Monte Carlo quasi-heat-bath by approximate inversion

被引：17
作者：
de Forcrand, P [1 ]
机构：
[1] Swiss Fed Inst Technol, Swiss Ctr Sci Comp, CH-8092 Zurich, Switzerland
来源：
PHYSICAL REVIEW E | 1999年 / 59卷 / 03期
关键词：
D O I：
10.1103/PhysRevE.59.3698
中图分类号：
O35 [流体力学]; O53 [等离子体物理学];
学科分类号：
070204 ; 080103 ; 080704 ;
摘要：
When sampling the distribution P(<(phi)over right arrow>)proportional to exp(-\A<(phi)over right arrow>\(2)), a global heat bath normally proceeds by solving the linear system A<(phi)over right arrow> = <(eta)over right arrow>, where <(eta)over right arrow> is a normal Gaussian vector, exactly. This paper shows how to preserve the distribution P(<(phi)over right arrow>) while solving the linear system with arbitrarily low accuracy. Generalizations are presented. [S1063-651X(99)13303-8].
引用
收藏
页码：3698 / 3701
页数：4
相关论文
共 5 条
1