mVMC - Open-source software for many-variable variational Monte Carlo method
Abstract
mVMC (many-variable Variational Monte Carlo) is an open-source software based on the variational Monte Carlo method applicable for a wide range of Hamiltonians for interacting fermion systems. In mVMC, we introduce more than ten thousands variational parameters and simultaneously optimize them by using the stochastic reconfiguration (SR) method. In this paper, we explain basics and user interfaces of mVMC. By using mVMC, users can perform the calculation by preparing only one input file of about ten lines for widely studied quantum lattice models, and can also perform it for general Hamiltonians by preparing several additional input files. We show the benchmark results of mVMC for the Hubbard model, the Heisenberg model, and the Kondo-lattice model. These benchmark results demonstrate that mVMC provides ground-state and low-energy-excited-state wave functions for interacting fermion systems with high accuracy.
Cite
@article{arxiv.1711.11418,
title = {mVMC - Open-source software for many-variable variational Monte Carlo method},
author = {Takahiro Misawa and Satoshi Morita and Kazuyoshi Yoshimi and Mitsuaki Kawamura and Yuichi Motoyama and Kota Ido and Takahiro Ohgoe and Masatoshi Imada and Takeo Kato},
journal= {arXiv preprint arXiv:1711.11418},
year = {2019}
}
Comments
19 pages, 8 figures, source code can be found at https://ma.issp.u-tokyo.ac.jp/en/app/518 and https://github.com/issp-center-dev/mVMC, accepted for publication in Computer Physics Communications