Polars 2.0 makes streaming and disk spilling the default
by Fla

Polars 2.0 makes streaming and disk spilling the default

Polars introduces version 2.0 that uses its streaming engine by default when collecting a LazyFrame, with spill-to-disk execution enabled for supported operations.

The release expands first-class SQL support with optimizer changes including join reordering and improved common-subplan elimination. It also introduces a native Map dtype and stricter dtype handling, including early schema validation through 'collect_schema()'.

Fla
Written by Fla
Follow
Add as a preferred source on Google
MORE ABOUTPolars
Polars iconPolars
  3
  • Free
  • Open Source
  • ...

Polars is a highly performant DataFrame library for manipulating structured data. The core is written in Rust, but the library is also available in Python. Its key features are:

No comments so far, maybe you want to be first?