📝 Add Narwhals

Author: veit

CHANGELOG.rst

@@ -23,6 +23,7 @@ emergencies when we need to start branches for older versions.
 Added
 ~~~~~
 
+* 📝 Add Narwhals
 * 📝 Add protobuf-py
 
 Changed

docs/performance/parallelise-pandas.rst

@@ -9,6 +9,8 @@ In  :doc:`pandas:user_guide/enhancingperf`, some possibilities are described for
 improving the performance of pandas. However, there are also special libraries
 that can parallelise the processing of data frames.
 
+.. _cudf:
+
 cuDF
 ----
 
@@ -23,6 +25,8 @@ cuDF is a GPU DataFrame library that implements a `pandas-like API
     * `Example notebooks
       <https://github.com/rapidsai-community/notebooks-contrib>`_
 
+.. _modin:
+
 Modin
 -----
 
@@ -54,3 +58,13 @@ of the pandas API, although there are minor changes.
     * `API docs <https://docs.dask.org/en/latest/dataframe-api.html>`_
     * `Example notebook <https://examples.dask.org/dataframe.html>`_
     * `Tutorial <https://tutorial.dask.org/01_dataframe.html>`_
+
+Narwhals
+--------
+
+`Narwhals <https://narwhals-dev.github.io/narwhals/>`_ provides a lightweight
+and extensible compatibility layer between various DataFrame libraries, with
+full API support for :ref:`cudf`, :ref:`modin`,
+:doc:`../workspace/pandas/index`, :ref:`polars` and `PyArrow
+<https://arrow.apache.org/docs/python/index.html>`_. :ref:`dask`, :ref:`duckdb`
+and others are also supported to a certain extent.

docs/workspace/pandas/index.rst

@@ -49,16 +49,25 @@ of workload:
 
 pandas
     is the canonical Python DataFrame library for analysis on a single machine.
+
+.. _polars:
+
 Polars
     is written in Rust and allows for powerful analysis on a single node or when
     `lazy evaluation <https://en.wikipedia.org/wiki/Lazy_evaluation>`_ and
     `expressions API
     <https://docs.pola.rs/api/python/stable/reference/expressions/index.html>`_
     are important.
+
+.. _dask:
+
 Dask
     is a Python library for parallel computing that scales familiar APIs,
     including pandas and `Scikit-Learn <https://scikit-learn.org/stable/>`_, to
     clusters.
+
+.. _duckdb:
+
 DuckDB
     is an in-process `OLAP
     <https://en.wikipedia.org/wiki/Online_analytical_processing>`_ database for