GoldenViz documentation
GoldenViz is a lightweight quality-checking library for Matplotlib charts. It focuses on 25 Golden Rules of data visualization and is built for Jupyter notebook workflows as well as explicit manual checks.
Why GoldenViz exists
The idea behind GoldenViz comes from teaching data visualization at Universite Paris Cite. In that setting, one observation became clear: generating a chart was rarely the main obstacle. With coding assistant tools, examples, and online documentation, students could usually produce a figure.
The harder part was understanding whether the chart was good, why it needed corrections, what was misleading or unclear, and what concrete changes would make it better.
A chart is not finished when it renders. It is finished when a reader can understand the message quickly, trust the encoding, and spot the important pattern without unnecessary effort.
Most visualization problems are not syntax problems. They are judgment problems: missing labels, misleading scales, cluttered legends, too many categories, unreadable text, inconsistent colors, or visual effects that distract from the data.
GoldenViz turns those judgment checks into repeatable feedback. It inspects a figure and reports issues as pass, warning, or fail results, so charts can be improved before they appear in notebooks, reports, dashboards, papers, or presentations.
The goal is not to replace the analyst or designer. The goal is to catch common mistakes early and make good visualization habits easier to apply.
The GoldenViz approach
GoldenViz is based on 25 practical rules covering chart purpose, titles, axes, scales, units, baselines, readability, visual clutter, color, legends, direct labeling, category handling, uncertainty, precision, dates, diverging values, and visual economy.
Each rule asks one question: does this chart help the reader understand the data clearly and honestly?
What makes a good chart?
A good chart should:
have a clear message
use the right visual encoding for the data
label what the reader needs to know
avoid decoration that competes with the signal
make comparisons easy
use color deliberately
show uncertainty or scale when it matters
respect the reader’s time
GoldenViz does not enforce one visual style. It encourages charts that are readable, honest, and useful.