state in cudf-polars tests by @wence- in #23340
Reconcile Index.union dtypes before the empty-operand short-circuits by @galipremsagar in #23318
Support the limit parameter in GroupBy.ffill and bfill by @galipremsagar in #23302
Stop forward-filling in GroupBy.pct_change by @galipremsagar in #23301
Return index labels from GroupBy idxmin/idxmax in agg and transform by @galipremsagar in #23298
Match pandas semantics for null reduction results and mode ordering by @galipremsagar in #23328
Fix incorrect DECIMAL128 groupby sums on Blackwell with CUDA 13.0 by @PointKernel in #23229
Fix null_precedence access in preprocessed_table::create by @davidwendt in #23238
Route GroupBy.transform size/cumcount/ngroup to their group-level implementations by @galipremsagar in #23297
Recognize named-aggregation lambdas as scans in GroupBy by @galipremsagar in #23300
Add timeout stacktrace utility for cudf-polars tests, remove pytest-timeout by @wence- in #23332
Stop mutating the source column dtype in equivalent-type numerical casts by @galipremsagar in #23364
Mirror cudf.pandas class-level monkeypatches onto the real type by @galipremsagar in #23001
Restore the original index in GroupBy transform broadcasts and ngroup by @galipremsagar in #23296
Honor level and sort_remaining in sort_index(axis=1) by @galipremsagar in #23367
Fix lazy quantifier priority handling in Glushkov regex engine by @davidwendt in #23381
Keep MultiIndex columns in GroupBy.agg for MultiIndex-column sources by @galipremsagar in #23366
Preserve masked integer dtypes in GroupBy cumsum/cumprod by @galipremsagar in #23299
Use new rapidsmpf utilities to handle cancellation in bloom filter awaitables by @wence- in #23293
RMM Library Linking Error Bug Fix by @abigalekim in #23341
Fix Parquet stats pinned host copy lifetime by @aryansri05 in #23219
Add cuFile wheel dependencies for Python tests by @bdice in #23392
Include sources.jar and javadoc.jar in the Java Maven repo output by @paul-aiyedun in #23389
Preserve MultiIndex column fidelity through ColumnAccessor round trips by @galipremsagar in #23365
[BUG] Fix ORC chunked writer root row statistics by @wjxiz1992 in #23118
Fix flaky cudf-polars Quent integration test by @TomAugspurger in #23482

📖 Documentation

Fix misleading Javadoc for Aggregation.nth(int) by @piotrrzysko in #22486
Remove wrong example from the docs by @findepi in #22575
Reorganize cuDF documentation to highlight all projects by @mroeschke in #22352
Drop caveat about strict channel priority by @jakirkham in #22688
Remove unneeded cudf classic documentation pages by @mroeschke in #22693
Fix duplicate and truncated copyright header in lists/detail/copying.hpp by @PointKernel in #22719
Fix cudf-polars MemoryResourceConfig example rendering by @bdice in #22751
Update cuDF Python user guides to current/pandas 3 behaviors by @mroeschke in #22720
Overhaul cudf Python developer docs by @mroeschke in #22689
Add blog links to appropriate cpp/examples readme files by @davidwendt in #22799
Document and test Parquet row-group output ordering for read_parquet by @qbacpey in #22733
Fix documentation for cudf::lists::sort_lists and the corresponding Java binding listSortRows by @ttnghia in #22916
Fix Python docs source links by @bdice in #22791
Clarify list sorting docs and replace stale disabled sort_lists test by @ttnghia in #22921
Document cudf.pandas Profiler usage in Python scripts and CLI line profiling by @gmartini2000 in #21991
Update docs to use NVIDIA Sphinx theme by @bdice in #23035
Add instructions on running PDS-H by @Matt711 in #23025
Fix doxygen to include appropriate include path names in generated output by @davidwendt in #23080
Add instructions on running PDS-H for pandas by @Matt711 in #23087
Add file descriptions for doxygen output by @davidwendt in #23083
Clarify C++ review approval requirements by @bdice in #23092
Align cuDF documentation branding by @bdice in #23079
Update outdated overflow/size guidance in libcudf developer guide by @PointKernel in #23127
Fix doxygen file descriptions by moving file/brief out of addtogroup blocks by @davidwendt in #23125
Document the include-what-you-use convention in developer and review guidelines by @PointKernel in #23245

🚀 New Features

Write variable bit-width keys for Parquet dictionary encoded pages by @mhaseeb123 in #22279
[FEA] ANSI SQL Operator JIT Support (1) : Refactor ROW IR by @lamarrr in #22511
Forward merge release/26.06 into main by @KyleFromNVIDIA in #22603
[FEA] LTO IR Support (1) - Introduce LibRTCX by @lamarrr in #21625
Extend SUM_WITH_OVERFLOW reduction to signed integers and decimals by @PointKernel in #22465
[FEA] LTO IR Support (2) - Cleanup CUDF's JIT Sources by @lamarrr in #22641
Rename build/probe to right/left in filtered_join by @PointKernel in #22464
Parquet utilities to fetch footer and page index buffers from multiple sources by @mhaseeb123 in #22613
Add hybrid scan multifile reader basics by @mhaseeb123 in #22616
[FEA] ANSI SQL Operator JIT Support (2) : Implement Operator Library by @lamarrr in #22514
