arXiv subjects

Werner Krauth

Publications and source records attributed to Werner Krauth.

At least 37 records · Page 2Linked to original sources

arXiv2021

Large-scale dynamics of event-chain Monte Carlo

Event-chain Monte Carlo (ECMC) accelerates the sampling of hard-sphere systems, and has been generalized to the potentials used in classical molecular simulation. Rather than imposing detailed balance on transition probabilities, the method enforces a weaker global-balance condition in order to guarantee convergence to equilibrium. In this paper we generalize the factor-field variant of ECMC to higher space dimensions. In the two-dimensional fluid phase, factor-field ECMC saturates the lower bound $z=0$ for the dynamical scaling exponent for local dynamics, whereas molecular dynamics is characterized by $z=1$ and local Metropolis Monte Carlo by $z=2$. In the presence of hexatic order, factor fields are not found to speed up the convergence. We indicate applications to the physics of glasses, and note that generalizations of factor fields could couple to orientational order.

cond-mat.stat-mech↗

arXiv2021

Sparse hard-disk packings and local Markov chains

We propose locally stable sparse hard-disk packings, as introduced by Böröczky, as a model for the analysis and benchmarking of Markov-chain Monte Carlo (MCMC) algorithms. We first generate such packings in a square box with periodic boundary conditions and analyze their properties. We then study how local MCMC algorithms, namely the Metropolis algorithm and several versions of event-chain Monte Carlo (ECMC), escape from configurations that are obtained by slightly reducing all disk radii by a relaxation parameter. A scaling analysis is confirmed by simulation results. We obtain two classes of ECMC, one in which the escape time varies algebraically with the relaxation parameter (as for the local Metropolis algorithm) and another in which the escape time scales as the logarithm of the relaxation parameter. We discuss the connectivity of the hard-disk sample space, the ergodicity of local MCMC algorithms, as well as the meaning of packings in the context of the NPT ensemble. Our work is accompanied by open-source, arbitrary-precision software for Böröczky packings (in Python) and for straight, reflective, forward, and Newtonian ECMC (in Go).

cond-mat.stat-mech↗

arXiv2021

Event-chain Monte Carlo: foundations, applications, and prospects

This review treats the mathematical and algorithmic foundations of non-reversible Markov chains in the context of event-chain Monte Carlo (ECMC), a continuous-time lifted Markov chain that employs the factorized Metropolis algorithm. It analyzes a number of model applications, and then reviews the formulation as well as the performance of ECMC in key models in statistical physics. Finally, the review reports on an ongoing initiative to apply the method to the sampling problem in molecular simulation, that is, to real-world models of peptides, proteins, and polymers in aqueous solution.

cond-mat.soft↗

arXiv2020

Multithreaded event-chain Monte Carlo with local times

We present a multithreaded event-chain Monte Carlo algorithm (ECMC) for hard spheres. Threads synchronize at infrequent breakpoints and otherwise scan for local horizon violations. Using a mapping onto absorbing Markov chains, we rigorously prove the correctness of a sequential-consistency implementation for small test suites. On x86 and ARM processors, a C++ (OpenMP) implementation that uses compare-and-swap primitives for data access achieves considerable speed-up with respect to single-threaded code. The generalized birthday problem suggests that for the number of threads scaling as the square root of the number of spheres, the horizon-violation probability remains small for a fixed simulation time. We provide C++ and Python open-source code that reproduces all our results.

physics.comp-ph↗

arXiv2019

JeLLyFysh-Version1.0 -- a Python application for all-atom event-chain Monte Carlo

We present JeLLyFysh-Version1.0, an open-source Python application for event-chain Monte Carlo (ECMC), an event-driven irreversible Markov-chain Monte Carlo algorithm for classical N-body simulations in statistical mechanics, biophysics and electrochemistry. The application's architecture closely mirrors the mathematical formulation of ECMC. Local potentials, long-ranged Coulomb interactions and multi-body bending potentials are covered, as well as bounding potentials and cell systems including the cell-veto algorithm. Configuration files illustrate a number of specific implementations for interacting atoms, dipoles, and water molecules.

physics.comp-ph↗

arXiv2018

A kinetic-Monte Carlo perspective on active matter

We study non-equilibrium phases for interacting two-dimensional self-propelled particles with isotropic pair-wise interactions using a persistent kinetic Monte Carlo (MC) approach. We establish the quantitative phase diagram, including the motility-induced phase separation (MIPS) that is a commonly observed collective phenomena in active matter. In addition, we demonstrate for several different potential forms the presence of two-step melting, with an intermediate hexatic phase, in regions far from equilibrium. Increased activity can melt a two-dimensional solid and the melting lines remain disjoint from MIPS. We establish this phase diagram for a range of the inter-particle potential stiffnesses, and identify the MIPS phase even in the hard-disk limit. We establish that the full description of the phase behavior requires three independent control parameters.

