/NNSpectrum

# Reconstructing spectral functions via automatic differentiation

Primary LanguageJupyter NotebookMIT LicenseMIT

Reconstructing spectral functions via automatic differentiation

Cite this work as,

L. Wang, S. Shi, and K. Zhou, Reconstructing Spectral Functions via Automatic Differentiation, ArXiv:2111.14760 [Hep-Lat, Physics:Hep-Ph] (2021).

Getting Started

The code requires Python >= 3.8 and PyTorch >= 1.2. You can configure on CPU machine and accelerate with a recent Nvidia GPU card.

Running the tests

Run juputer notebook to generate mock data. Using Index and noise to specify propagator data.

python NNspectrum1202.py  --Index 4 --noise 5

Authors

  • Lingxiao Wang - Construct codes and write the preprint paper - Homepage
  • Shuzhe Shi - Check results and provide physics guidance
  • Kai Zhou - Lead the project and complete the article.

License

This project is licensed under the MIT License - see the LICENSE file for details