Support reading from file:// URIs in cudf-polars by @Matt711 in #22749
Create new libcudf_streaming C++ library by @vyasr in #22747
Add multifile row group pruning with stats and byte ranges by @mhaseeb123 in #22715
Add Python/Cython bindings for libcudf_streaming by @vyasr in #22760
[FEA] LTO IR Support (3) - Replace JITIFY usage with LIBRTCX by @lamarrr in #22654
[BUG][FEA] Implement Static Linking for RTCX's Dependencies by @lamarrr in #22823
Add filter_join_indices_output_size by @PointKernel in #22694
Expose max_dictionary_size and dictionary_policy in Java/JNI Parquet writer bindings by @thirtiseven in #22656
Add multi-column ORDER BY support to range rolling windows by @pramodsatya in #22627
Prepare librtcx for standalone usage by @vyasr in #22856
Field extraction for Parquet VARIANT columns (scalar + nested objects) by @vuule in #22416
Add hybrid scan multifile single-step materializers by @mhaseeb123 in #22793
Add array-index ("[N]") support to VARIANT field extraction by @vuule in #22877
Add row mask builder APIs for multifile hybrid scan by @mhaseeb123 in #22792
Add multifile hybrid scan pass construction API by @mhaseeb123 in #22794
[MLIR] Scaffold MLIR based UDF backend by @brandon-b-miller in #22766
Split libcudf-streaming tests and benchmarks into a dedicated conda package by @pentschev in #22871
Added Concurrent Polynomial Transform Benchmark by @lamarrr in #19199
Support converting column to a python list in pylibcudf by @Matt711 in #19446
[FEA] ANSI SQL Operator JIT Support (3) : Refactor operator library by @lamarrr in #22836
Support cudf-polars epoch by @brandon-b-miller in #18443
Support cudf-polars total_xxx datetime extraction methods by @brandon-b-miller in #18171
Add SUM_OVERFLOW in sort groupby by @PointKernel in #22832
Skill to compare performance of a branch or PR with main by @mhaseeb123 in #22725
Add dynamic join key prefilter planning by @pentschev in #22995
Add hybrid scan multifile two-step materializers by @mhaseeb123 in #22833
Add hybrid scan multifile chunked materializers by @mhaseeb123 in #23013
[MLIR] MaskedType: core type, constructor, accessors by @brandon-b-miller in #22884
Prepend source index column in Parquet reader by @mhaseeb123 in #22879
[FEA] Improve cuDF's Kernel Cache Directory Resolution by @lamarrr in #22975
[FEA] ANSI SQL Operator JIT Support (4) : Implement IR for ANSI and other JIT Extensions by @lamarrr in #22602
Pin Polars<1.43 by @Matt711 in #22993
[FEA] Support input of dictionary-encoded types in Transforms by @lamarrr in #21552
[FEA] LTO IR Support (4) - Implement LTO Transform Kernels by @lamarrr in #22680
Add config option for JIT parquet filtering by @Matt711 in #21460
[MLIR] MaskedType: NA handling by @brandon-b-miller in #22885
Implement item aggregation for LazyFrame.pivot by @rjzamora in #22941
Support fused arithmetic expressions in cudf-polars by @Matt711 in #23136
JNI support for multiple order-by columns for RANGE-based window functions by @mythrocks in #22778
Add Java Variant extraction support by @nartal1 in #23069
Add Java Bindings for Hybrid Scan Parquet Reader by @paul-aiyedun in #22456
Add cuDF Java option for loading libraries from a user provided directory by @paul-aiyedun in #22592
Propagate OrderScheme metadata through aliases and timestamp truncation by @rjzamora in #23148
Prepend row index column in Parquet reader by @mhaseeb123 in #23077
Add memory_resources and reusable validation foundations by @bdice in #23028
Select Parquet columns by field ID by @mhaseeb123 in #22955
Prepend source and row index columns in Hybrid scan reader by @mhaseeb123 in #22878
Add Glushkov regex fast-path to libcudf by @lingyany-nv in #21936
Add memory resource parameter to cudf::hash_join constructors by @PointKernel in #23223
Add adjust_ordering utility by @rjzamora in #22628
Support non-Col shuffle keys in streaming Join (version 2) by @rjzamora in #23262
Add fixed-size rolling window support to cudf-polars by @tolleybot in #21964
Add cuDF JAR build support for all Maven classifiers by @paul-aiyedun in #23261
Add plan optimizer pass that extracts join keys and uses them to construct pre-filters for inner joins by @pentschev in #22996
Experimental Parquet reader supports retention vectors by @mhaseeb123 in #23402
Add array indexing support to Parquet variant field extraction by @vuule in #22895
Upgrade to nvcomp 5.3.0.16 by @robertmaynard in #23456

🛠️ Improvements

Trim CI temporarily for pandas3 feature branch by @galipremsagar in #20892
Ignore UserWarning from pandas 3.0 about obj.round on temporal types by @mroeschke in #20900
Remove expect_warning_if in series/test_np_ufuncs.py by @mroeschke in #20899
Remove unneeded pandas 3 FutureWarning filters from test_binops.py/test_apply.py by @mroeschke in #20917
Remove expected FutureWarning in test_date_range_raise_unsupported for pandas 3 by @mroeschke in #20931
Remove most pandas 2.x version checks for pandas 3.0 by @mroeschke in #20932
Remove more pandas<3 version checks in cuDF by @mroeschke in #20948
Drop warnings in test_reductions.py by @galipremsagar in #20911
