yusizhuo/QMagen

Learning Effective Hamiltonian from thermodynamic data

we wish to propose a systematic approach of fitting thermodynamic data which is unbiased and more efficient.

Now we have a first version demo where we show a proof-of-principle case by learning the Hamiltonian of a XXZ Heisenberg chain.

jupyter notebook Demo.ipynb

Too run the demo, make sure you have these packages properly intstalled.

[1] Pytorch https://pytorch.org/get-started/locally/

pip install torch torchvision torchaudio

[2] Bayesian Optimization https://github.com/fmfn/BayesianOptimization

pip install bayesian-optimization

or

conda install -c conda-forge bayesian-optimization