behaviz

behaviz is a modular scientific plotting library that renders the same high-level call on matplotlib, seaborn, or bokeh without changing your code. You describe what a plot should look like with a declarative spec; behaviz translates it to whichever backend is active.

import behaviz as bv

fig, ax = bv.plot_line(x, y)        # matplotlib by default
bv.set_renderer("bokeh")
fig, ax = bv.plot_line(x, y)        # same call, interactive bokeh figure

Why behaviz?

• Backend-agnostic. Write once, switch renderer with one line. Prototype in matplotlib, ship interactive bokeh, never rewrite plotting code.
• A single styling vocabulary. Scales, limits, ticks, grids, spines, fonts, legends — one PlotSpec describes them and every backend honors it.
• Loud, helpful errors. A declarative channel layer validates inputs and raises clear BehavizDataError messages instead of cryptic backend tracebacks.
• Reach the metal when you need to. Any native backend property is reachable through keyword overrides or a post_hook.

Where to go next

You want to…	Read
Install it	Installation
Make your first plot	Quickstart
Understand the model	Core concepts
See every plot type	Plotting overview
Plot from a DataFrame	Data input
One line per category	Grouping
Style a figure	The spec system
Reuse a look	Presets
Full API	API reference

---

behaviz is GPLv3. Source: github.com/kaancet/behaviz.