Random Cluster Model Research Articles

Complete graph asymptotics for the Ising and random-cluster models on five-dimensional grids with a cyclic boundary.

The finite-size scaling behavior for the Ising model in five dimensions, with either free or cyclic boundary, has been the subject of a long-running debate. The older papers have been based on ideas from, e.g., field theory or renormalization. In this paper we propose a detailed and exact scaling picture for critical region of the model with cyclic boundary. Unlike the previous papers our approach is based on a comparison with the existing exact and rigorous results for the FK-random-cluster model on a complete graph. Based on those results we predict several distinct scaling regions in an L-dependent window around the critical point. We test these predictions by comparing with data from Monte Carlo simulations and find a good agreement. The main feature which differs between the complete graph and the five-dimensional model with free boundary is the existence of a bimodal energy distribution near the critical point in the latter. This feature was found by the same authors in an earlier paper in the form of a quasi-first-order phase transition for the same Ising model.

Perfect simulation for the infinite random cluster model, Ising and Potts models at low or high temperature

In this article we create a new algorithm for the perfect simulation of the infinite random cluster model for a sufficiently small or a sufficiently high value of the parameters. This implies the simulation of the Ising and Potts models with free boundary conditions.

Infinite volume continuum random cluster model

The continuum random cluster model is defined as a Gibbs modification of the stationary Boolean model in $\mathbb{R}^d$ with intensity $z \gt 0$ and the law of radii $Q$. The formal unormalized density is given by $q^{N_{cc}}$ where $q \gt 0$ is a fixed parameter and $N_{cc}$ the number of connected components in the random germ-grain structure. In this paper we prove the existence of the model in the infinite volume regime for a large class of parameters including the case $q \lt 1$ or distributions $Q$ without compact support. In the extreme setting of non integrable radii (i.e. $\int R^d Q(dR)=\infty$) and $q$ is an integer larger than 1, we prove that for $z$ small enough the continuum random cluster model is not unique; two different probability measures solve the DLR equations. We conjecture that the uniqueness is recovered for $z$ large enough which would provide a phase transition result. Heuristic arguments are given. Our main tools are the compactness of level sets of the specific entropy, a fine study of the quasi locality of the Gibbs kernels and a Fortuin-Kasteleyn representation via Widom-Rowlinson models with random radii.

Delaunay sets and condensed matter: The dialogue continues

Since the discovery of quasicrystals began to percolate through the scientific community, thirty years ago, Delaunay sets have been the tool of choice for describing their structures geometrically. These descriptions have gradually evolved from tiling vertex models to random cluster models, as the structures of real and simulated quasicrystals have been clarified experimentally. In this paper I outline these developments and explain why this productive dialogue between mathematicians and materials scientists will continue.

Finite size scaling of the 5D Ising model with free boundary conditions

There has been a long running debate on the finite size scaling for the Ising model with free boundary conditions above the upper critical dimension, where the standard picture gives an L2 scaling for the susceptibility and an alternative theory has promoted an L5/2 scaling, as would be the case for cyclic boundary. In this paper we present results from simulation of the far largest systems used so far, up to side L=160 and find that this data clearly supports the standard scaling. Further we present a discussion of why rigorous results for the random-cluster model provide both supports for the standard scaling picture and a clear explanation of why the scalings for free and cyclic boundary should be different.

The Random Cluster Model for Robust Geometric Fitting.

Random hypothesis generation is central to robust geometric model fitting in computer vision. The predominant technique is to randomly sample minimal subsets of the data, and hypothesize the geometric models from the selected subsets. While taking minimal subsets increases the chance of successively "hitting" inliers in a sample, hypotheses fitted on minimal subsets may be severely biased due to the influence of measurement noise, even if the minimal subsets contain purely inliers. In this paper we propose Random Cluster Models, a technique used to simulate coupled spin systems, to conduct hypothesis generation using subsets larger than minimal. We show how large clusters of data from genuine instances of the model can be efficiently harvested to produce accurate hypotheses that are less affected by the vagaries of fitting on minimal subsets. A second aspect of the problem is the optimization of the set of structures that best fit the data. We show how our novel hypothesis sampler can be integrated seamlessly with graph cuts under a simple annealing framework to optimize the fitting efficiently. Unlike previous methods that conduct hypothesis sampling and fitting optimization in two disjoint stages, our algorithm performs the two subtasks alternatingly and in a mutually reinforcing manner. Experimental results show clear improvements in overall efficiency.

