Understanding and Improving the Wang-Landau Algorithm

TL;DR

This paper provides a rigorous mathematical analysis of the Wang-Landau algorithm, proving its convergence, identifying error sources, and suggesting optimization strategies to enhance its efficiency.

Contribution

It offers the first detailed convergence proof and error analysis of the Wang-Landau algorithm, along with practical optimization insights.

Findings

Histogram increases uniformly with small fluctuations after initial stage

Statistical error scales as sqrt(ln f) with modification factor f

Strategies for faster convergence are proposed

Abstract

We present a mathematical analysis of the Wang-Landau algorithm, prove its convergence, identify sources of errors and strategies for optimization. In particular, we found the histogram increases uniformly with small fluctuation after a stage of initial accumulation, and the statistical error is found to scale as $\sqrt{\ln f}$ with the modification factor $f$. This has implications for strategies for obtaining fast convergence.