PyTorch Wavelet Toolbox (ptwt)

ptwt brings wavelet transforms to PyTorch. The code is open-source, follow the link above to go to the source. This package is listed in the Python Package Index (PyPI). Its 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 systems hardware setup. Once PyTorch ist 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 the following thesis:

@article{JMLR:v25:23-0636,
author  = {Moritz Wolter and Felix Blanke and Jochen Garcke and Charles Tapley Hoyt},
title   = {ptwt - The PyTorch Wavelet Toolbox},
journal = {Journal of Machine Learning Research},
year    = {2024},
volume  = {25},
number  = {80},
pages   = {1--7},
url     = {http://jmlr.org/papers/v25/23-0636.html}
}