Skip to content
/ gf2ML Public

Predicting interacting Green functions with ML

Notifications You must be signed in to change notification settings

EgorcaA/gf2ML

Repository files navigation

About The Project

This tool is designed to predict interacting Green functions by noninteracting ones.
One can build a database of Green's functions, train different models and use them to predict G, obtain density of states.

(back to top)

Prerequisites

  • triqs_cthyb
    mamba install conda-forge::triqs_cthyb
  • triqs
mamba install -c conda-forge triqs
  • PyTorch
    conda create --name triqs python pytorch cpuonly -c pytorch
  • triqs-maxent
pip install triqs-maxent

(back to top)

Usage

Notebooks:

  • *sample_analize *:Plot G, G_0 and other features of single sample
  • gf2ML: Build, calculate and plot samples from the database
  • train_model: Train ML model to predict G by G_0

sample.png

(back to top)

Contact

Egor Agapov - agapov.em@phystech.edu

Project Link: https://github.com/EgorcaA/gf2ML

(back to top)

About

Predicting interacting Green functions with ML

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published