misc/py-narwhals 2.25.0 NetBSD

2.27.1 fix: return a numeric dtype from str.len_chars on all-null object columns fix: count characters, not bytes, in str.zfill on Polars < 2.0

2.26.0 Performance improvements perf: Avoid getting columns to read dtypes, route bare selectors through simple_select perf: Speedup pandas-like is_finite, sqrt, kurtosis and skew Enhancements feat: Improve error message on unsupported length-changing methods Bug fixes fix(pyarrow): strip_chars with empty characters no longer strips whitespace ci: fix downstream pins for pointblank fix: use dtype.categories for pl.Enum in get_categories fix(typing): Stableify selectors fix(typing): Stableify concat with the stable frame classes fix(pyarrow): shift preserves length when n exceeds the series length fix(pyarrow): keep nulls through clip instead of replacing them with a bound fix: Deprecate cat.get_categories and stop relying on Polars' removed method …

2.25.0 Performance improvements perf: remove redundant work in eager when/then, schema resolution, selectors, and module lookups Enhancements feat: add str.strip_chars_{start,end} feat: Add maintain_order argument to {Expr, Series}.list.unique feat(typing): Introduce PluginName NewType so plugin backends can type check feat(ibis): Add support for Expr.list.median for ibis backend Bug fixes fix: follow left-hand-index alignment rule for pandas-like concat fix: null propagation for str boolean methods fix: consistently place null's in with_row_index(order_by=...) fix: honor schema column selection/order in pandas-like from_dicts fix: use exact quantile for median on the pyarrow backend fix: order Categorical columns by value in the pyarrow backend …

2.24.0 Enhancements feat: Widen IntoSchema to accept IntoDType values and (name, dtype) pair sequences feat: add nw.list feat: Support plugin native objects in is_into_* feat: stabilize dtype conversion across pandas, pyarrow, and modin test constructors Bug fixes fix: avoid potential sql injection in sink_parquet fix(pyarrow): Adapt for pyarrow v25.0.0 fix(ci): sync all dependencies in altair downstream test fix: Fix fill_null for Dask with expressions fix: avoid potential sql injection for join_asof fix assert_equal_data for objects where __eq__ returns non-bool Documentation docs: Add 'AI-assisted contributions' section in CONTRIBUTING.md and PR field

Extremely lightweight and extensible compatibility layer between dataframe libraries! * Full API support: cuDF, Modin, pandas, Polars, PyArrow. * Lazy-only support: Daft, Dask, DuckDB, Ibis, PySpark, SQLFrame.