Polars package¶
Polars is a "blazingly fast" DataFrame library for manipulating structured data.
Website: https://pola.rs/ Docs: https://docs.pola.rs/
Features:
- Among the fastest data.frame engines with DuckDB, according to their own benchmarks. Polars is significantly faster than pandas, and even beats R's excellent data.table package in some scenarios.
- Consistent API (unlike, e.g., pandas)
- Built-in query optimization, made even more powerful through lazy evaluation (see LazyFrame)
- Native parallel computation support for many operations
- Support for common data storage formats and layers.
- Streaming support: Can handle datasets larger than RAM allows
- Uses Apache Arrow arrays to represent data in memory while Pandas uses Numpy arrays, allows for easy integration with other tools that support Apache Arrow.
- Can import and export from pandas