main
PERF: avoid materializing set/list in isin for common dtype cases (#65284)
PERF: parse read_csv integers with fast_float::from_chars (#66459)
STY: Use ruff's pyupgrade (#66484)
CLN: Remove ipython 2.x compat (#25150)
CLN: Reduce and cleanup benchmarks/algorithms (#65651)
PERF: datetimelike addition (#56373)
STY: Enable B904 (#56941)
PERF: speed up tab completion for DataFrame/Series with a large index (#65277)
ENH: Implement Kleene logic for BooleanArray (#29842)
CLN: Enable ruff rule RUF005 (collection-literal-concatenation) (#63698)
PERF: faster to_datetime parsing of timezone-aware strings (#66123)
TST/CLN: Remove makeStringIndex (#56155)
DEPR: deprecate strings T, S, L, U, and N in offsets frequencies, resolution abbreviations, _attrname_to_abbrevs (#54061)
PERF: Added finalize benchmark (#33275)
CI: Fix numpy nightly failures from generic timedelta64 deprecation (#64993)
[pre-commit.ci] pre-commit autoupdate (#61802)
PERF: add an Arrow/NumPy hybrid groupby path for decimal and string types. (#63416)
PERF: RangeIndex.append with the same index (#57252)
PERF: typed non-boxing path for get_indexer_non_unique (#66125)
MNT: Bump dev pin on NumPy (#60987)
PERF: single-pass composite-key factorization for multi-key merges (#66124)
Bump ruff to latest version (#57766)
ASV: garbage collect as part of setup (#65681)
CLN/DOC: fix test_register_entrypoint (#46302)
ASV: Run inplace fillna once per setup (#65688)
CLN/ASV: Remove NumPy/built-in aliases in ASVs (#54316)
ENH: Add Rolling.nunique() (#61087)
STY: Add strict=True in zip() in \_testing and \asv_bench (#62793)
PERF: speed up nanops pearson correlation (#65502)
CLN: Remove pandas.value_counts, Index.format, is_interval, is_period (#57296)
REF: get rid of StringArrayNumpySemantics (#62149)
PERF: use Arrow native path in duration component calculation for Arrow-backed array (#63470)
DEPR: Enforce datetimelike deprecations (#57999)