38 Pull requests merged by 21 people

• remove WheelHelpers.cmake

  #17276 merged Nov 8, 2024

• Make constructor of DeviceMemoryBufferView public

  #17265 merged Nov 8, 2024

• Wrap custom iterator result

  #17251 merged Nov 8, 2024

• Implement inequality joins by translation to conditional joins

  #17000 merged Nov 8, 2024

• Rewrite Java API Table.readJSON to return the output from libcudf read_json directly

  #17180 merged Nov 8, 2024

• Add IWYU to CI

  #17078 merged Nov 8, 2024

• Plumb pylibcudf datetime APIs through cudf python

  #17275 merged Nov 8, 2024

• Fix data_type ctor call in JSON_TEST

  #17273 merged Nov 8, 2024

• Allow generating large strings in benchmarks

  #17224 merged Nov 8, 2024

• Add optional column_order in JSON reader

  #17029 merged Nov 8, 2024

• Mark column chunks in a PQ reader pass as large strings when the cumulative offsets exceeds the large strings threshold.

  #17207 merged Nov 7, 2024

• Use pylibcudf.search APIs in cudf python

  #17271 merged Nov 7, 2024

• Use pylibcudf.strings.convert.convert_integers.is_integer in cudf python

  #17270 merged Nov 7, 2024

• Move strings to date/time types benchmarks to nvbench

  #17229 merged Nov 7, 2024

• Process parquet bools with microkernels

  #17157 merged Nov 7, 2024

• Add support for pyarrow-18

  #17256 merged Nov 7, 2024

• Add io.text APIs to pylibcudf

  #17232 merged Nov 7, 2024

• AWS S3 IO through KvikIO

  #16499 merged Nov 7, 2024

• Refactor gather/scatter benchmarks for strings

  #17223 merged Nov 7, 2024

• Fix extract-datetime deprecation warning in ndsh benchmark

  #17254 merged Nov 7, 2024

• cudf-polars string/numeric casting

  #17076 merged Nov 7, 2024

• Added ast tree to simplify expression lifetime management

  #17156 merged Nov 7, 2024

• Move strings/numeric convert benchmarks to nvbench

  #17255 merged Nov 7, 2024

• Fix the example in documentation for get_dremel_data()

  #17242 merged Nov 7, 2024

• Put a ceiling on cuda-python

  #17264 merged Nov 7, 2024

• KvikIO shared library

  #17239 merged Nov 7, 2024

• Disallow cuda-python 12.6.1 and 11.8.4

  #17253 merged Nov 6, 2024

• Search for kvikio with lowercase

  #17243 merged Nov 6, 2024

• Deprecate single component extraction methods in libcudf

  #17221 merged Nov 5, 2024

• Separate evaluation logic from IR objects in cudf-polars

  #17175 merged Nov 5, 2024

• Refactor Dask cuDF legacy code

  #17205 merged Nov 4, 2024

• Fix discoverability of submodules inside pd.util

  #17215 merged Nov 4, 2024

• Use more pylibcudf.io.types enums in cudf._libs

  #17237 merged Nov 4, 2024

• Expose mixed and conditional joins in pylibcudf

  #17235 merged Nov 4, 2024

• Make HostMemoryBuffer call into the DefaultHostMemoryAllocator

  #17204 merged Nov 4, 2024

• Expose stream-ordering in subword tokenizer API

  #17206 merged Nov 4, 2024

• Add num_iterations axis to the multi-threaded Parquet benchmarks

  #17231 merged Nov 2, 2024

• Change default KvikIO parameters in cuDF: set the thread pool size to 4, and compatibility mode to ON

