Python implementation of the Split Bregman algorithm.
Sustainable Use License from n8n.
- Python >= 3.10
- NumPy
- SciPy
- PyWavelets
- CuPy (required, not installed automatically by this package)
Install the CuPy package that matches your CUDA version (e.g.,
cupy-cuda12x,cupy-cuda11x, etc.). - gradops (required; from GitHub)
Install the CuPy wheel matching your CUDA version.
This package depends on gradops, which is hosted on GitHub under kejang.
pip install "gradops @ git+https://github.com/kejang/gradops.git"If you are installing from GitHub:
pip install "splitbregman @ git+https://github.com/kejang/splitbregman.git"