Vectorized matrix, quaternion, euler, axis-angle and vector math for CG.
Everything is numpy-native, broadcasts over arrays of transforms, and is
numba-accelerated where it matters. No DCC required.
Install
pip install https://github.com/Eric-Vignola/transforms/releases/download/1.0.0/transforms-1.0.0-py3-none-any.whl
Or straight from the repository:
pip install git+https://github.com/Eric-Vignola/transforms
Python 3.9 or newer.
Dependencies
numpy, scipy, numba.
What moved in this release
The tag now points at the packaged source. Previously the repository carried no
build metadata at all, so neither pip install git+… nor a wheel was possible;
it could only be used by dropping the folder onto sys.path. It is now a
normal installable distribution.
Line endings are pinned to LF through a committed .gitattributes, so a
checkout reads the same on Windows, macOS and Linux.