Polarise

Style your data to explore. Style your results to present.

A Polars-native DataFrame styling tool for HTML visualization

• Fast, expressive styling with a clean, chainable API
• Turn Polars DataFrames into clear, expressive HTML views
• Style using native Polars expressions
• Built for data inspection, debugging, and exploration
• Ready for reports, presentations, and sharing

---

Documentation

Full documentation: https://egayer.github.io/polarise/

---

Installation

pip install polarise

---

Quickstart

import polars as pl
import polarise

df = pl.DataFrame({
    "date": ["2024-01-01", "2024-01-02", "2024-01-03", "2024-01-04"],
    "region": ["EU", "EU", "US", "US"],
    "sales": [120, 85, 210, 250],
    "profit": [20, -15, 45, 55]
})

(
    df.style()
      .highlight_when(
          in_col="date",
          when=pl.col("profit") < 0,
          then_fill="alert_orange"
      )
      .gradient("sales", cmap="greens")
      .bar("profit", fill_pos="alert_green", fill_neg="alert_orange")
      .fashion_zebra()
      .show()
)

---

Where Polarise fits

Polarise is inspired by the styling capabilities of pandas, but built for a Polars workflow.

While Great Tables provides a rich and highly customizable system for building publication-quality tables, it comes with a more structured and declarative approach.

Polarise takes a different path:

• Lightweight and fast
• Fully aligned with Polars expressions
• Designed for quick inspection and clean presentation

It started as a simple tool to explore Polars DataFrames visually, and grew into a practical way to produce clear, styled HTML tables for reports and sharing — with optional LaTeX export for simple use cases.

At a glance

Feature	pandas Styler	Great Tables	Polarise
Ecosystem	pandas	Polars	Polars
Philosophy	Flexible, built-in	Rich, declarative	Lightweight, expressive
Best for	General styling	Publication workflows	Inspection & quick presentation
Syntax	pandas-based	Table grammar	Polars expressions
Complexity	Medium	High	Low
Speed (iteration)	Medium	Slower	Fast

---

License

This project is licensed under the GNU General Public License v3.0 — see the LICENSE file for details.