Write dataframe-agnostic Python code that runs across multiple backends.
npx skills add https://github.com/anam-org/metaxy --skill narwhals
Please help me install this Agent Skill. Skill: narwhals Source: https://github.com/anam-org/metaxy/tree/main/.claude/skills/narwhals Command: npx skills add https://github.com/anam-org/metaxy --skill narwhals
Narwhals enables writing dataframe-agnostic code that works across multiple Python dataframe libraries, eliminating backend-specific silos and reducing maintenance overhead.
Install Narwhals with pip install narwhals. Then import narwhals as nw and annotate functions with @nw.narwhalify to enable cross-backend compatibility.
High-intent search queries and answers about installing and using this skill.
Dataframe-agnostic code runs across pandas and polars by using the Narwhals library to provide a unified API. You wrap library functions with the @narwhalify decorator to enable automatic conversions between native DataFrames.
Supporting multiple dataframe backends without duplicating logic requires a backend-agnostic DataFrame API. Narwhals eliminates backend-specific silos by allowing you to write code once that runs on pandas, polars, and cuDF.
Static typing works with dataframe-agnostic code across backends using Narwhals. It provides full static typing support so your type-checked Narwhals code remains compatible across pandas, polars, and other supported libraries.
To make existing pandas functions compatible with polars, install Narwhals via pip and import it as nw. Annotate your functions with @nw.narwhalify to enable cross-backend compatibility and automatic native DataFrame conversion.
Narwhals works with cuDF for feature engineering tasks in data pipelines. It provides a backend-agnostic DataFrame API designed for data analysis where code must run seamlessly across pandas, polars, and cuDF environments.
Maintaining separate codebases for different dataframe libraries creates maintenance overhead and backend-specific silos. Narwhals solves this by enabling developers to write dataframe-agnostic Python code that runs across multiple backends without duplication.
Ritabrata-ChakrabortyProcess large datasets with Polars DataFrame operations and lazy evaluation.
Marwan78888Automate pandas data cleaning, groupby aggregations, and feature engineering on large DataFrames.
ManfronEnricoProcess in-memory data with Polars DataFrames using lazy evaluation.
JonathanMitchell1234Perform data cleaning, feature engineering, and statistical modeling with pandas and scikit-learn.