  #17185 merged Nov 1, 2024

24 Pull requests opened by 14 people

• Add read_parquet_metadata to pylibcudf

  #17245 opened Nov 5, 2024

• Migrate contiguous split APIs to pylibcudf

  #17246 opened Nov 5, 2024

• Add breaking change workflow trigger

  #17248 opened Nov 5, 2024

• POC: Implement `HOST_UDF` aggregations

  #17249 opened Nov 5, 2024

• [WIP] Add new ``dask_cudf.read_parquet`` API

  #17250 opened Nov 5, 2024

• Add write_parquet to pylibcudf

  #17252 opened Nov 6, 2024

• Expose streams in public quantile APIs

  #17257 opened Nov 6, 2024

• Add type stubs for pylibcudf

  #17258 opened Nov 6, 2024

• Always prefer `device_read`s when kvikio is enabled

  #17260 opened Nov 6, 2024

• [DNM][WIP] Single-partition Dask executor for cuDF-Polars

  #17262 opened Nov 7, 2024

• Add write_parquet to pylibcudf

  #17263 opened Nov 7, 2024

• Expose delimiter character in JSON reader options to JSON reader APIs

  #17266 opened Nov 7, 2024

• Add `catboost` to the third-party integration tests

  #17267 opened Nov 7, 2024

• Raise errors on specific types of fallback in `cudf.pandas`

  #17268 opened Nov 7, 2024

• Move strings filter benchmarks to nvbench

  #17269 opened Nov 7, 2024

• Follow up making Python tests more deterministic

  #17272 opened Nov 7, 2024

• Add `cudf::calendrical_month_sequence` to pylibcudf

  #17277 opened Nov 7, 2024

• Optimize distinct inner join to use set `find` instead of `retrieve`

  #17278 opened Nov 8, 2024

• Add compute_column_expression to pylibcudf for transform.compute_column

  #17279 opened Nov 8, 2024

• Use numba-cuda<0.0.18

  #17280 opened Nov 8, 2024

• Java JNI for Multiple contains

  #17281 opened Nov 8, 2024

• test PR

  #17283 opened Nov 8, 2024

• WIP: [DO NOT MERGE] enforce wheel size limits, README formatting in CI

  #17284 opened Nov 8, 2024

• Switch to using `TaskSpec`

  #17285 opened Nov 8, 2024

23 Issues closed by 7 people

• [FEA] Check why we need `__iter__` special overrides for `cudf.pandas`

  #14481 closed Nov 8, 2024

• [BUG] Incorrect dtype when iterating over dtypes in cudf.pandas

  #17165 closed Nov 8, 2024

• [FEA] Support conditional joins in cudf-polars

  #16926 closed Nov 8, 2024

• [FEA] Implement a better JNI function to assemble the output columns from `cudf::read_json`

  #17002 closed Nov 8, 2024

• Compile times are growing significantly

  #581 closed Nov 8, 2024

• [BUG] `cudf::io::read_json` does not verify output column structures with the input schema

  #16799 closed Nov 8, 2024

• [FEA] `read_json` should output all-nulls columns for the schema columns that do not exist in the input

  #17091 closed Nov 8, 2024

• [FEA] The output columns of `read_json` need to follow depth-first-search order as in the input schema

  #17090 closed Nov 8, 2024

• [BUG] JSON parser still returns outdated schema structure for strings column

  #17240 closed Nov 8, 2024

• [BUG] Chunked parquet reader incorrect results for large string columns

  #17158 closed Nov 7, 2024

• [BUG] The read_json() method of cudf can't parse the string like "5-0"

  #14670 closed Nov 7, 2024

• The example in the documentation for `get_dremel_data()` seems incorrect at line#1764

  #11396 closed Nov 7, 2024

• Reduce IO when `byte_range` option is used in `read_json`

  #15185 closed Nov 5, 2024

• Memory mapped datasource does not allow reading data beyond the mapped range

  #15186 closed Nov 5, 2024

• [FEA] [Proposal] Separate IR evaluation logic from the IR object in cudf-polars

  #17127 closed Nov 5, 2024

• [FEA] Support datetime64[D]

  #16803 closed Nov 5, 2024

• [BUG] Series.plot method populates pd.util.version when run under cudf.pandas

  #17166 closed Nov 4, 2024

• [FEA] support max and min aggregations for nested structs

  #8974 closed Nov 4, 2024

• [FEA] Add nested struct support for comparison operations

  #8964 closed Nov 4, 2024

• [BUG] new parquet writer code checks for `nullable` not `has_nulls`

  #7654 closed Nov 4, 2024

• [BUG] JSON reader metadata contains in an extra child for string column inside (deeply) nested structs/lists

  #17108 closed Nov 4, 2024

• [FEA] Adjust libcudf to not load cuFile by default

  #16512 closed Nov 1, 2024

• [BUG] `cudf::io::json::detail::get_token_stream` does not respect `normalize_single_quotes` option

  #17230 closed Nov 1, 2024

9 Issues opened by 7 people

• [BUG] pd.api.interchange.from_dataframe fails with simple cuDF dataframe

  #17282 opened Nov 8, 2024

• [BUG] AST_TEST TransformTest.DeeplyNestedArithmeticLogicalExpression fails when run in a debug build

  #17274 opened Nov 7, 2024

• [BUG] `cudf::io::json::detail::normalize_single_quotes` outputs incorrect result when the input has `\n` character

  #17261 opened Nov 7, 2024

• [FEA] Adjust libcudf to use kvikIO for small host reads

  #17259 opened Nov 6, 2024

• [FEA] Compare performance of decompression engine (HW) versus decompression kernels (SW) on Blackwell

