PyTorch Wavelet Toolbox (ptwt)

ptwt brings wavelet transforms to PyTorch. The code is open-source. Follow the GitHub link above to go to the source. This package is listed in the Python Package Index (PyPI). It’s best installed via pip. GPU support depends on PyTorch. To take advantage of GPU-processing, follow the PyTorch install instructions. Install the version that best suits your system’s hardware setup. Once PyTorch is set up, type the following to get started:

pip install ptwt

This documentation aims to explain the foundations of wavelet theory, introduce the ptwt package by example, and deliver a complete documentation of all functions. Readers who are already familiar with the theory should directly jump to the examples or the API documentation using the navigation on the left.

ptwt is built to be PyWavelets compatible. It should be possible to switch back and forth with relative ease.

If you use this work in a scientific context, please cite us! We have a BibTeX entry on the Citation-page.