Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

Journal of Physics Conference Series

Flat-histogram methods in quantum Monte Carlo simulations: Application to the t-J model

Nikolaos G. Diamantis, Efstratios Manousakis

Published Sep 1, 2016• 0 citations
Publisher

Abstract

We discuss that flat-histogram techniques can be appropriately applied in the sampling of quantum Monte Carlo simulation in order to improve the statistical quality of the results at long imaginary time or low excitation energy. Typical imaginary-time correlation functions calculated in quantum Monte Carlo are subject to exponentially growing errors as the range of imaginary time grows and this smears the information on the low energy excitations. We show that we can extract the low energy physics by modifying the Monte Carlo sampling technique to one in which configurations which contribute to making the histogram of certain quantities flat are promoted. We apply the diagrammatic Monte Carlo (diag-MC) method to the motion of a single hole in the t-J model and we show that the implementation of flat-histogram techniques allows us to calculate the Green's function in a wide range of imaginary-time. In addition, we show that applying the flat-histogram technique alleviates the "sign"-problem associated with the simulation of the single-hole Green's function at long imaginary time.

DOI
10.1088/1742-6596/750/1/012002

Related papers

Flat histogram quantum Monte Carlo for analytic continuation to real time

Diamantis, Nikolaos G., Efstratios Manousakis

The Quantum Monte Carlo (QMC) method can yield the imaginary-time dependence of a correlation function $C(τ)$ of an operator $\hat O$. The analytic continuation to real-time proceeds by means of a "numerical inversion" of these data to find the response function or spectral density $A(ω)$ corresponding to $\hat O$. Such a technique is very sensitive to the statistical errors in $C(τ)$ especially for large values of $τ$, when we are interested in the low-energy excitations. In this paper, we find that if we use the flat histogram technique in the QMC method, in such a way to make the {\it histogram of} $C(τ)$ flat, the results of the analytic continuation for low-energy excitations improve using the same amount of computational time. To demonstrate the idea we select an exactly soluble version of the single-hole motion in the $t-J$ model and the diagrammatic Monte Carlo technique.

arXiv:1411.5984•Nov 21, 2014•0 citations

Flat histogram diagrammatic Monte Carlo method: Calculation of the Green's function in imaginary time

Nikolaos G. Diamantis, Efstratios Manousakis

The diagrammatic Monte Carlo (DiagMC) method is a numerical technique which samples the entire diagrammatic series of the Green's function in quantum many-body systems. In this work, we incorporate the flat histogram principle in the diagrammatic Monte Carlo method, and we term the improved version the "flat histogram diagrammatic Monte Carlo" method. We demonstrate the superiority of this method over the standard DiagMC in extracting the long-imaginary-time behavior of the Green's function, without incorporating any a priori knowledge about this function, by applying the technique to the polaron problem.

arXiv:1306.6320•Physical Review E•Oct 9, 2013•5 citations

Matsubara Green's Function and its Analytic Continuation using Flat Histogram Quantum Monte Carlo Simulation

Nikolaos G. Diamantis, Efstratios Manousakis

We demonstrate that by incorporating flat histogram techniques in the quantum Monte Carlo simulation, we can calculate the Matsubara Green's function for significantly larger values of the imaginary time. This allows us to obtain more accurate results for the spectral function, particularly for low energy excitations, using known techniques for analytic continuation to real time.

Physics Procedia•Jan 1, 2015•2 citations

Application of Flat Histogram Methods to Extract the Long Imaginary-time Behavior in Diagrammatic Monte Carlo

Nikolaos G. Diamantis, Efstratios Manousakis

We demonstrate that using flat histogram methods we can extract the imaginary-time behavior of the Green's function G from diagrammatic Monte Carlo simulations very accurately even when G changes by many orders of magnitude.

Physics Procedia•Jan 1, 2014•0 citations

Reducing autocorrelation time in determinant quantum Monte Carlo using the Wang-Landau algorithm: Application to the Holstein model

Meng Yao, Da Wang, Qiang-Hua Wang

When performing a Monte Carlo calculation, the running time should, in principle, be much longer than the autocorrelation time in order to get reliable results. Among different lattice fermion models, the Holstein model is notorious for its particularly long autocorrelation time. In this paper, we employ the Wang-Landau algorithm in the determinant quantum Monte Carlo to achieve the flat-histogram sampling in the "configuration weight space," which can greatly reduce the autocorrelation time by sacrificing some sampling efficiency. The proposal is checked in the Holstein model on both square and honeycomb lattices. Based on such a Wang-Landau assisted determinant quantum Monte Carlo method, some models with long autocorrelation times can now be simulated possibly.