Modify/Remove cuDF classic pandas-2 related xfails by @mroeschke in #20988
Modify unit tests that use deprecated copy keyword in pandas 3 by @mroeschke in #21020
Fix miscellaneous cuDF test failures from pandas 3 pt 2 by @mroeschke in #21005
Fix miscellaneous cuDF test failures from pandas 3 by @mroeschke in #21001
Implement .groups return dictionary keys FutureWarning for pandas 3 by @mroeschke in #21021
Chage more misc cuDF classic unit tests for pandas 3 by @mroeschke in #21031
Fix DatetimeIndex docstrings for pandas 3 by @mroeschke in #21035
Fix Numba row udf compat with pandas.StringDtype by @mroeschke in #21063
Adjust cuDF unit tests for existing pandas object types in pandas 3 by @mroeschke in #21080
Bump to pandas-3.0rc2 by @galipremsagar in #21097
Adjust parquet unit tests for existing pandas object types in pandas 3 by @mroeschke in #21109
Adjust reset_index and categorial tests for existing pandas object types in pandas 3 by @mroeschke in #21111
Adjust string accessor, setitem, constructor tests for existing pandas object types in pandas 3 by @mroeschke in #21127
Adjust json, csv tests for existing pandas object types in pandas 3 by @mroeschke in #21128
Fix pandas3 related test failures in test_fillna by @galipremsagar in #21142
Avoid asserting object in test_csv for pandas 3 by @mroeschke in #21150
Fully assert the result for test_join_multi in pandas 3 by @mroeschke in #21149
Adjust categorical tests for existing pandas object types in pandas 3 by @mroeschke in #21148
Switch to release version of pandas 3.0 by @galipremsagar in #21121
Fix pytest failures in test_explode and test_repr by @galipremsagar in #21158
Adjust misc tests for existing pandas object types in pandas 3 by @mroeschke in #21171
Fix most tests that raise a Pandas4Warning by @mroeschke in #21167
Adjust more misc tests for existing pandas object types in pandas 3 by @mroeschke in #21193
Fix test_index_append_error for pandas 3 by @mroeschke in #21211
Fix docstrings for pandas 3 by @mroeschke in #21209
Forward merge main into pandas3 by @Matt711 in #21215
Forward merge main into pandas3 by @Matt711 in #21233
Forward merge main into pandas3 by @Matt711 in #21288
Bump pandas 3 branch's latest version to 3.0.1 by @mroeschke in #21593
Fix some minor, pandas 3 branch backport issues by @mroeschke in #21595
Fix strftime callers in pandas3 by @galipremsagar in #21624
Fix groupby value_counts pytest by @galipremsagar in #21689
Fix all udf pytests by @galipremsagar in #21731
Fix all pytest failures in test_apply.py by @galipremsagar in #21835
Fix pandas failures in fillna tests by @galipremsagar in #21830
Update test_attributes for new better cudf/pandas alignment by @vyasr in #21833
Use arrow type to fix test issue by @vyasr in #21878
Remove now unnecessary env var flags by @vyasr in #21877
Fix type inconsistencies in diff pytests by @galipremsagar in #21893
Update loc pytests by @galipremsagar in #21983
Fix reduction test failures by @galipremsagar in #21986
Bump latest pandas to 3.0.2 by @mroeschke in #21987
Avoid pytz in test_holidays_within_dates by @mroeschke in #22139
Test different pandas option in test_options_mode for pandas 3 by @mroeschke in #22160
Pass writable numpy array to pytorch third party test with pandas 3 by @mroeschke in #22161
Drop all xpassing tests in pandas-3 by @galipremsagar in #22163
Fix custreamz tests for pandas 3 by @mroeschke in #22195
Fix some DatetimeIndex doctests for pandas 3 by @mroeschke in #22198
Adjust pylibcudf test_read_csv_parse_options for pandas 3 large string by @mroeschke in #22200
Fix more misc cuDF unit tests for pandas 3 by @mroeschke in #22201
Fix cudf interval test for pandas 3 and numpy < 2 by @mroeschke in #22205
Disable csv tests temporarily & fix dask_cudf failure by @galipremsagar in #22203
Remove unnecessary older pandas checks in pandas 3 by @mroeschke in #22215
Change pandas links from version/2.3.3 to stable by @mroeschke in #22218
Move skippable pandas tests in run-pandas-test.sh to conftest-patch.py by @mroeschke in #22260
Remove pandas tests that no longer exist by @galipremsagar in #22262
Implement Flags in cudf by @galipremsagar in #22225
Consolidate all pandas test skips to new NODEIDS_TO_SKIP dictionary by @mroeschke in #22282
Drop passing tests from conftest-patch.py by @galipremsagar in #22298
Consolidate all pandas test to NODEIDS_THAT_FAIL dictionary by @mroeschke in #22311
Add remaining pandas test failures to NODEIDS_THAT_FAIL for pandas 3 by @mroeschke in #22339
Remove dtype workaround in Column.create by @galipremsagar in #22396
Add ability to set freq on DatetimeIndex by @galipremsagar in #22440
Bump pandas to 3.0.3 by @mroeschke in #22484
Handle all dtypes in convert_dtypes by @galipremsagar in #22202
Update conftest-patch with latest xfails by @galipremsagar in #22489
Enable full ci for pandas3 by @galipremsagar in #22502
Use cudf::tables_equal API in libcudf examples by @mhaseeb123 in #22428
Add additional AI review guidelines for pylibcudf/cudf_polars by @mroeschke in #22340
Require CMake 4.0 by @KyleFromNVIDIA in #22492
