An Analysis of Swendsen–Wang and Related Sampling Methods

George S. Fishman

Journal of the Royal Statistical Society Series B, 1999, vol. 61, issue 3, 623-641

Abstract: Convergence rates, statistical efficiency and sampling costs are studied for the original and extended Swendsen–Wang methods of generating a sample path {Sj, j≥1} with equilibrium distribution π, with r distinct elements, on a finite state space X of size N1. Given Sj‐1, each method uses auxiliary random variables to identify the subset of X from which Sj is to be randomly sampled. Let πmin and πmax denote respectively the smallest and largest elements in π and let Nr denote the number of elements in π with value πmax. For a single auxiliary variable, uniform sampling from the subset and (N1−Nr)πmin+Nrπmax≈1, our results show rapid convergence and high statistical efficiency for large πmin/πmaxorNr/N1 and slow convergence and poor statistical efficiency for small πmin/πmaxandNr/N1. Other examples provide additional insight. For extended Swendsen–Wang methods with non‐uniform subset sampling, the analysis identifies the properties of a decomposition of π(x) that favour fast convergence and high statistical efficiency. In the absence of exploitable special structure, subset sampling can be costly regardless of which of these methods is employed.

Date: 1999
References: Add references at CitEc
Citations: View citations in EconPapers (1)

Downloads: (external link)
https://doi.org/10.1111/1467-9868.00197

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:bla:jorssb:v:61:y:1999:i:3:p:623-641

Ordering information: This journal article can be ordered from
http://ordering.onli ... 1111/(ISSN)1467-9868

Access Statistics for this article

Journal of the Royal Statistical Society Series B is currently edited by P. Fryzlewicz and I. Van Keilegom

More articles in Journal of the Royal Statistical Society Series B from Royal Statistical Society Contact information at EDIRC.
Bibliographic data for series maintained by Wiley Content Delivery ().