py-1.42.1
📦 polarsView on GitHub →
✨ 3 features🐛 10 fixes⚡ 1 deprecations🔧 6 symbols
Summary
This release introduces performance improvements, including faster small dtype sums and non-blocking path expansion, while deprecating the `strict` parameter in `pl.concat`. Numerous bugs related to panics, data casting, and resource leaks have also been resolved.
Migration Steps
- When using `pl.concat`, replace the `strict` parameter with `how='horizontal_extend'`.
✨ New Features
- Added a sampled resolve mode for multi-file parquet metadata.
- Speed up small dtype series sums with an upcast sum kernel.
- Don't block on path expansion.
🐛 Bug Fixes
- Fixed incorrect sorted fast-path returning NaN for grouped max.
- Fixed panic on scan_parquet filter of fixed-size binary column.
- Fixed decimal dynamic float cast.
- Fixed pyiceberg key dot HDFS prefix filter.
- Fixed projection pushdown panic on `select(len())` after `groupby`.
- Fully resolved remaining `ResourceWarning` leaks in database/iceberg tests.
- Fixed expiry time loading for AWS `DeferredRefreshableCredentials`.
- Closed `ResourceWarning` leaks in database tests.
- Avoided panic on temporal extraction for datetime columns with nulls.
- Prevented panic in `replace` when `old`/`new` contain Expr or object dtype values.
Affected Symbols
⚡ Deprecations
- The `strict` parameter of `pl.concat` is deprecated and should be replaced with the new `how='horizontal_extend'` argument.