cond-mat.soft↗

arXiv2018

Event-chain Monte Carlo with factor fields

We study the dynamics of one-dimensional (1D) interacting particles simulated with the event-chain Monte Carlo algorithm (ECMC). We argue that previous versions of the algorithm suffer from a mismatch in the factor potential between different particle pairs (factors) and show that in 1D models, this mismatch is overcome by factor fields. ECMC with factor fields is motivated, in 1D, for the harmonic model, and validated for the Lennard-Jones model as well as for hard spheres. In 1D particle systems with short-range interactions, autocorrelation times generally scale with the second power of the system size for reversible Monte Carlo dynamics, and with its first power for regular ECMC and for molecular-dynamics. We show, using numerical simulations, that they grow only with the square root of the systems size for ECMC with factor fields. Mixing times, which bound the time to reach equilibrium from an arbitrary initial configuration, grow with the first power of the system size.

cond-mat.stat-mech↗

arXiv2018

Mixing and perfect sampling in one-dimensional particle systems

We study the approach to equilibrium of the event-chain Monte Carlo (ECMC) algorithm for the one-dimensional hard-sphere model. Using the connection to the coupon-collector problem, we prove that a specific version of this local irreversible Markov chain realizes perfect sampling in O(N^2 log N) events, whereas the reversible local Metropolis algorithm requires O(N^3 log N) time steps for mixing. This confirms a special case of an earlier conjecture about O(N^2 log N) scaling of mixing times of ECMC and of the forward Metropolis algorithm, its discretized variant. We furthermore prove that sequential ECMC (with swaps) realizes perfect sampling in O(N^2) events. Numerical simulations indicate a cross-over towards O(N^2 log N) mixing for the sequential forward swap Metropolis algorithm, that we introduce here. We point out open mathematical questions and possible applications of our findings to higher-dimensional statistical-physics models.

cond-mat.stat-mech↗

arXiv2018

All-atom computations with irreversible Markov chains

The event-chain Monte Carlo (ECMC) method is an irreversible Markov process based on the factorized Metropolis filter and the concept of lifted Markov chains. Here, ECMC is applied to all-atom models of multi-particle interactions that include the long-ranged Coulomb potential. We discuss a line-charge model for the Coulomb potential and demonstrate its equivalence with the standard Coulomb model with tin-foil boundary conditions. Efficient factorization schemes for the potentials used in all-atom water models are presented, before we discuss the best choice for lifting schemes for factors of more than three particles. The factorization and lifting schemes are then applied to simulations of point-charge and charged-dipole Coulomb gases, as well as to small systems of liquid water. For a locally charge-neutral system in three dimensions, the algorithmic complexity is O(N log N) in the number N of particles. In ECMC, a Particle-Particle method, it is achieved without the interpolating mesh required for the efficient implementation of other modern Coulomb algorithms. An event-driven, cell-veto-based implementation samples the equilibrium Boltzmann distribution using neither time-step approximations nor spatial cutoffs on the range of the interaction potentials. We discuss prospects and challenges for ECMC in soft condensed-matter and biological physics.

cond-mat.stat-mech↗

arXiv2018

Thermodynamic phases in two-dimensional active matter

Active matter has been intensely studied for its wealth of intriguing properties such as collective motion, motility-induced phase separation (MIPS), and giant fluctuations away from criticality. However, the precise connection of active materials with their equilibrium counterparts has remained unclear. For two-dimensional (2D) systems, this is also because the experimental and theoretical understanding of the liquid, hexatic, and solid equilibrium phases and their phase transitions is very recent. Here, we use self-propelled particles with inverse-power-law repulsions (but without alignment interactions) as a minimal model for 2D active materials. A kinetic Monte Carlo (MC) algorithm allows us to map out the complete quantitative phase diagram. We demonstrate that the active system preserves all equilibrium phases, and that phase transitions are shifted to higher densities as a function of activity. The two-step melting scenario is maintained. At high activity, a critical point opens up a gas-liquid MIPS region. We expect that the independent appearance of two-step melting and of MIPS is generic for a large class of two-dimensional active systems.

cond-mat.stat-mech↗

arXiv2018

Irreversible local Markov chains with rapid convergence towards equilibrium

We study the continuous one-dimensional hard-sphere model and present irreversible local Markov chains that mix on faster time scales than the reversible heatbath or Metropolis algorithms. The mixing time scales appear to fall into two distinct universality classes, both faster than for reversible local Markov chains. The event-chain algorithm, the infinitesimal limit of one of these Markov chains, belongs to the class presenting the fastest decay. For the lattice-gas limit of the hard-sphere model, reversible local Markov chains correspond to the symmetric simple exclusion process (SEP) with periodic boundary conditions. The two universality classes for irreversible Markov chains are realized by the totally asymmetric simple exclusion process (TASEP), and by a faster variant (lifted TASEP) that we propose here. Lifted Markov chains and the recently introduced factorized Metropolis acceptance rule extend the irreversible Markov chains discussed here to general pair interactions and to higher dimensions.

