Showing 1–50 of 69 results for author: Kennedy, A D

Geometry and Dynamics for Markov Chain Monte Carlo

Authors: Alessandro Barp, Francois-Xavier Briol, Anthony D. Kennedy, Mark Girolami

Abstract: Markov Chain Monte Carlo methods have revolutionised mathematical computation and enabled statistical inference within many previously intractable models. In this context, Hamiltonian dynamics have been proposed as an efficient way of building chains which can explore probability densities efficiently. The method emerges from physics and geometry and these links have been extensively studied by a… ▽ More Markov Chain Monte Carlo methods have revolutionised mathematical computation and enabled statistical inference within many previously intractable models. In this context, Hamiltonian dynamics have been proposed as an efficient way of building chains which can explore probability densities efficiently. The method emerges from physics and geometry and these links have been extensively studied by a series of authors through the last thirty years. However, there is currently a gap between the intuitions and knowledge of users of the methodology and our deep understanding of these theoretical foundations. The aim of this review is to provide a comprehensive introduction to the geometric tools used in Hamiltonian Monte Carlo at a level accessible to statisticians, machine learners and other users of the methodology with only a basic understanding of Monte Carlo methods. This will be complemented with some discussion of the most recent advances in the field which we believe will become increasingly relevant to applied scientists. △ Less

Submitted 8 May, 2017; originally announced May 2017.

Comments: Submitted to "Annual Review of Statistics and Its Applications"

Investigation of New Methods for Numerical Stochastic Perturbation Theory in $\varphi^4$ Theory

Authors: Mattia Dalla Brida, Marco Garofalo, A. D. Kennedy

Abstract: Numerical stochastic perturbation theory is a powerful tool for estimating high-order perturbative expansions in lattice field theory. The standard algorithms based on the Langevin equation, however, suffer from several limitations which in practice restrict the potential of this technique. In this work we investigate some alternative methods which could in principle improve on the standard approa… ▽ More Numerical stochastic perturbation theory is a powerful tool for estimating high-order perturbative expansions in lattice field theory. The standard algorithms based on the Langevin equation, however, suffer from several limitations which in practice restrict the potential of this technique. In this work we investigate some alternative methods which could in principle improve on the standard approach. In particular, we present a study of the recently proposed Instantaneous Stochastic Perturbation Theory, as well as a formulation of numerical stochastic perturbation theory based on Generalized Hybrid Molecular Dynamics algorithms. The viability of these methods is investigated in $\varphi^4$ theory. △ Less

Submitted 8 September, 2017; v1 submitted 13 March, 2017; originally announced March 2017.

Comments: 45 pages, 12 figures. Added new section on cost comparison with Langevin NSPT. Matches published version

Journal ref: Phys. Rev. D 96, 054502 (2017)

Numerical Stochastic Perturbation Theory and Gradient Flow in φ^4 Theory

Authors: Mattia Dalla Brida, Marco Garofalo, Anthony D. Kennedy

Abstract: In this contribution we present an exploratory study of several novel methods for numerical stochastic perturbation theory. For the investigation we consider observables defined through the gradient flow in the simple φ^4 theory. In this contribution we present an exploratory study of several novel methods for numerical stochastic perturbation theory. For the investigation we consider observables defined through the gradient flow in the simple φ^4 theory. △ Less

Submitted 27 December, 2015; originally announced December 2015.

Comments: Talk given at the 33rd International Symposium on Lattice Field Theory, 14 -18 July, 2015, Kobe, Japan; LaTeX source, 7 pages, 7 figures

Renormalizability of the Schrödinger Functional

Authors: A. D. Kennedy, Stefan Sint

Abstract: Symanzik showed that quantum field theory can be formulated on a space with boundaries by including suitable surface interactions in the action to implement boundary conditions. We show that to all orders in perturbation theory all the divergences induced by these surface interactions can be absorbed by a renormalization of their coefficients. Symanzik showed that quantum field theory can be formulated on a space with boundaries by including suitable surface interactions in the action to implement boundary conditions. We show that to all orders in perturbation theory all the divergences induced by these surface interactions can be absorbed by a renormalization of their coefficients. △ Less

Submitted 26 November, 2015; originally announced November 2015.

