The repository provides Python codes for the numerical examples illustrated in the paper titled ‘Harnessing physics-informed operators for high-dimensional reliability analysis problems’ Please go through the paper to understand the implemented algorithm. Requirements:
-
Install Python packages pytorch, numpy, pandas, matplotlib, etc, and other prerequisite packages (for more details, visit: https://github.com/TapasTripura/WNO).
-
There are four separate folders containing data sets (data generation codes) and implementation codes, where the folders are named as ‘Diffusion-reaction system’, ‘Impulse transmission in nerve’, ‘Fluid flow through a porous medium’, and ‘Phase transitions in alloys’.
-
Add the data (generated data) path to load the data and use the run(PIWNO.py) file to execute the program.
-
If you find the code helpful, please cite the papers. @article{navaneeth2024harnessing, title={Harnessing physics-informed operators for high-dimensional reliability analysis problems}, author={Navaneeth, N and Chakraborty, Souvik and others}, journal={arXiv preprint arXiv:2409.04708}, year={2024} }
@article{navaneeth2024physics, title={Physics informed WNO}, author={Navaneeth, N and Tripura, Tapas and Chakraborty, Souvik}, journal={Computer Methods in Applied Mechanics and Engineering}, volume={418}, pages={116546}, year={2024}, publisher={Elsevier} }