quatplot

Interactive 3D plots for Jupyter with quaternion-based view controls — no gimbal lock.

Quaternion math is powered by ganja.js (Algebra(0,2) ≅ ℍ) when its CDN is reachable, with a built-in quaternion fallback otherwise. Rendering is plain HTML5 canvas inside an <iframe srcdoc>, so plots are self-contained: multiple plots per notebook can't collide on IDs, CSS, or globals, and no notebook-side JavaScript is required.

Features

• Surface plots from 2D mesh data (colored fill and/or wireframe, toggleable)
• 3D scatter plots with colored markers and hover coordinate tooltips
• Optional color bar for scatter values
• Isometric ⇄ perspective projection toggle
• Plotly-style mouse controls: drag to rotate, wheel to zoom, shift-drag / right-drag to pan
• Top view button (straight down, x right / y up — isometric top view reads as a 2D plot)
• Auto-scaled, labeled axes with nice-number tick marks
• Light / dark themes; Viridis, Plasma, Jet, Coolwarm, Grayscale colormaps
• PNG screenshot button
• Resizable plots in Jupyter: drag the bottom-right corner
• Standalone HTML export

Install

pip install quatplot

Or the latest development version straight from GitHub:

pip install git+https://github.com/drscotthawley/quatplot.git

Usage

import numpy as np
import quatplot as qp

x = np.linspace(-3, 3, 41)
y = np.linspace(-3, 3, 41)
X, Y = np.meshgrid(x, y)
Z = np.sin(X) * np.cos(Y)

qp.plot(qp.Surface(x, y, Z))                       # in a notebook cell

qp.plot(qp.Scatter3(px, py, pz, v=values),         # scatter, colored by v
        colorbar=True, cmap="Plasma")

qp.plot(qp.Surface(x, y, Z), qp.Scatter3(px, py, pz),  # overlay traces
        theme="light", proj="perspective", height=600)

qp.save("plot.html", qp.Surface(x, y, Z))          # standalone HTML file

plot() options: theme ("dark"/"light"), proj ("iso"/"perspective"), cmap, fill, wire, colorbar, height (initial px; plots are corner-drag resizable afterward).

See the demo notebook for a working tour.

Screenshot

Development

git clone https://github.com/drscotthawley/quatplot.git
cd quatplot
pip install -e .

The entire library is one file, quatplot.py; the HTML/JS template embedded in it is the single source of truth for both notebook rendering and HTML export.

License

MIT — see LICENSE.