Bump run_cudf_polars_pytests timeout by @TomAugspurger in #22505
Add Node.get_stable_plan_id to cudf-polars by @TomAugspurger in #22513
Capture command line on pds-h benchmark runner by @TomAugspurger in #22508
Fix mypy issue by @wence- in #22549
Add nvtx annotations to more cudf-polars IO by @TomAugspurger in #22573
Reduce some input data sizes for cudf_polars unit tests by @mroeschke in #22562
pylibcudf read_parquet accepts pre-materialized FileMetadata by @TomAugspurger in #22587
Add compression type axis to the parquet reader benchmarks by @qbacpey in #22471
Excercise assert_collect_raises with engine fixture by @mroeschke in #22563
Require engine argument in assert_ir_translation_raises in cudf_polars tests by @mroeschke in https://github.com/rapidsai/cudf/pull/22565
Pass memory resource to exec_policy_nosync in remaining modules by @bdice in https://github.com/rapidsai/cudf/pull/22045
Pass engine everywhere in assert_sink_ir_translation_raises by @mroeschke in https://github.com/rapidsai/cudf/pull/22609
Add pylibucdf bindings for FileMetadata.row_groups by @TomAugspurger in https://github.com/rapidsai/cudf/pull/22599
Expose libcudf tables_equal in pylibcudf by @mroeschke in https://github.com/rapidsai/cudf/pull/22611
Use CCCL overflow utils instead of custom overflow logic by @PointKernel in https://github.com/rapidsai/cudf/pull/22593
Improve devcontainer startup times by @trxcllnt in https://github.com/rapidsai/cudf/pull/22594
Reduce mutex scoped region in parquet io utilities by @mhaseeb123 in https://github.com/rapidsai/cudf/pull/22550
Clone rapidsmpf before devcontainer env creation by @bdice in https://github.com/rapidsai/cudf/pull/22579
Minimize Dask resource acquisition in cudf_polars tests by @mroeschke in https://github.com/rapidsai/cudf/pull/22646
Convert conftest-patch.py to a pytest plugin for pandas testing by @mroeschke in https://github.com/rapidsai/cudf/pull/22643
Clean up Polars unit tests setup by @mroeschke in https://github.com/rapidsai/cudf/pull/22605
Use nvcompGetStatusString when available by @vuule in https://github.com/rapidsai/cudf/pull/22566
Add LZ4 to ORC/Parquet IO benchmark compression axes by @vuule in https://github.com/rapidsai/cudf/pull/22564
Add row group size/stripe options to Parquet and ORC benchmarks by @qbacpey in https://github.com/rapidsai/cudf/pull/22503
Workaround nvcc compiler hangs in libcudf debug build by @davidwendt in https://github.com/rapidsai/cudf/pull/22675
Remove deprecated mark_join overloads without join_prefilter by @PointKernel in https://github.com/rapidsai/cudf/pull/22670
Add SECURITY.md by @jameslamb in https://github.com/rapidsai/cudf/pull/22684
Various cuDF CI fixes by @TomAugspurger in https://github.com/rapidsai/cudf/pull/22677
Reduce verbosity of cudf-polars-polars tests by @TomAugspurger in https://github.com/rapidsai/cudf/pull/22543
Drop the GIL while making datasources in read_parquet_footers by @TomAugspurger in https://github.com/rapidsai/cudf/pull/22685
Use CUB bulk for row hashing by @bdice in https://github.com/rapidsai/cudf/pull/22632
Minimize Ray resource acquisition in cudf_polars tests by @mroeschke in https://github.com/rapidsai/cudf/pull/22661
Adjust cudf-polars test timeouts by @TomAugspurger in https://github.com/rapidsai/cudf/pull/22691
Parallelize statistics collection by @TomAugspurger in https://github.com/rapidsai/cudf/pull/22645
Reduce MixedInnerJoinTest.LargeDataMultiBlockCoordination to a single type by @davidwendt in https://github.com/rapidsai/cudf/pull/22698
Add noinline to n_table_comparator::operator() by @davidwendt in https://github.com/rapidsai/cudf/pull/22699
Add debug-cudf-pandas skill for fixing pandas test compatibility failures by @vyasr in https://github.com/rapidsai/cudf/pull/22625
Remove stale python filterwarnings and skipped tests by @mroeschke in https://github.com/rapidsai/cudf/pull/22706
PERF: Skip header check in parquet reader by @TomAugspurger in https://github.com/rapidsai/cudf/pull/22679
docs: fix linked locations for notebooks after docs refactor by @gforsyth in https://github.com/rapidsai/cudf/pull/22710
Pass value to timeout parameters in subprocess/threading APIs in cudf_polars tests by @mroeschke in https://github.com/rapidsai/cudf/pull/22669
Drop the GIL in more places for parquet reads by @TomAugspurger in https://github.com/rapidsai/cudf/pull/22692
Add cuDF agent skill by @kjmph in https://github.com/rapidsai/cudf/pull/22721
Add peak_memory_usage tracking to nvbench benchmarks by @lhlRahman in https://github.com/rapidsai/cudf/pull/22548
Add more patterns to libcudf regex benchmarks by @davidwendt in https://github.com/rapidsai/cudf/pull/22213
Remove deprecated cudf::detail::copy_if_safe by @davidwendt in https://github.com/rapidsai/cudf/pull/22682
More miscellaneous cuDF Python testing cleanups by @mroeschke in https://github.com/rapidsai/cudf/pull/22713
Remove rapidsmpf from all dependencies by @bdice in https://github.com/rapidsai/cudf/pull/22655
Fix various pandas issues by @vyasr in https://github.com/rapidsai/cudf/pull/22705
