Installation¶
Pre-built wheels¶
pyg4ometry runs on all mainstream platforms (Linux, MacOS and Windows) and requires at least Python 3.7.
The easiest and recommended way to install it is from PyPI:
pip install pyg4ometry
All C++ libraries required by pyg4ometry at runtime are bundled with the Python wheel.
A Docker image with the latest stable version of pyg4ometry can be downloaded from Docker Hub
$ docker run g4edge/pyg4ometry:latest python
Python 3.10.12 (main, Jun 11 2023, 05:26:28) [GCC 11.4.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import pyg4ometry
>>>
Building from source¶
In case you need an unstable version of pyg4ometry or you are interested in developing, some non-Python dependencies are required.
The following C/C++ libraries (and headers) are required to build the package from source:
The Visualisation toolkit (VTK) (including Python bindings)
Additionally, CMake is required for building pyg4ometry’s C++ extensions.
Tip
Check out the platform-specific scripts we use to obtain dependencies in our CI/CD workflow.
If you want to save yourself the hassle of installing these dependencies, we recommend building in our Ubuntu Docker image, available on Docker Hub:
$ docker run -it g4edge/ubuntu:latest
root@b260417746a1:~#
All Python dependencies will be automatically obtained by pip.
Note
It is worth choosing a Python version for which VTK bindings are available (see https://pypi.org/project/vtk/#files). For example, there are limited builds for M1 Mac (ARM64).
To install the package, simply run pip install from the root directory
$ git clone https://github.com/g4edge/pyg4ometry
$ cd pyg4ometry
$ pip install .
Tip
The building uses CMake through scikit-build-core. Build errors or warnings are typically from cmake and can help determining the missing dependency.
Note
If you are building for windows you are going to need Visual Studio and the appropriate command line tools.
Warning
the package is built in parallel by default and is memory intensive! Make sure you have enough computational resources. If not, consult the scikit-build-core documentation to learn how to reduce the number of threads.