We propose a method to remove the large statistical fluctuations coming from the core regions when applying an all-electron Variational Monte Carlo method to a molecule. It is based on an efficient subsampling approach which performs sidewalks in these regions. The gain in variance is displayed for a series of atoms and we show that this gain is transferable to molecules on alkane chains and clusters of silicon. For these systems the gain in numerical efficiency is presented, and can be extrapolated to ∼ 22 and ∼ 420 for larger systems when computing the variational energy with a Slater determinant. These results are a proof of concept for numerically affordable all-electron quantum Monte Carlo calculations on molecules with large atomic charges.
We propose a method to remove the large statistical fluctuations coming from the core regions when applying an all-electron Variational Monte Carlo method to a molecule. It is based on an efficient subsampling approach which performs sidewalks in these regions. The gain in variance is displayed for a series of atoms and we show that this gain is transferable to molecules on alkane chains and clusters of silicon. For these systems the gain in numerical efficiency is presented, and can be extrapolated to ∼ 22 and ∼ 420 for larger systems when computing the variational energy with a Slater determinant. These results are a proof of concept for numerically affordable all-electron quantum Monte Carlo calculations on molecules with large atomic charges.
We propose a Monte Carlo method which exploits that core regions are physically independent in a molecule to almost remove their contribution to the numerical cost. The method is tantamount to computing an effective core potential on the fly, by efficiently subsampling the core regions with independent sidewalks. The removal of fluctuations in the core region enables also the dynamic in the valence region to be accelerated using a process with two time steps. As a function of the total number of electrons N the numerical overhead O ( N ) is negligible in comparison to the overall scaling O ( N 3 ) (due to the evaluation of determinants). Tests are presented on atoms, alkane chains, and clusters of silicons. We report a transferability of the parameters of the method from atoms to molecules, enabling a calibration using only single atoms. These tests display a gain in numerical efficiency between one and two orders of magnitude for large N .
We propose to compute physical properties by Monte Carlo calculations using conditional expectation values. The latter are obtained on top of the usual Monte Carlo sampling by partitioning the physical space in several subspaces or fragments, and subsampling each fragment (i.e., performing side walks) while freezing the environment. No bias is introduced and a zero-variance principle holds in the limit of separability, i.e., when the fragments are independent. In practice, the usual bottleneck of Monte Carlo calculations-the scaling of the statistical fluctuations as a function of the number of particles N-is relieved for extensive observables. We illustrate the method in variational Monte Carlo on the two-dimensional Hubbard model and on metallic hydrogen chains using Jastrow-Slater wave functions. A factor O(N) is gained in numerical efficiency.
A method is presented for using truncated, maximally localized Wannier functions to introduce sparsity into the Slater determinant part of the trial wave function in quantum Monte Carlo calculations. When combined with an efficient numerical evaluation of these localized orbitals, the dominant cost in the calculation, namely, the evaluation of the Slater determinant, scales linearly with system size. This technique is applied to accurate total energy calculation of hydrogenated silicon clusters and carbon fullerenes containing 20-1000 valence electrons.
In this work, we introduce three algorithmic improvements to reduce the cost and improve the scaling of orbital space variational Monte Carlo (VMC). First, we show that, by appropriately screening the one- and two-electron integrals of the Hamiltonian, one can improve the efficiency of the algorithm by several orders of magnitude. This improved efficiency comes with the added benefit that the cost of obtaining a constant error per electron scales as the second power of the system size O ( N 2 ), down from the fourth power O ( N 4 ). Using numerical results, we demonstrate that the practical scaling obtained is, in fact, O ( N 1.5 ) for a chain of hydrogen atoms. Second, we show that, by using the adaptive stochastic gradient descent algorithm called AMSGrad, one can optimize the wave function energies robustly and efficiently. Remarkably, AMSGrad is almost as inexpensive as the simple stochastic gradient descent but delivers a convergence rate that is comparable to that of the Stochastic Reconfiguration algorithm, which is significantly more expensive and has a worse scaling with the system size. Third, we introduce the use of the rejection-free continuous time Monte Carlo (CTMC) to sample the determinants. Unlike the first two improvements, CTMC does come at an overhead that the local energy must be calculated at every Monte Carlo step. However, this overhead is mitigated to a large extent because of the reduced scaling algorithm, which ensures that the asymptotic cost of calculating the local energy is equal to that of updating the walker. The resulting algorithm allows us to calculate the ground state energy of a chain of 160 hydrogen atoms using a wave function containing ∼2 × 10 5 variational parameters with an accuracy of 1 mE h /particle at a cost of just 25 CPU h, which when split over 2 nodes of 24 processors each amounts to only about half hour of wall time. This low cost coupled with embarrassing parallelizability of the VMC algorithm and great freedom in the forms of usable wave functions, represents a highly effective method for calculating the electronic structure of model and ab initio systems.
Variational quantum Monte Carlo (VMC) method applied to interacting electron systems has played a pivotal role for the studies of electronic states of atoms, molecules and condensed matters. First, the many-body wave functions optimized by the VMC method provide good guiding functions in more accurate, diffusion quantum Monte Carlo (DMC) calculations for identical systems. Second, the VMC calculations themselves can give a good upper bound for the exact ground-state energy of many-electron systems when an appropriate form for the variational wave function is employed. In addition to a conventional form for the variational wave function that is expressed by the product of the Jastrow factor and the single Slater determinant, recent investigations have found that the expression in terms of many Slater determinants can provide a very accurate description of electronic states, especially for intrinsically multi-reference systems, which corresponds to the configuration interaction (CI) expansion method in conventional quantum chemistry. Although the CI method needs an extremely large number of expansion coefficients as variational parameters, the number of parameters can be reduced efficiently by using an idea of (molecular) orbital pair correlations. The purpose of the present paper is to examine the feasibility and usefulness of this idea within the framework of VMC method with the Jastrow–Slater wave function, where the ground state of a water molecule is investigated as a test case. In this study we rely on a variational wave function composed of a product of the Jastrow and Slater parts as 1⁄4 J S; ð1Þ
In this work, we introduce three algorithmic improvements to reduce the cost and improve the scaling of orbital space variational Monte Carlo (VMC). First, we show that by appropriately screening the one- and two-electron integrals of the Hamiltonian one can improve the efficiency of the algorithm by several orders of magnitude. This improved efficiency comes with the added benefit that the resulting algorithm scales as the second power of the system size O(N2), down from the fourth power O(N4). Using numerical results, we demonstrate that the practical scaling obtained is in fact O(N1.5) for a chain of Hydrogen atoms, and O(N1.2) for the Hubbard model. Second, we introduce the use of the rejection-free continuous time Monte Carlo (CTMC) to sample the determinants. CTMC is usually prohibitively expensive because of the need to calculate a large number of intermediates. Here, we take advantage of the fact that these intermediates are already calculated during the evaluation of the local energy and consequently, just by storing them one can use the CTCM algorithm with virtually no overhead. Third, we show that by using the adaptive stochastic gradient descent algorithm called AMSGrad one can optimize the wavefunction energies robustly and efficiently. The combination of these three improvements allows us to calculate the ground state energy of a chain of 160 hydrogen atoms using a wavefunction containing ∼2×105 variational parameters with an accuracy of 1 mEh/particle at a cost of just 25 CPU hours, which when split over 2 nodes of 24 processors each amounts to only about half hour of wall time. This low cost coupled with embarrassing parallelizability of the VMC algorithm and great freedom in the forms of usable wavefunctions, represents a highly effective method for calculating the electronic structure of model and \emph{ab initio} systems.