Remove duplicated logic that does parquet metadata read by @Matt711 in https://github.com/rapidsai/cudf/pull/22741
fix(ci): fix configuration for breaking change notification workflow by @jameslamb in https://github.com/rapidsai/cudf/pull/22732
Define pandas and polars compatibility testing versions in dependencies.yaml by @Matt711 in https://github.com/rapidsai/cudf/pull/22189
Drop stale tests in cudf.pandas by @galipremsagar in https://github.com/rapidsai/cudf/pull/22618
Remove deprecated filtered_join set_as_build_table API by @PointKernel in https://github.com/rapidsai/cudf/pull/22671
Support polars.Expr.truncate by @mroeschke in https://github.com/rapidsai/cudf/pull/22754
Asynchronously fetch byte ranges from multiple sources by @mhaseeb123 in https://github.com/rapidsai/cudf/pull/22586
fix(ci): add missing permissions to pandas-tests workflow by @vyasr in https://github.com/rapidsai/cudf/pull/22761
Improve Parquet case-insensitive column selection semantics by @qbacpey in https://github.com/rapidsai/cudf/pull/22729
Unpin numba-cuda upper bound on main by @brandon-b-miller in https://github.com/rapidsai/cudf/pull/22730
Condense Decimal32/64/128Column to DecimalColumn, remove other unused code by @mroeschke in https://github.com/rapidsai/cudf/pull/22738
Bump polars upper bound to <1.42 by @madsbk in https://github.com/rapidsai/cudf/pull/22745
Refactor dynamic Scan node lowering by @TomAugspurger in https://github.com/rapidsai/cudf/pull/22758
Add benchmarks for reshaping operations by @Matt711 in https://github.com/rapidsai/cudf/pull/21212
Rename build/probe to left/right in mark_join by @PointKernel in https://github.com/rapidsai/cudf/pull/22723
Update cudf.pandas skill with recent changes by @galipremsagar in https://github.com/rapidsai/cudf/pull/22787
Rename build/probe to right/left in mixed_join_semi by @PointKernel in https://github.com/rapidsai/cudf/pull/22790
refactor(cudf_polars): use cudf_streaming for cudf-specific streaming imports by @vyasr in https://github.com/rapidsai/cudf/pull/22786
Make publish job search for abi3 wheels by @vyasr in https://github.com/rapidsai/cudf/pull/22815
Add fused expression IR tests by @Matt711 in https://github.com/rapidsai/cudf/pull/22805
refactor(rtcx): decouple librtcx from cudf and add standalone CMake build by @vyasr in https://github.com/rapidsai/cudf/pull/22744
Add remaining cudf_streaming tests and benchmarks by @vyasr in https://github.com/rapidsai/cudf/pull/22814
Treat \r\n as a single line terminator for ^/$ under EXT_NEWLINE by @wjxiz1992 in https://github.com/rapidsai/cudf/pull/22763
Add rolling window tests that exercise overflow-like edge cases for SUM by @wence- in https://github.com/rapidsai/cudf/pull/22660
Add send_chunk utility for repeated trace-then-send pattern by @rjzamora in https://github.com/rapidsai/cudf/pull/22651
Add unstable decorator to cudf-polars by @madsbk in https://github.com/rapidsai/cudf/pull/22804
Add CUDA error checks after every kernel launch by @PointKernel in https://github.com/rapidsai/cudf/pull/22755
Add FusedScan and nxtx annotations for FusedScan and SplitScan by @Matt711 in https://github.com/rapidsai/cudf/pull/22838
Improve performance of strings count_matches utility with specific expression patterns by @davidwendt in https://github.com/rapidsai/cudf/pull/22429
Increase polars IR coverage in tests by @Matt711 in https://github.com/rapidsai/cudf/pull/22847
Derive per-chunk memory reservation from scan statistics by @Matt711 in https://github.com/rapidsai/cudf/pull/22846
Reenable benchmarks when rapidsmpf supports MPI by @vyasr in https://github.com/rapidsai/cudf/pull/22853
Build libcudf_streaming comm benchmarks with partial MPI/UCXX availability by @pentschev in https://github.com/rapidsai/cudf/pull/22860
Don't capture GPU hardware information during CPU benchmarking runs by @mroeschke in https://github.com/rapidsai/cudf/pull/22862
Use google cloud read-only Maven Central mirror by @davidwendt in https://github.com/rapidsai/cudf/pull/22875
Collect CPU hardware info during benchmarking by @mroeschke in https://github.com/rapidsai/cudf/pull/22873
Pin matplotlib<3.11 for pandas-tests CI job by @vyasr in https://github.com/rapidsai/cudf/pull/22882
ci: reduce timeout for cudf_streaming C++ tests to 5m by @vyasr in https://github.com/rapidsai/cudf/pull/22880
Improve parquet footer reading performance by speculatively reading footer bytes by @TomAugspurger in https://github.com/rapidsai/cudf/pull/22782
Add cudf_streaming to clang-format IncludeCategories by @vyasr in https://github.com/rapidsai/cudf/pull/22881
refactor: switch to rapids-artifact-name for consistent artifact naming by @gforsyth in https://github.com/rapidsai/cudf/pull/22327
Refactor streaming scan construction into StreamingScan static methods by @Matt711 in https://github.com/rapidsai/cudf/pull/22889
Revert unreviewed VARIANT array-index extraction changes by @vuule in https://github.com/rapidsai/cudf/pull/22893
chore(cudf.pandas): remove stale xfail entries for tests no longer in pandas by @vyasr in https://github.com/rapidsai/cudf/pull/22883
