- Clone Repository
git clone https://github.com/Aditya001-max/Denoising-CFD-Simulation-Data-using-Deep-Learning.git
cd Denoising-CFD-Simulation-Data-using-Deep-Learning
- Install Requirements Make sure you have Python 3 and necessary dependencies (e.g., numpy, scipy, tensorflow/keras, or pytorch). Install using:
pip install -r requirements.txt
- Obtain and Organize Data Follow the repository’s instructions.md to download sample CFD datasets. Place the datasets into the data/ folder, mirroring the example structure. Your folder hierarchy should look like:
data/
noisy_simulation.npy
clean_simulation.npy
...
-
Run the Example Script Open example in python conda environment. Check and update file paths so that they point to your data in data/. Adjust RPCA parameters (like thresholds or decomposition settings) as per question.
-
Run the Example In MATLAB, run:
matlab
Or in Python:
python example.py
This script applies RPCA to noisy CFD data and visualizes the clean/low-rank output. 6. Apply to Your Own Data
- Add your CFD output (e.g., .npy, .mat) to the same folder.
- Ensure example.m or example.py references your files.
- Run the script to denoise your data and save outputs.
- Analyze Results Outputs typically include:
- Recovered low-rank fields (denoised phenomenon)
- Sparse components (noise or anomalies)
- Visualization plots or animated flows
- Save or Export Outputs