Biased Metropolis-Heat-Bath Algorithm for Fundamental-Adjoint SU(2) Lattice Gauge Theory
High Energy Physics - Lattice
2009-11-11 v2
Abstract
For SU(2) lattice gauge theory with the fundamental-adjoint action an efficient heat-bath algorithm is not known so that one had to rely on Metropolis simulations supplemented by overrelaxation. Implementing a novel biased Metropolis-heat-bath algorithm for this model, we find improvement factors in the range 1.45 to 2.06 over conventionally optimized Metropolis simulations. If one optimizes further with respect to additional overrelaxation sweeps, the improvement factors are found in the range 1.3 to 1.8.
Keywords
Cite
@article{arxiv.hep-lat/0510108,
title = {Biased Metropolis-Heat-Bath Algorithm for Fundamental-Adjoint SU(2) Lattice Gauge Theory},
author = {Alexei Bazavov and Bernd A. Berg and Urs M. Heller},
journal= {arXiv preprint arXiv:hep-lat/0510108},
year = {2009}
}
Comments
4 pages, 2 figures; minor changes and one reference added; accepted for publication in PRD