Use conservative column-size estimates in cudf-polars by @rjzamora in https://github.com/rapidsai/cudf/pull/22796
fix(artifacts): correctly label non-arch-dependent but cuda-dependent artifacts by @gforsyth in https://github.com/rapidsai/cudf/pull/22902
Fix Parquet column mis-resolution when allow_mismatched_pq_schemas is enabled by @qbacpey in https://github.com/rapidsai/cudf/pull/22803
Rename build/probe to right/left in key_remapping by @PointKernel in https://github.com/rapidsai/cudf/pull/22789
CI Add third-party testing of cuml by @betatim in https://github.com/rapidsai/cudf/pull/21897
Expand chunked Parquet writer benchmark also add formatted throughput by @Matt711 in https://github.com/rapidsai/cudf/pull/19049
Update pre-commit hooks by @vyasr in https://github.com/rapidsai/cudf/pull/22905
Expose the table_view version of unique/distinct_count in pylibcudf by @wence- in https://github.com/rapidsai/cudf/pull/22487
Revert xfail_strict=false for pandas tests by @vyasr in https://github.com/rapidsai/cudf/pull/22913
Cudf_streaming dedup tests from rapidsmpf by @nirandaperera in https://github.com/rapidsai/cudf/pull/22858
Add nvbench for libcudf dictionary concatenate by @davidwendt in https://github.com/rapidsai/cudf/pull/22798
Adapt to rapidsmpf changes by @wence- in https://github.com/rapidsai/cudf/pull/22919
Build and test with CUDA 13.3.0 by @bdice in https://github.com/rapidsai/cudf/pull/22825
Fix deprecation warning in transform_polynomials_concurrent benchmarks by @davidwendt in https://github.com/rapidsai/cudf/pull/22937
Test only oldest and newest supported polars versions in PR CI by @Matt711 in https://github.com/rapidsai/cudf/pull/22939
Default the chunked Parquet reader's pass_read_limit to reduce surprise OOM by @qbacpey in https://github.com/rapidsai/cudf/pull/22779
Add reproduce-ci agent skill for local CI failure reproduction by @vyasr in https://github.com/rapidsai/cudf/pull/22872
Adding cudf streaming dist tests by @nirandaperera in https://github.com/rapidsai/cudf/pull/22864
Hide absorbed rapids_logger from cudf exports by @bdice in https://github.com/rapidsai/cudf/pull/22538
fix(cudf.pandas): remove stale xfail entries from pandas-testing-plugin by @vyasr in https://github.com/rapidsai/cudf/pull/22904
Update CONTRIBUTING guide and build.sh script to include cudf streaming libraries by @Matt711 in https://github.com/rapidsai/cudf/pull/22943
Add string_view -> string normalization in pylibcudf test utilities by @vyasr in https://github.com/rapidsai/cudf/pull/22958
[FEA] Use the conda package for croaring instead of fetching via CPM by @vyasr in https://github.com/rapidsai/cudf/pull/22953
Enable deprecation warnings by default in build.sh by @vyasr in https://github.com/rapidsai/cudf/pull/22959
Use main shared-workflows branch by @bdice in https://github.com/rapidsai/cudf/pull/22960
java: use GCS mirror for Maven Central, other small changes by @jameslamb in https://github.com/rapidsai/cudf/pull/22969
Relax librdkafka pinning by @vyasr in https://github.com/rapidsai/cudf/pull/22961
Pin Cython <3.3.0a0 to avoid prerelease regression by @vyasr in https://github.com/rapidsai/cudf/pull/22977
Use matplotlib-base instead of matplotlib in conda test deps by @vyasr in https://github.com/rapidsai/cudf/pull/22974
Add row-group statistics availability check by @qbacpey in https://github.com/rapidsai/cudf/pull/22664
Fix mypy arg-type for from_buffers after pyarrow-stubs bump by @madsbk in https://github.com/rapidsai/cudf/pull/22979
Make HostColumnVector constructors for non-nested types public by @paul-aiyedun in https://github.com/rapidsai/cudf/pull/22498
Add the predicate in physical plan explain output by @Matt711 in https://github.com/rapidsai/cudf/pull/22984
Adjust verbosity of cudf-polars-polars-tests by @TomAugspurger in https://github.com/rapidsai/cudf/pull/22980
Omit Parquet min/max statistics for float/double columns containing NaN by @wjxiz1992 in https://github.com/rapidsai/cudf/pull/22818
Add a I/O partition planning information to benchmark runner by @Matt711 in https://github.com/rapidsai/cudf/pull/22945
Add additional regex gtests for contains, count, findall, and replace by @davidwendt in https://github.com/rapidsai/cudf/pull/22874
Support building and testing cudf-java on JDK 17/21 by @igorpeshansky in https://github.com/rapidsai/cudf/pull/23006
Add pandas-compatible args and caching to RangeIndex.to_numpy by @rpathade in https://github.com/rapidsai/cudf/pull/21896
Add regex-flags member variable to internal libcudf reprog class by @davidwendt in https://github.com/rapidsai/cudf/pull/22994
Remove nogil from pylibcudf view()/mutable_view() and hoist calls out of nogil blocks by @vyasr in https://github.com/rapidsai/cudf/pull/23002
Fuse multi-column range window offset generation by @mhaseeb123 in https://github.com/rapidsai/cudf/pull/22863
PERF/API: Specify datasource size to avoid HEAD requests for S3 endpoints by @TomAugspurger in https://github.com/rapidsai/cudf/pull/22739
Improve performance for concatenate on dictionary columns by @davidwendt in https://github.com/rapidsai/cudf/pull/22807