Dynamic connectivity algorithms for Monte Carlo simulations of the random-cluster model

We review Sweeny's algorithm for Monte Carlo simulations of the random cluster model. Straightforward implementations suffer from the problem of computational critical slowing down, where the computational effort per edge operation scales with a power of the system size. By using a tailored dynamic connectivity algorithm we are able to perform all operations with a poly-logarithmic computational effort. This approach is shown to be efficient in keeping online connectivity information and is of use for a number of applications also beyond cluster-update simulations, for instance in monitoring droplet shape transitions. As the handling of the relevant data structures is non-trivial, we provide a Python module with a full implementation for future reference.

A Combination of X‐Ray Tomography and Carbon Binder Modeling: Reconstructing the Three Phases of LiCoO2 Li‐Ion Battery Cathodes

X‐ray tomography allows the active‐material domain (LiCoO2) of Li‐ion battery cathodes to be imaged, but it is unable to resolve the carbon‐binder domain (CBD). Here, a new method for creating a complete 3D representation (virtual design) of all three phases of a cathode is provided; this includes the active‐material domain, the CBD, and the electrolyte‐filled pore space. It combines X‐ray tomographic data of active material with a statistically modeled CBD. Two different statistical CBD morphology models are compared as examples: i) a random cluster model representing a standard mixture of carbon black and polyvenylidene fluoride (PVDF) and ii) a fiber model. The transport parameters are compared in a charged and a discharged cathode. The results demonstrate that the CBD content and morphology changes the ionic and electronic transport parameters dramatically and thus cannot be neglected. Calculations yield that the fiber model shows up to three times higher electrical conductivity at the same CBD content (discharged case) and better ionic diffusion conditions for all CBD contents. In the charged case, the morphology impact on electrical conduction is small. This effective method to generate transport parameters for different CBDs can be transferred to other CBD morphologies and electrodes.

Rapid mixing of Swendsen–Wang dynamics in two dimensions

We prove comparison results for the Swendsen–Wang (SW) dynamics, the heat-bath (HB) dynamics for the Potts model and the single-bond (SB) dynamics for the random-cluster model on arbitrary graphs. In particular, we prove that rapid (i.e. polynomia

Swendsen--Wang Is Faster than Single-Bond Dynamics

We prove that the spectral gap of the Swendsen-Wang dynamics for the random-cluster model is larger than the spectral gap of a single-bond dynamics, that updates only a single edge per step. For this we give a representation of the algorithms on the joint (Potts/random-cluster) model. Furthermore we obtain upper and lower bounds on the mixing time of the single-bond dynamics on the discrete $d$-dimensional torus of side length $L$ at the Potts transition temperature for $q$ large enough that are exponential in $L^{d-1}$, complementing a result of Borgs, Chayes and Tetali.

Efficient simulation of the random-cluster model

The simulation of spin models close to critical points of continuous phase transitions is heavily impeded by the occurrence of critical slowing down. A number of cluster algorithms, usually based on the Fortuin-Kasteleyn representation of the Potts model, and suitable generalizations for continuous-spin models have been used to increase simulation efficiency. The first algorithm making use of this representation, suggested by Sweeny in 1983, has not found widespread adoption due to problems in its implementation. However, it has been recently shown that it is indeed more efficient in reducing critical slowing down than the more well-known algorithm due to Swendsen and Wang. Here, we present an efficient implementation of Sweeny's approach for the random-cluster model using recent algorithmic advances in dynamic connectivity algorithms.

