PAUXY is a collection of Python implementations of AUXilliarY field quantum Monte Carlo algorithms with a focus on simplicity rather than speed.
PAUXY can currently:
- estimate ground state properties of real (ab-initio) and model (Hubbard + UEG) systems.
- perform phaseless and constrained path AFQMC.
- calculate expectation values and correlation functions using back propagation.
- calculate imaginary time correlation functions.
- perform simple data analysis.
Clone the repository
$ git clone https://github.com/pauxy-qmc/pauxy.git
and run the following in the top-level pauxy directory
$ python setup.py build_ext --inplace
You may also need to set your PYTHONPATH appropriately.
- python (>= 3.6)
- numpy (>= 0.19.1)
- scipy (>= 1.13.3)
- h5py (>= 2.7.1)
- mpi4py (>= 3.0.1)
- cython (>= 0.29.2)
To run the tests you will need pytest and pandas. To perform error analysis you will also need pyblock <https://github.com/jsspencer/pyblock>
Pauxy contains unit tests and short deterministic longer tests of full calculations. To the tests you can do:
$ pytest
In the main repository.
Notes on the underlying theory as well as documentation and tutorials are available at readthedocs.