Bound mypy pre-commit dependencies by @vyasr in https://github.com/rapidsai/cudf/pull/23020
Move lists_column_device_view and structs_column_device_view to cudf public namespace by @davidwendt in https://github.com/rapidsai/cudf/pull/23008
Upgrade nanoarrow to 0.8.0 by @vyasr in https://github.com/rapidsai/cudf/pull/23019
Cull unused CudaStreamPoolConfig and dependents by @wence- in https://github.com/rapidsai/cudf/pull/23026
fix(csv): correct to_csv quoting documentation and expose QUOTE_NONE by @a-hirota in https://github.com/rapidsai/cudf/pull/21517
Avoid unnecessary copy when TableChunk data is already packed by @Matt711 in https://github.com/rapidsai/cudf/pull/23034
Update CuPy and NumPy dependency specs by @bdice in https://github.com/rapidsai/cudf/pull/22962
Document and Migrate host_span vs span public APIs by @joannamooon in https://github.com/rapidsai/cudf/pull/22588
Rename sum_with_overflow aggregation to sum_overflow and deprecate old APIs by @PointKernel in https://github.com/rapidsai/cudf/pull/23014
Add braces to prefetch_noexcept location var to fix clang-tidy reported error by @davidwendt in https://github.com/rapidsai/cudf/pull/23066
Update dictionary::get_index to use cub FindIf by @davidwendt in https://github.com/rapidsai/cudf/pull/22841
Skewed-Length String Dataset Benchmark by @abigalekim in https://github.com/rapidsai/cudf/pull/22876
Add additional regex pytests derived from cudf-spark integration tests by @davidwendt in https://github.com/rapidsai/cudf/pull/23065
cudf-classic and CI cleanups for NumPy/CuPy 2/14 version bump by @mroeschke in https://github.com/rapidsai/cudf/pull/23053
Enforce east const via clang-format QualifierAlignment by @PointKernel in https://github.com/rapidsai/cudf/pull/22834
cudf-polars PythonScan support with rank-aware IO sources by @madsbk in https://github.com/rapidsai/cudf/pull/22867
Add roles to cudf-polars benchmark run outputs. by @TomAugspurger in https://github.com/rapidsai/cudf/pull/23068
Improve scan partitioning planning to prefer sizes closer to target by @Matt711 in https://github.com/rapidsai/cudf/pull/23011
Improve chunk sampling in dynamic actors by @rjzamora in https://github.com/rapidsai/cudf/pull/23051
fix(deps): try to dynamically load nvrtc from wheel by @gforsyth in https://github.com/rapidsai/cudf/pull/23089
Add copyright header to .clang-format and enforce it via pre-commit by @nethum529 in https://github.com/rapidsai/cudf/pull/23073
Replace numeric_scalar parameters in cudf::strings::slice_strings with std::optional by @davidwendt in https://github.com/rapidsai/cudf/pull/23031
Adopt cuco's parametric bloom filter policy in Parquet and streaming by @PointKernel in https://github.com/rapidsai/cudf/pull/23049
Prefetch parquet metadata for scan tasks by @TomAugspurger in https://github.com/rapidsai/cudf/pull/22700
switch from 'cuda-python' to specific components (e.g. 'cuda-bindings') by @jameslamb in https://github.com/rapidsai/cudf/pull/23091
Emit Quent traces for Engine, Lifecycle, and Plans by @TomAugspurger in https://github.com/rapidsai/cudf/pull/22337
Add row-level JSON schema mismatch diagnostics by @wjxiz1992 in https://github.com/rapidsai/cudf/pull/22915
Add Quent trace export to .gitignore by @TomAugspurger in https://github.com/rapidsai/cudf/pull/23152
Support single-partition pl.Expr.is_empty by @mroeschke in https://github.com/rapidsai/cudf/pull/23170
Support single-partition pl.Expr.gather_every by @mroeschke in https://github.com/rapidsai/cudf/pull/23166
Support single-partition pl.Expr.drop_nans by @mroeschke in https://github.com/rapidsai/cudf/pull/23165
Support single-partition pl.Expr.cum_count by @mroeschke in https://github.com/rapidsai/cudf/pull/23163
Support pl.Expr.str.escape_regex by @mroeschke in https://github.com/rapidsai/cudf/pull/23171
Support pl.Expr.dt.century by @mroeschke in https://github.com/rapidsai/cudf/pull/23153
Support single-partition pl.Expr.has_nulls by @mroeschke in https://github.com/rapidsai/cudf/pull/23168
Add agent skill for cudf_polars expressions by @mroeschke in https://github.com/rapidsai/cudf/pull/23078
Support single-partition pl.Expr.extend_constant by @mroeschke in https://github.com/rapidsai/cudf/pull/23167
Refactor host_span to use cuda::std::span internally and remove span_base by @PointKernel in https://github.com/rapidsai/cudf/pull/23072
Support single-partition pl.Expr.product in cudf_polars by @mroeschke in https://github.com/rapidsai/cudf/pull/23169
Support pl.Expr.dt.millennium by @mroeschke in https://github.com/rapidsai/cudf/pull/23154
Support pl.Expr.dt.days_in_month by @mroeschke in https://github.com/rapidsai/cudf/pull/23156
Support pl.Expr.dt.quarter by @mroeschke in https://github.com/rapidsai/cudf/pull/23157
Propagate parent nulls into struct children in mergeAndSetValidity by @rishic3 in https://github.com/rapidsai/cudf/pull/22650
Support pl.Expr.rechunk by @mroeschke in https://github.com/rapidsai/cudf/pull/23192
Support remote storage URLs for PDS-H validation by @TomAugspurger in https://github.com/rapidsai/cudf/pull/23202