arXiv:2107.14454•Physical review. E•Aug 16, 2021•4 citations

Single-hole dynamics in the t−J model on a square lattice

Michael Brunner, Fakher F. Assaad, Alejandro Muramatsu

We present quantum Monte Carlo (QMC) simulations for a single hole in a $t\ensuremath{-}J$ model from $J=0.4t$ to $J=4t$ on square lattices with up to $24\ifmmode\times\else\texttimes\fi{}24$ sites. The lower edge of the spectrum is directly extracted from the imaginary time Green's function. In agreement with earlier calculations, we find flat bands around $(0,\ifmmode\pm\else\textpm\fi{}\ensuremath{\pi}),$ $(\ifmmode\pm\else\textpm\fi{}\ensuremath{\pi},0)$ and the minimum of the dispersion at $(\ifmmode\pm\else\textpm\fi{}\ensuremath{\pi}/2,\ifmmode\pm\else\textpm\fi{}\ensuremath{\pi}/2).$ For small J, both self-consistent Born approximation and series expansions give a bandwidth for the lower edge of the spectrum in agreement with the simulations, whereas for $J/t>1,$ only series expansions agree quantitatively with our QMC results. This band corresponds to a coherent quasiparticle. This is shown by a finite-size scaling of the quasiparticle weight $Z(\stackrel{\ensuremath{\rightarrow}}{k})$ that leads to a finite result in the thermodynamic limit for the considered values of $J/t.$ The spectral function $A(\stackrel{\ensuremath{\rightarrow}}{k},\ensuremath{\omega})$ is obtained from the imaginary time Green's function via the maximum entropy method. Resonances above the lowest edge of the spectrum are identified, whose J dependence is quantitatively described by string excitations up to $J/t=2.$

arXiv:cond-mat/0002321•Physical review. B, Condensed matter•Dec 15, 2000•110 citations
We discuss that flat-histogram techniques can be appropriately applied in the sampling of quantum Monte Carlo simulation in order to improve the statistical quality of the results at long imaginary time or low excitation energy. Typical imaginary-time correlation functions calculated in quantum Monte Carlo are subject to exponentially growing errors as the range of imaginary time grows and this smears the information on the low energy excitations. We show that we can extract the low energy physics by modifying the Monte Carlo sampling technique to one in which configurations which contribute to making the histogram of certain quantities flat are promoted. We apply the diagrammatic Monte Carlo (diag-MC) method to the motion of a single hole in the t-J model and we show that the implementation of flat-histogram techniques allows us to calculate the Green's function in a wide range of imaginary-time. In addition, we show that applying the flat-histogram technique alleviates the "sign"-problem associated with the simulation of the single-hole Green's function at long imaginary time.
The Quantum Monte Carlo (QMC) method can yield the imaginary-time dependence of a correlation function C(τ)C(τ)C(τ) of an operator O^\hat OO^. The analytic continuation to real-time proceeds by means of a "numerical inversion" of these data to find the response function or spectral density A(ω)A(ω)A(ω) corresponding to O^\hat OO^. Such a technique is very sensitive to the statistical errors in C(τ)C(τ)C(τ) especially for large values of τττ, when we are interested in the low-energy excitations. In this paper, we find that if we use the flat histogram technique in the QMC method, in such a way to make the {\it histogram of} C(τ)C(τ)C(τ) flat, the results of the analytic continuation for low-energy excitations improve using the same amount of computational time. To demonstrate the idea we select an exactly soluble version of the single-hole motion in the t−Jt-Jt−J model and the diagrammatic Monte Carlo technique.
The diagrammatic Monte Carlo (DiagMC) method is a numerical technique which samples the entire diagrammatic series of the Green's function in quantum many-body systems. In this work, we incorporate the flat histogram principle in the diagrammatic Monte Carlo method, and we term the improved version the "flat histogram diagrammatic Monte Carlo" method. We demonstrate the superiority of this method over the standard DiagMC in extracting the long-imaginary-time behavior of the Green's function, without incorporating any a priori knowledge about this function, by applying the technique to the polaron problem.
When performing a Monte Carlo calculation, the running time should, in principle, be much longer than the autocorrelation time in order to get reliable results. Among different lattice fermion models, the Holstein model is notorious for its particularly long autocorrelation time. In this paper, we employ the Wang-Landau algorithm in the determinant quantum Monte Carlo to achieve the flat-histogram sampling in the "configuration weight space," which can greatly reduce the autocorrelation time by sacrificing some sampling efficiency. The proposal is checked in the Holstein model on both square and honeycomb lattices. Based on such a Wang-Landau assisted determinant quantum Monte Carlo method, some models with long autocorrelation times can now be simulated possibly.
We present quantum Monte Carlo (QMC) simulations for a single hole in a $t\ensuremath{-}J$ model from J=0.4tJ=0.4tJ=0.4t to J=4tJ=4tJ=4t on square lattices with up to $24\ifmmode\times\else\texttimes\fi{}24$ sites. The lower edge of the spectrum is directly extracted from the imaginary time Green's function. In agreement with earlier calculations, we find flat bands around $(0,\ifmmode\pm\else\textpm\fi{}\ensuremath{\pi}),$ $(\ifmmode\pm\else\textpm\fi{}\ensuremath{\pi},0)$ and the minimum of the dispersion at $(\ifmmode\pm\else\textpm\fi{}\ensuremath{\pi}/2,\ifmmode\pm\else\textpm\fi{}\ensuremath{\pi}/2).$ For small J, both self-consistent Born approximation and series expansions give a bandwidth for the lower edge of the spectrum in agreement with the simulations, whereas for J/t>1,J/t>1,J/t>1, only series expansions agree quantitatively with our QMC results. This band corresponds to a coherent quasiparticle. This is shown by a finite-size scaling of the quasiparticle weight $Z(\stackrel{\ensuremath{\rightarrow}}{k})$ that leads to a finite result in the thermodynamic limit for the considered values of J/t.J/t.J/t. The spectral function $A(\stackrel{\ensuremath{\rightarrow}}{k},\ensuremath{\omega})$ is obtained from the imaginary time Green's function via the maximum entropy method. Resonances above the lowest edge of the spectrum are identified, whose J dependence is quantitatively described by string excitations up to J/t=2.J/t=2.J/t=2.