  #17247 opened Nov 5, 2024

• [FEA] Support `strict=False` casting in `cudf-polars`

  #17244 opened Nov 4, 2024

• Support Hash-based group by aggregations for min/max with nesting

  #17241 opened Nov 4, 2024

• [FEA] read_csv optimizations with streaming multiprocessors and MGPU

  #17238 opened Nov 2, 2024

• [FEA] Add function for "deduplicate map" to libcudf

  #17236 opened Nov 1, 2024

37 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• Add cudf::strings::contains_multiple

  #16900 commented on Nov 8, 2024 • 19 new comments

• [FEA] Report all unsupported operations for a query in cudf.polars

  #16960 commented on Nov 8, 2024 • 15 new comments

• Add new nvtext minhash_permuted API

  #16756 commented on Nov 8, 2024 • 13 new comments

• Reading multi-source compressed JSONL files

  #17161 commented on Nov 8, 2024 • 7 new comments

• Migrate CSV writer to pylibcudf

  #17163 commented on Nov 7, 2024 • 6 new comments

• Improve the performance of low cardinality groupby

  #16619 commented on Nov 8, 2024 • 5 new comments

• Polars: DataFrame Serialization

  #17062 commented on Nov 8, 2024 • 4 new comments

• Benchmarking JSON reader for compressed inputs

  #17219 commented on Nov 7, 2024 • 2 new comments

• Support hyper log log plus plus(HLL )

  #17133 commented on Nov 6, 2024 • 2 new comments

• Implement `cudf-polars` chunked parquet reading

  #16944 commented on Nov 8, 2024 • 1 new comment

• add telemetry setup to test

  #16924 commented on Nov 4, 2024 • 0 new comments

• Use block per string for super long strings in cudf::strings::find()

  #17011 commented on Nov 6, 2024 • 0 new comments

• Use `libcudf_exception_handler` throughout `pylibcudf.libcudf`

  #17109 commented on Nov 7, 2024 • 0 new comments

• WIP: Support decimal dtypes

  #17113 commented on Nov 7, 2024 • 0 new comments

• Improve cudf::io::datasource::create().

  #17115 commented on Nov 3, 2024 • 0 new comments

• [DO NOT MERGE/REVIEW] GDS debugging

  #17116 commented on Nov 5, 2024 • 0 new comments

• [test] compression benchmarks

  #17145 commented on Nov 8, 2024 • 0 new comments

• Added Arrow Interop Benchmarks

  #17194 commented on Nov 7, 2024 • 0 new comments

• Fix `Dataframe.__setitem__` slow-downs

  #17222 commented on Nov 8, 2024 • 0 new comments

• Precompute AST arity

  #17234 commented on Nov 5, 2024 • 0 new comments

• [FEA] Support COUNT_VALID and COUNT_ALL for scan and group by scan

  #8710 commented on Nov 4, 2024 • 0 new comments

• [FEA] aggregation each list in a column to a single value using a user supplied function

  #8020 commented on Nov 4, 2024 • 0 new comments

• [FEA] Full coverage of datetime methods in cudf-polars

  #16481 commented on Nov 5, 2024 • 0 new comments

• [FEA] Confirm correct and efficient dictionary processing in libcudf

  #15199 commented on Nov 5, 2024 • 0 new comments

• [FEA] Use bloom filters in Parquet reader to filter row groups with equality predicates

  #17164 commented on Nov 5, 2024 • 0 new comments

• [Story] Enable multithreading in cuIO and libcudf

  #17119 commented on Nov 5, 2024 • 0 new comments

• [FEA] Global thread pool in benchmarks

  #16801 commented on Nov 5, 2024 • 0 new comments

• [FEA] Add cudf-polars to test.yaml

  #16383 commented on Nov 5, 2024 • 0 new comments

• [QST] cudf.pandas prefer using CPU over GPU in some cases

  #14500 commented on Nov 5, 2024 • 0 new comments

• [FEA] Offer more control over CPU fallback in cudf.pandas

  #14975 commented on Nov 5, 2024 • 0 new comments

• [FEA] Replace internal usage of std::string with std::string_view

  #15907 commented on Nov 6, 2024 • 0 new comments

• [FEA] Provide type stubs for pylibcudf package

  #15190 commented on Nov 6, 2024 • 0 new comments

• [FEA] Expose public stream-ordered C APIs

  #13744 commented on Nov 6, 2024 • 0 new comments

• [BUG] read_json fails when reading multiple files but works with the individual files

  #14268 commented on Nov 7, 2024 • 0 new comments

• [QST] Does the read_json() method support GPU acceleration?

  #14669 commented on Nov 7, 2024 • 0 new comments

• [FEA] Implement all libcudf modules required by cuDF Python in pylibcudf

  #15162 commented on Nov 8, 2024 • 0 new comments

• Fix reading of single-row unterminated CSV files

  #16055 commented on Nov 8, 2024 • 0 new comments