Random Graphs from a Weighted Minor-Closed Class

There has been much recent interest in random graphs sampled uniformly from the n-vertex graphs in a suitable minor-closed class, such as the class of all planar graphs. Here we use combinatorial and probabilistic methods to investigate a more general model. We consider random graphs from a 'well-behaved' class of graphs: examples of such classes include all minor-closed classes of graphs with 2-connected excluded minors (such as forests, series-parallel graphs and planar graphs), the class of graphs embeddable on any given surface, and the class of graphs with at most $k$ vertex-disjoint cycles. Also, we give weights to edges and components to specify probabilities, so that our random graphs correspond to the random cluster model, appropriately conditioned.We find that earlier results extend naturally in both directions, to general well-behaved classes of graphs, and to the weighted framework, for example results concerning the probability of a random graph being connected; and we also give results on the 2-core which are new even for the uniform (unweighted) case.

Modular invariant partition function of critical dense polymers

A lattice model of critical dense polymers is solved exactly for arbitrary system size on the torus. More generally, an infinite family of lattice loop models is studied on the torus and related to the corresponding Fortuin–Kasteleyn random cluster models. Starting with a cylinder, the commuting periodic single-row transfer matrices are built from the periodic Temperley–Lieb algebra extended by the shift operators Ω±1. In this enlarged algebra, the non-contractible loop fugacity is α and the contractible loop fugacity is β. The torus is formed by gluing the top and bottom of the cylinder. This gives rise to a variety of non-contractible loops winding around the torus. Because of their nonlocal nature, the standard matrix trace does not produce the proper geometric torus. Instead, we introduce a modified matrix trace for this purpose. This is achieved by using a representation of the enlarged periodic Temperley–Lieb algebra with a parameter v that keeps track of the winding of defects on the cylinder. The transfer matrix representatives and their eigenvalues thus depend on v. The modified trace is constructed as a linear functional on planar connectivity diagrams in terms of matrix traces Trd (with a fixed number of defects d) and Chebyshev polynomials of the first kind. For critical dense polymers, where β=0, the transfer matrix eigenvalues are obtained by solving a functional equation in the form of an inversion identity. The solution depends on d and is subject to selection rules which we prove. Simplifications occur if all non-contractible loop fugacities are set to α=2 in which case the traces are evaluated at v=1. In the continuum scaling limit, the corresponding conformal torus partition function obtained from finite-size corrections agrees with the known modular invariant partition function of symplectic fermions.

Smirnov’s fermionic observable away from criticality

In a recent and celebrated article, Smirnov [Ann. of Math. (2) 172 (2010) 1435-1467] defines an observable for the self-dual random-cluster model with cluster weight q = 2 on the square lattice $\mathbb{Z}^2$, and uses it to obtain conformal invariance in the scaling limit. We study this observable away from the self-dual point. From this, we obtain a new derivation of the fact that the self-dual and critical points coincide, which implies that the critical inverse temperature of the Ising model equals $1/2\log(1+\sqrt{2})$. Moreover, we relate the correlation length of the model to the large deviation behavior of a certain massive random walk (thus confirming an observation by Messikh [The surface tension near criticality of the 2d-Ising model (2006) Preprint]), which allows us to compute it explicitly.

Phase separation in random cluster models II: The droplet at equilibrium, and local deviation lower bounds

We study the droplet that results from conditioning the planar subcritical Fortuin–Kasteleyn random cluster model on the presence of an open circuit Γ0 encircling the origin and enclosing an area of at least (or exactly) n2. We consider local deviation of the droplet boundary, measured in a radial sense by the maximum local roughness, MLR(Γ0), this being the maximum distance from a point in the circuit Γ0 to the boundary ∂ conv(Γ0) of the circuit’s convex hull; and in a longitudinal sense by what we term maximum facet length, namely, the length of the longest line segment of which the polygon ∂ conv(Γ0) is formed. We prove that there exists a constant c > 0 such that the conditional probability that the normalised quantity n−1/3(log n)−2/3 MLR(Γ0) exceeds c tends to 1 in the high n-limit; and that the same statement holds for n−2/3(log n)−1/3 MFL(Γ0). To obtain these bounds, we exhibit the random cluster measure conditional on the presence of an open circuit trapping high area as the invariant measure of a Markov chain that resamples sections of the circuit boundary. We analyse the chain at equilibrium to prove the local roughness lower bounds. Alongside complementary upper bounds provided in [14], the fluctuations MLR(Γ0) and MFL(Γ0) are determined up to a constant factor.

