PyQMC: An all-Python real-space quantum Monte Carlo module in PySCF

Abstract

We describe a new open-source Python-based package for high accuracy correlated electron calculations using quantum Monte Carlo (QMC) in real space: PyQMC. PyQMC implements modern versions of QMC algorithms in an accessible format, enabling algorithmic development and easy implementation of complex workflows. Tight integration with the PySCF environment allows for a simple comparison between QMC calculations and other many-body wave function techniques, as well as access to high accuracy trial wave functions.

Ensemble variational Monte Carlo for optimization of correlated excited state wave functions

“…where F is a Fisher information matrix [17,18] and τ is the descent step length that can be optimized using correlated sampling [24] or reduced on a schedule. In this version of the algorithm, where the parameter gradients are given by equation ( 30), we can consider the time step to be scaled by the weight, τ = τ /w j .…”

“…The HF and CASCI calculations were carried out in PySCF. We note that by using a Jastrow factor, we achieve accurate wave functions with orders of magnitude fewer determinants than typically needed in CASCI calculations [24], allowing us to use unusually small CASCI calculations in this step. We initialize our three states for ensemble optimization (equation ( 1)) from the CASCI determinant coefficients and a pre-optimized Jastrow factor.…”

“…Vertical excitation energies are evaluated from the optimized wave functions using VMC with 10 000 walkers. The Jastrow optimization and all ensemble optimization calculations were carried out in PyQMC [24].…”

QMCTorch: a PyTorch Implementation of Real-Space Quantum Monte Carlo Simulations of Molecular Systems

“…Many software such as QMCPack (Kim et al, 2018), QMC=Chem (Scemama et al, 2013), CHAMP (C. Filippi, 2019) provide high-quality implementation of advanced QMC methodologies in low-level languages (C++/Fortran). Python implementations of QMC such as PAUXY (Fionn Malone, n.d.) and PyQMC (Wheeler et al, 2023) have also been proposed to facilitate the use and development of QMC techniques. Large efforts have been made to leverage recent development of deep learning techniques for QMC simulations with for example the creation of neural-network based wave-function ansatz (Choo et al, 2020;Han et al, 2019;Hermann et al, 2020;Inui et al, 2021;Kessler et al, 2021;Lin et al, 2023;Pfau et al, 2020;Schätzle et al, 2021;Yang et al, 2020) that have lead to very interesting results.…”