v0.3

What's Changed

• Use normalisation from paper and add PyTorch notebook example by @nikolasborrel in #9
• Use physical quantities in settings, improve data generators by @nikolasborrel in #10
• Fix missing init.py and package discovery for subpackages

Full Changelog: v0.2...v0.3

v0.2

What's Changed

• Refactor code to handle source directivity by @nikolasborrel in #1
• Restructure project setup using TOML by @nikolasborrel in #2
• Rename deeponet_room_acoustics -> deeponet_acoustics by @nikolasborrel in #4
• Add ruff linter and format code by @nikolasborrel in #5
• Uniform 1D/2D and 3D implementations by @nikolasborrel in #6
• Add test suite by @nikolasborrel in #7
• Normalize u pressure fields from data instead of assuming fixed range by @nikolasborrel in #8

Full Changelog: v0.1...v0.2

Code compatible with original publication datasets

All results from the paper 'Sound propagation in realistic interactive 3D scenes with parameterized sources using deep neural operators' by Borrel-Jensen et al. can be reproduced with this version.
If the trained models from https://doi.org/10.11583/DTU.24812004 are used, remember to comment out the lines

U = self.activation(U)
V = self.activation(V)

in line 191-192 inside the file models/networks_flax.py.