DSpace
 

RRI Digital Repository >
07. Theoretical Physics >
Research Papers (TP) >

Please use this identifier to cite or link to this item: http://hdl.handle.net/2289/4521

Title: Performance limitations of flat-histogram methods
Authors: Dayal, P.
Trebst, S.
Wessel, S.
Würtz, D.
Troyer, M.
Sabhapandit, Sanjib
Coppersmith, S.N.
Keywords: classical spin models
general statistical methods
order-disorder transformations
statistical mechanics of model systems
Issue Date: Mar-2004
Publisher: American Physical Society
Citation: Physical Review Letters, 2004, Vol.92, p097201
Abstract: We determine the optimal scaling of local-update flat-histogram methods with system size by using a perfect flat-histogram scheme based upon the exact density of states of 2D Ising models. The typical tunneling time needed to sample the entire bandwidth does not scale with the number of spins N as the minimal N2 of an unbiased random walk in energy space. While the scaling is power law for the ferromagnetic and fully frustrated Ising model, for the ±J nearest-neighbor spin glass the distribution of tunneling times is governed by a fat-tailed Fréchet extremal value distribution that obeys exponential scaling. Furthermore, the shape parameters of these distributions indicate that statistical sample means become ill defined already for moderate system sizes within these complex energy landscapes.
Description: Restricted Access. An open-access version is available at arXiv.org (one of the alternative locations)
URI: http://hdl.handle.net/2289/4521
ISSN: 0031-9007
1079-7114 (Online)
Alternative Location: http://arxiv.org/abs/cond-mat/0306108
http://dx.doi.org/10.1103/PhysRevLett.92.097201
http://adsabs.harvard.edu/abs/2004PhRvL..92i7201D
Copyright: 2004 The American Physical Society
Appears in Collections:Research Papers (TP)

Files in This Item:

File Description SizeFormat
2004_PhysRevLett_V92_p097201.pdfRestricted Access246.82 kBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

 

    RRI Library DSpace