Awesome

Capytaine standalone distribution

This repository stores the recipe to bundle a full Python distribution with Capytaine into a single file executable file.

Download

Choose the file corresponding to your OS in the "Release" section of the Github page.

After downloading the file, you might need to give it permission to be executed as a program. On Linux, the command is as follows:

chmod +x ipython-with-capytaine-linux

Usage

In a terminal, use the following command to open an IPython shell with Capytaine pre-loaded:

./ipython-with-capytaine-linux

In this shell, you can run any Python command, including commands using Capytaine:

In [1]: print("Hello world")
Hello world

In [2]: mesh = cpt.mesh_sphere()
Out[2]: Mesh(vertices=[[... 92 vertices ...]], faces=[[... 100 faces ...]], name="sphere_0")

To execute a whole Python script, you can pass it as argument to the command line

./ipython-with-capytaine-linux radiation_cylinder.py

or, from within the IPython shell, use the following "magic" command:

In [3]: %run radiation_cylinder.py

Limitations

• Interactive display of Matplotlib plots (plt.show()) seems to be broken. But saving the plot to a file (plt.savefig('example.pdf')) works.

• You cannot install other Python libraries in the bundled environment. If you have specific need, you will need to use pip or conda to install Capytaine in a virtual environment that you can customize.

• On MacOS, the standalone version is only built for the ARM architecture (Apple Silicon processors), since Capytaine 2.2. Packages for the legacy MacOS Intel architecture are still available on pip or conda.

Support

Please report issues to Capytaine's main issue tracker while mentionning [capytaine-standalone] in the issue title.