cond-mat.stat-mech↗

arXiv2017

Irreversible Markov chains in spin models: Topological excitations

We analyze the convergence of the irreversible event-chain Monte Carlo algorithm for continuous spin models in the presence of topological excitations. In the two-dimensional XY model, we show that the local nature of the Markov-chain dynamics leads to slow decay of vortex-antivortex correlations while spin waves decorrelate very quickly. Using a Frechet description of the maximum vortex-antivortex distance, we quantify the contributions of topological excitations to the equilibrium correlations, and show that they vary from a dynamical critical exponent z \sim 2 at the critical temperature to z \sim 0 in the limit of zero temperature. We confirm the event-chain algorithm's fast relaxation (corresponding to z = 0) of spin waves in the harmonic approximation to the XY model. Mixing times (describing the approach towards equilibrium from the least favorable initial state) however remain much larger than equilibrium correlation times at low temperatures. We also describe the respective influence of topological monopole-antimonopole excitations and of spin waves on the event-chain dynamics in the three-dimensional Heisenberg model.

cond-mat.stat-mech↗

arXiv2017

Solid--liquid transition of skyrmions in a two-dimensional chiral magnet

We study the melting of skyrmions in a two-dimensional Heisenberg chiral magnet with bi-axial Dzyaloshinskii--Moriya interactions. These topological excitations may form at zero temperature a triangular crystal with long-range positional order. However, we show using large-scale Monte Carlo simulations that at small finite temperature, the skyrmions rather form a typical two-dimensional solid: Positional correlations decay with distance as power laws while the orientational correlations remain finite. At higher temperature, we observe a direct transition from this two-dimensional solid to a liquid with short-range correlations. This differs from generic two-dimensional homogeneous particle systems, where a hexatic phase is realized between the solid and the liquid.

cond-mat.stat-mech↗

arXiv2016

Momentum distribution in the unitary Bose gas from first principles

We consider a realistic bosonic N-particle model with unitary interactions relevant for Efimov physics. Using quantum Monte Carlo methods, we find that the critical temperature for Bose-Einstein condensation is decreased with respect to the ideal Bose gas. We also determine the full momentum distribution of the gas, including its universal asymptotic behavior, and compare this crucial observable to recent experimental data. Similar to the experiments with different atomic species, differentiated solely by a three-body length scale, our model only depends on a single parameter. We establish a weak influence of this parameter on physical observables. In current experiments, the thermodynamic instability of our model from the atomic gas towards an Efimov liquid could be masked by the dynamical instability due to three-body losses.

cond-mat.quant-gas↗

arXiv2016

Cell-veto Monte Carlo algorithm for long-range systems

We present a rigorous efficient event-chain Monte Carlo algorithm for long-range interacting particle systems. Using a cell-veto scheme within the factorized Metropolis algorithm, we compute each single-particle move with a fixed number of operations. For slowly decaying potentials such as Coulomb interactions, screening line charges allow us to take into account periodic boundary conditions. We discuss the performance of the cell-veto Monte Carlo algorithm for general inverse-power-law potentials, and illustrate how it provides a new outlook on one of the prominent bottlenecks in large-scale atomistic Monte Carlo simulations.

cond-mat.stat-mech↗

arXiv2015

Event-chain algorithm for the Heisenberg model: Evidence for $z \simeq 1$ dynamic scaling

We apply the event-chain Monte Carlo algorithm to the three-dimensional ferromagnetic Heisenberg model. The algorithm is rejection-free and also realizes an irreversible Markov chain that satisfies global balance. The autocorrelation functions of the magnetic susceptibility and the energy indicate a dynamical critical exponent $z \approx 1$ at the critical temperature, while that of the magnetization does not measure the performance of the algorithm. This seems to be the first report that the event-chain Monte Carlo algorithm substantially reduces the dynamical critical exponent from the conventional value of $z\simeq 2$.

cond-mat.stat-mech↗

arXiv2015

Hard-sphere melting and crystallization with event-chain Monte Carlo

We simulate crystallization and melting with local Monte Carlo (LMC), event-chain Monte Carlo (ECMC), and with event-driven molecular dynamics (EDMD) in systems with up to one million three-dimensional hard spheres. We illustrate that our implementations of the three algorithms rigorously coincide in their equilibrium properties. We then study nucleation in the NVE ensemble from the fcc crystal into the homogeneous liquid phase and from the liquid into the homogeneous crystal. ECMC and EDMD both approach equilibrium orders of magnitude faster than LMC. ECMC is also notably faster than EDMD, especially for the equilibration into a crystal from a disordered initial condition at high density. ECMC can be trivially implemented for hard-sphere and for soft-sphere potentials, and we suggest possible applications of this algorithm for studying jamming and the physics of glasses, as well as disordered systems.

cond-mat.stat-mech↗

arXiv2015

Event-chain Monte Carlo for classical continuous spin models

We apply the event-chain Monte Carlo algorithm to classical continuum spin models on a lattice and clarify the condition for its validity. In the two-dimensional XY model, it outperforms the local Monte Carlo algorithm by two orders of magnitude, although it remains slower than the Wolff cluster algorithm. In the three-dimensional XY spin glass at low temperature, the event-chain algorithm is far superior to the other algorithms.

cond-mat.stat-mech↗