Phase Separation in Random Cluster Models I: Uniform Upper Bounds on Local Deviation

This is the first in a series of three papers that addresses the behaviour of the droplet that results, in the percolating phase, from conditioning the planar Fortuin-Kasteleyn random cluster model on the presence of an open dual circuit Γ0 encircling the origin and enclosing an area of at least (or exactly) n2. (By the Fortuin-Kasteleyn representation, the model is a close relative of the droplet formed by conditioning the Potts model on an excess of spins of a given type.) We consider local deviation of the droplet boundary, measured in a radial sense by the maximum local roughness, MLR(Γ0), this being the maximum distance from a point in the circuit Γ0 to the boundary ∂ of the circuit’s convex hull; and in a longitudinal sense by what we term maximum facet length, MLF(Γ0), namely, the length of the longest line segment of which the polygon ∂ is formed. The principal conclusion of the series of papers is the following uniform control on local deviation: that there are constants 0 < c < C < ∞ such that the conditional probability that the normalized quantity n−1/3(log n )−2/3MLR lies in the interval [c, C] tends to 1 in the high n-limit; and that the same statement holds for n−2/3 (log n )−1/3 MLF. In this way, we confirm the anticipated n1/3 scaling of maximum local roughness, and provide a sharp logarithmic power-law correction. This local deviation behaviour occurs by means of locally Gaussian effects constrained globally by curvature, and we believe that it arises in many radially defined stochastic interface models, including growth models belonging to the Kardar-Parisi-Zhang universality class.

Dynamic Critical Behavior of the Chayes–Machta Algorithm for the Random-Cluster Model, I. Two Dimensions

We study, via Monte Carlo simulation, the dynamic critical behavior of the Chayes-Machta dynamics for the Fortuin-Kasteleyn random-cluster model, which generalizes the Swendsen-Wang dynamics for the q-state Potts ferromagnet to non-integer q \ge 1. We consider spatial dimension d=2 and 1.25 \le q \le 4 in steps of 0.25, on lattices up to 1024^2, and obtain estimates for the dynamic critical exponent z_{CM}. We present evidence that when 1 \le q \lesssim 1.95 the Ossola-Sokal conjecture z_{CM} \ge \beta/\nu is violated, though we also present plausible fits compatible with this conjecture. We show that the Li-Sokal bound z_{CM} \ge \alpha/\nu is close to being sharp over the entire range 1 \le q \le 4, but is probably non-sharp by a power. As a byproduct of our work, we also obtain evidence concerning the corrections to scaling in static observables.

Potts q-color field theory and scaling random cluster model

We study structural properties of the q-color Potts field theory which, for real values of q, describes the scaling limit of the random cluster model. We show that the number of independent n-point Potts spin correlators coincides with that of independent n-point cluster connectivities and is given by generalized Bell numbers. Only a subset of these spin correlators enters the determination of the Potts magnetic properties for q integer. The structure of the operator product expansion of the spin fields for generic q is also identified. For the two-dimensional case, we analyze the duality relation between spin and kink field correlators, both for the bulk and boundary cases, obtaining in particular a sum rule for the kink–kink elastic scattering amplitudes.

Probability on graphs: random processes on graphs and lattices

Preface 1. Random walks on graphs 2. Uniform spanning tree 3. Percolation and self-avoiding walk 4. Association and influence 5. Further percolation 6. Contact process 7. Gibbs states 8. Random-cluster model 9. Quantum Ising model 10. Interacting particle systems 11. Random graphs 12. Lorentz gas References Index.

The self-dual point of the two-dimensional random-cluster model is critical for q ≥ 1

We prove a long-standing conjecture on random-cluster models, namely that the critical point for such models with parameter q ≥ 1 on the square lattice is equal to the self-dual point \({p_{sd}(q) = \sqrt{q} / (1+\sqrt{q})}\). This gives a proof that the critical temperature of the q-state Potts model is equal to \({\log (1+\sqrt q)}\) for all q ≥ 2. We further prove that the transition is sharp, meaning that there is exponential decay of correlations in the sub-critical phase. The techniques of this paper are rigorous and valid for all q ≥ 1, in contrast to earlier methods valid only for certain given q. The proof extends to the triangular and the hexagonal lattices as well.