Expressions and contexts

Leverage Polars’ Domain Specific Language to write complex, human-readable data transformations. Utilize expressions as modular, flexible building blocks for creating complex logic, such as computing BMI via column operations. Since expressions are lazy representations, they perform no computation until executed within a context. Contexts like select, with_columns, filter, and group_by determine execution results, meaning identical expressions yield different outputs based on their environment. This design enables the Polars query engine to optimize queries for maximum speed while maintaining readability. Save expressions in variables for further manipulation or printing, allowing abstract computations to be deferred until necessary. Scope: Polars. Reviewed October 4, 2026.