cudf · Issues· 1342 open
Open on GitHubLocally synced open issues (discussions stay on GitHub)
- #24233
[BUG] Ambiguous union_member::get overloads with GCC 16 break ORC/Parquet writer builds
Updated Sep 18, 2026 - #24232
[BUG] Missing #include <variant> in cudf/ast/expressions.hpp breaks build with GCC 16
Updated Sep 18, 2026 - #24222
[BUG] HashCSR increases retained hash_join memory by 82% for large unique builds
buglibcudfUpdated Sep 18, 2026 - #24200
[FEA] Explore lto fragments in cudf-polars
feature requestcudf-polarsUpdated Sep 17, 2026 - #24040
[BUG] groupby().sum() changes uint64 input to int64
bug? - Needs TriageUpdated Sep 17, 2026 - #24166
[FEA] Improving read parquet + udf + partitioned parquet write at scale
feature request? - Needs TriageUpdated Sep 17, 2026 - #22748
[FEA] Java: Table.mixedInnerJoinGatherMaps variant accepting a pre-built HashJoin
feature requestJavaUpdated Sep 17, 2026 - #24096
[FEA] Expose filter_join_indices in Java
feature requestJava? - Needs TriageUpdated Sep 17, 2026 - #23422
[FEA] ORC writer: allow setting writerTimezone instead of hardcoding "UTC"
feature requestUpdated Sep 17, 2026 - #23519
[STORY] Support page-level I/O in hybrid scan
feature requestlibcudfcuIOUpdated Sep 17, 2026 - #22128
[Story][FEA] Order-aware streaming execution in cuDF-Polars
feature requestcudf-polarsUpdated Sep 16, 2026 - #24194
[FEA] Decode Parquet decimals directly into prescribed per-column types
Updated Sep 16, 2026 - #24002
[BUG] chunked_parquet_reader hangs or hits cudaErrorIllegalAddress on zstd DataPageV2 files: nvCOMP temp-size query is given the uncompressed level bytes
2 - In ProgressUpdated Sep 15, 2026 - #23963
[Story]: Trim cudf-polars's own test suite down to what upstream polars can't cover
improvementcudf-polarsUpdated Sep 14, 2026 - #22949
Use Quent bindings in cudf-polars
Pythonimprovementcudf-polarsUpdated Sep 14, 2026