Support pl.Expr.clip by @mroeschke in https://github.com/rapidsai/cudf/pull/23162
Update benchmark data-generator to allow nullable with no nulls by @davidwendt in https://github.com/rapidsai/cudf/pull/23198
Rename some benchmark .cu files to .cpp by @davidwendt in https://github.com/rapidsai/cudf/pull/23186
Plumb statistics object into ProgressThread for all engines by @wence- in https://github.com/rapidsai/cudf/pull/23082
Support pl.Expr.reinterpret/degrees/radians by @mroeschke in https://github.com/rapidsai/cudf/pull/23210
Pin pyarrow < 25 in Polars tests by @mroeschke in https://github.com/rapidsai/cudf/pull/23218
Optimize RLE decoding by staging in shared memory by @vyasr in https://github.com/rapidsai/cudf/pull/23090
Fix IO benchmark naming consistency by @Tjindl in https://github.com/rapidsai/cudf/pull/23180
Error when parquet decode string corruption detected by @pmattione-nvidia in https://github.com/rapidsai/cudf/pull/23203
Add dictionary benchmarks for match-keys and join by @davidwendt in https://github.com/rapidsai/cudf/pull/23185
Use rtcx from the librtcx repository by @vyasr in https://github.com/rapidsai/cudf/pull/23137
Support pl.Expr.cot/log1p by @mroeschke in https://github.com/rapidsai/cudf/pull/23191
Start async device writes and default host reads eagerly instead of deferring I/O to future wait by @Emillock in https://github.com/rapidsai/cudf/pull/23231
Move function definitions from generate_input.hpp to generate_input.cu by @davidwendt in https://github.com/rapidsai/cudf/pull/23199
Add fast-path logic to some libcudf regex APIs by @davidwendt in https://github.com/rapidsai/cudf/pull/22178
Fix parquet preprocess strings page bounds check for pages with zero rows by @pmattione-nvidia in https://github.com/rapidsai/cudf/pull/23241
Support pl.Expr.truncate/unique_counts by @mroeschke in https://github.com/rapidsai/cudf/pull/23244
Use batched memcpy instead of a custom copy kernel by @mhaseeb123 in https://github.com/rapidsai/cudf/pull/23266
Add multifile dictionary pruning support for hybrid scan parquet reader by @qbacpey in https://github.com/rapidsai/cudf/pull/22866
Replace docs.rapids.ai links that 400 by @mroeschke in https://github.com/rapidsai/cudf/pull/23270
Support pl.Expr.index_of/search_sorted by @mroeschke in https://github.com/rapidsai/cudf/pull/23252
Support pl.Expr.arg_max/arg_min/arg_unique/arg_sort by @mroeschke in https://github.com/rapidsai/cudf/pull/23212
Support pl.Expr.round_sig_figs/sign by @mroeschke in https://github.com/rapidsai/cudf/pull/23250
Add engine.execute() returning a PersistedQueryResult by @madsbk in https://github.com/rapidsai/cudf/pull/23114
No longer link Arrow in CSV and JSON tests by @mroeschke in https://github.com/rapidsai/cudf/pull/23283
Support pl.Expr.replace/replace_strict by @mroeschke in https://github.com/rapidsai/cudf/pull/23253
cudf-polars: preserve zero-column row counts by @madsbk in https://github.com/rapidsai/cudf/pull/23234
Improve column selection validation under mismatched PQ schemas by @qbacpey in https://github.com/rapidsai/cudf/pull/23000
Support pl.argwhere/pl.Expr.arg_true by @mroeschke in https://github.com/rapidsai/cudf/pull/23284
Update upstream cuml tests run by @jcrist in https://github.com/rapidsai/cudf/pull/23335
Empty commit to trigger a build by @bdice in https://github.com/rapidsai/cudf/pull/23376
Remove rmm::device_buffer forward declaration from types.hpp by @davidwendt in https://github.com/rapidsai/cudf/pull/23373
Bump cudf-polars test suite timeouts by @TomAugspurger in https://github.com/rapidsai/cudf/pull/23399
Bump cudf-polars CI timeouts by @TomAugspurger in https://github.com/rapidsai/cudf/pull/23410
Disable join filter pushdown rewrites by default by @TomAugspurger in https://github.com/rapidsai/cudf/pull/23407
Add profitability guards for join filter pushdown by @pentschev in https://github.com/rapidsai/cudf/pull/22997

New Contributors

@piotrrzysko made their first contribution in #22486
@findepi made their first contribution in #22575
@0ax1 made their first contribution in #22620
@kjmph made their first contribution in https://github.com/rapidsai/cudf/pull/22721
@lhlRahman made their first contribution in https://github.com/rapidsai/cudf/pull/22548
@betatim made their first contribution in https://github.com/rapidsai/cudf/pull/21897
@anthonyguerrera made their first contribution in #22436
@igorpeshansky made their first contribution in https://github.com/rapidsai/cudf/pull/23006
@rpathade made their first contribution in https://github.com/rapidsai/cudf/pull/21896
@utkarshparekh made their first contribution in #23021
@gmartini2000 made their first contribution in #21991
@joannamooon made their first contribution in https://github.com/rapidsai/cudf/pull/22588
@dttutty made their first contribution in #23195
@Emillock made their first contribution in https://github.com/rapidsai/cudf/pull/23231
@lingyany-nv made their first contribution in #21936
@tolleybot made their first contribution in #21964

Full Changelog: v26.08.00a...v26.08.00