https://github.com/alisiahkoohi/groundwater
Groundwater flow equation: Finite-difference forward, adjoint, and gradient operators with PyTorch wrapper
https://github.com/alisiahkoohi/groundwater
automatic-differentiation darcy-flow finite-difference groundwater-flow inverse-problems pytorch
Last synced: about 1 year ago
JSON representation
Groundwater flow equation: Finite-difference forward, adjoint, and gradient operators with PyTorch wrapper
- Host: GitHub
- URL: https://github.com/alisiahkoohi/groundwater
- Owner: alisiahkoohi
- License: mit
- Created: 2024-09-16T23:33:16.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2024-09-26T21:09:49.000Z (over 1 year ago)
- Last Synced: 2025-03-15T16:42:02.501Z (about 1 year ago)
- Topics: automatic-differentiation, darcy-flow, finite-difference, groundwater-flow, inverse-problems, pytorch
- Language: Python
- Homepage:
- Size: 35.2 KB
- Stars: 0
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
Groundwater flow equation: Finite-difference forward, adjoint, and gradient operators with PyTorch wrapper
This repository contains the code for implementing the finite-difference
forward, adjoint, and gradient operators for the groundwater flow equation.
The operators are implemented via
[Devito](https://github.com/devitocodes/devito), a finite-difference
domain-specific language for solving partial differential equations that
generates optimized C code depending on the target architecture.
## Installation
Clone the repository and install the package in editable mode.
```bash
# Create a new conda environment (optional).
conda create --name groundwater python
conda activate groundwater
# Clone the repository and install the package in your Python environment.
git clone ttps://github.com/alisiahkoohi/groundwater
cd groundwater/
pip install -e .
```
Visit the [Devito installation
guide](https://www.devitoproject.org/download.html) for more information
on setting up the environment variables to fully utilize the
parallelization capabilities of Devito.
## Questions
Please contact [alisk@rice.edu](mailto:alisk@rice.edu) for questions.
## Authors
Ali Siahkoohi and Mathias Louboutin