If you’ve been following the typesetting world lately, you know Typst is the cool new kid on the block, giving LaTeX a serious run for its money. But while Typst is fast and modern, data scientists have been waiting for a truly declarative way to build charts within it. Enter Gribouille. Developed by Mickaël Canouil, this package brings the "Grammar of Graphics" philosophy—the same logic that powers R’s legendary ggplot2—directly into the Typst ecosystem.
A Layered Approach to Data
Most plotting tools feel like a series of disconnected commands. Gribouille changes that by treating a chart as a series of layers. You define your data, map variables to aesthetics (like x-coordinates, y-coordinates, or colors), and add "geoms" like points, lines, or bars.
With the release of version 0.3.0, this workflow is getting even more refined. It’s not just about drawing lines; it’s about a structured language for visualization that lives natively within your document source code. Instead of exporting a PNG from another tool and hoping the fonts match, Gribouille lets you render high-quality, themed graphics as part of your Typst compilation process.
What’s New in 0.3.0: Precision and Theming
The latest update isn't just a minor patch; it’s a significant quality-of-life boost for power users. Canouil has introduced much finer control over guides and axes, allowing users to toggle them on or off with surgical precision. This is a huge win for those of us trying to create clean, minimalist layouts where every pixel of whitespace matters.
One of the most practical changes is that geom-area() now stacks by default, making it easier to create stacked area charts without extra boilerplate code. Furthermore, the compose() function can now carry its own theme. This means your complex, multi-plot layouts can maintain visual consistency effortlessly, even if you’re mixing different types of data visualizations in a single figure.
The Future of Technical Publishing
As Typst itself matures—with the recent 0.15 update adding variable fonts and multi-file output—packages like Gribouille are turning it into a legitimate powerhouse for technical publishing. We’re moving toward a world where your papers, reports, and data visualizations are all part of one seamless, high-performance pipeline. If you’re looking to ditch the clunkiness of old-school plotting for something more elegant, Gribouille 0.3.0 is a fantastic place to start.
Sources
Media
