Decorrelation steps in Wolff algorithm

I
Thread starter IWantToLearn
Start date Jul 24, 2024
Tags

Simulation

In summary, the decorrelation steps in the Wolff algorithm are designed to reduce correlations between spins in a lattice during Monte Carlo simulations of the Ising model. This is achieved by flipping clusters of spins based on a probability that depends on the interaction strength and temperature. By sequentially updating clusters, the algorithm efficiently samples configurations, improving convergence and reducing autocorrelation times. The decorrelation process enhances the effectiveness of the simulation, allowing for more accurate thermodynamic measurements.

Jul 24, 2024

IWantToLearn

TL;DR Summary: Decorrelation steps in Wolff algorithm

Hello all,

I am doing a monte Carlo simulation using Wolff algorithm for the 2D/3D Ising and classical 3D Heisenberg model, I have been told that the rule of thumb is to make your decorrelation steps = 10*L where L is the size of lattice, and this is would be sufficient enough to make independent measurements for observables even near the critical temperature

So what is recommendation really ?