JupyThree

A Jupyter widget based on pythreejs to display massive pointclouds (millions of points) with excellent interactive speed. Also handles triangular meshes, with easy per-vertex color or scalar field.

import numpy as np
from jupythree.pointcloud import pointcloud

N = 2000000 #<-- two million points
x = np.random.random((N,3))/2 - 0.25 # cube

pointcloud(x, color=x[:,1]).show()

Installation

Clone this repo and install requirements

git clone https://github.com/groscot/jupythree
cd jupythree
pip install -r requirements.txt

Installation through pip (pip install ...) will come soon 🙏

Usage

Run the included Notebook Jupythree Demo.ipynb

Requirements

• pythreejs
• numpy
• ipywidgets
• matplotlib (for colormaps)

Some nice features:

1. Handles millions of points while keeping interactive rotation and zoom
2. RGB colors + real values with matplotlib colormaps
3. Slider controlling the radius: pointcloud(x).with_slider().show()
4. Realtime animations: create and display the pointcloud with one cell (handle = pointcloud(x1); handle.show()), animate it from any other cell of your notebook (handle.update(pc=x2))
5. Triangular meshes with vertex colors: mesh(V, F, vertex_color=V[:,0]).show()

Known issues:

In recent versions of Jupyter, you will have a warning when displaying the first pointcloud of a session. It affects other libraries based on WebGL (such as plotly). You can safely ignore this message.

UserWarning: Message serialization failed with:
Out of range float values are not JSON compliant
Supporting this message is deprecated in jupyter-client 7, please make sure your message is JSON-compliant
  content = self.pack(content)