jupyturtle

Python Turtle graphics for Jupyter notebooks

For a quick demo, open the lab notebook.

To use on a local Jupyter Notebook, just download the jupyturtle.py file to the same folder where your notebook is saved, and import it.

(I need to figure out how to deploy on Binder because Jupyturtle is not on conda forge yet )

jupyturtle offers only a small subset of the features of Python's standard turtle module. But it has all the features required to solve the exercises in the book Think Python, Third Edition, by Allen Downey.

Credits

The idea and some of the code for this module came from Tolga Atam's ColabTurtle, which I discovered while reviewing a pre-print version of Downey's Think Python, Third Edition (O'Reilly, 2024).

Design

Atam's best idea was to use SVG for drawing, which makes the code simple and lightweight, requiring only the Python standard library and the ipython module that is always available in Jupyter.

This is a rewrite from scratch, using classes to model the turtle and the drawing. My goal was to make jupyturtle easier to understand and extend by encapsulating the state and avoiding code duplication.

I used metaprogramming techniques to build the procedural API with global functions like fd() to move the turtle. The techniques are easier to understand in the didactic project abacus.

Turtle on a canvas

jupyturtle2 is a fork that uses ipycanvas to draw pixels on an HTML canvas, instead of generating SVG, so it handles complex drawings better.

But jupyturtle2 has two main drawbacks:

• ipycanvas requires the binary dependencies numpy and pillow, which may be harder to install in some environments.

• the generated drawings are not saved with the notebook like most other output cells; this is also a limitation of ipycanvas

@ramalho