Comments: Parallel talk presented at Lattice 2015, Kobe, Japan

Numerical determination of partial spectrum of Hermitian matrices using a Lanczos method with selective reorthogonalization

Authors: Chris Johnson, A. D. Kennedy

Abstract: We introduce a new algorithm for finding the eigenvalues and eigenvectors of Hermitian matrices within a specified region, based upon the LANSO algorithm of Parlett and Scott. It uses selective reorthogonalization to avoid the duplication of eigenpairs in finite-precision arithmetic, but uses a new bound to decide when such reorthogonalization is required, and only reorthogonalizes with respect to… ▽ More We introduce a new algorithm for finding the eigenvalues and eigenvectors of Hermitian matrices within a specified region, based upon the LANSO algorithm of Parlett and Scott. It uses selective reorthogonalization to avoid the duplication of eigenpairs in finite-precision arithmetic, but uses a new bound to decide when such reorthogonalization is required, and only reorthogonalizes with respect to eigenpairs within the region of interest. We investigate its performance for the Hermitian Wilson--Dirac operator (γ_5D) in lattice quantum chromodynamics, and compare it with previous methods. △ Less

Submitted 6 November, 2012; originally announced November 2012.

Shadow Hamiltonians, Poisson Brackets, and Gauge Theories

Authors: A. D. Kennedy, P. J. Silva, M. A. Clark

Abstract: Numerical lattice gauge theory computations to generate gauge field configurations including the effects of dynamical fermions are usually carried out using algorithms that require the molecular dynamics evolution of gauge fields using symplectic integrators. Sophisticated integrators are in common use but are hard to optimise, and force-gradient integrators show promise especially for large latti… ▽ More Numerical lattice gauge theory computations to generate gauge field configurations including the effects of dynamical fermions are usually carried out using algorithms that require the molecular dynamics evolution of gauge fields using symplectic integrators. Sophisticated integrators are in common use but are hard to optimise, and force-gradient integrators show promise especially for large lattice volumes. We explain why symplectic integrators lead to very efficient Monte Carlo algorithms because they exactly conserve a shadow Hamiltonian. The shadow Hamiltonian may be expanded in terms of Poisson brackets, and can be used to optimize the integrators. We show how this may be done for gauge theories by extending the formulation of Hamiltonian mechanics on Lie groups to include Poisson brackets and shadows, and by giving a general method for the practical computation of forces, force-gradients, and Poisson brackets for gauge theories. △ Less

Submitted 24 October, 2012; originally announced October 2012.

Monte Carlo Integration with Subtraction

Authors: Rudy Arthur, A. D. Kennedy

Abstract: This paper investigates a class of algorithms for numerical integration of a function in d dimensions over a compact domain by Monte Carlo methods. We construct a histogram approximation to the function using a partition of the integration domain into a set of bins specified by some parameters. We then consider two adaptations; the first is to subtract the histogram approximation, whose integral w… ▽ More This paper investigates a class of algorithms for numerical integration of a function in d dimensions over a compact domain by Monte Carlo methods. We construct a histogram approximation to the function using a partition of the integration domain into a set of bins specified by some parameters. We then consider two adaptations; the first is to subtract the histogram approximation, whose integral we may easily evaluate explicitly, from the function and integrate the difference using Monte Carlo; the second is to modify the bin parameters in order to make the variance of the Monte Carlo estimate of the integral the same for all bins. This allows us to use Student's t-test as a trigger for rebinning, which we claim is more stable than the χ-squared test that is commonly used for this purpose. We provide a program that we have used to study the algorithm for the case where the histogram is represented as a product of one-dimensional histograms. We discuss the assumptions and approximations made, as well as giving a pedagogical discussion of the myriad ways in which the results of any such Monte Carlo integration program can be misleading. △ Less

Submitted 4 September, 2012; originally announced September 2012.

Comments: Code PANIC included as a set of ancillary files

Bounds on the convergence of Ritz values from Krylov subspaces to interior eigenvalues of Hermitean matrices

Authors: Chris Johnson, A. D. Kennedy

Abstract: We consider bounds on the convergence of Ritz values from a sequence of Krylov subspaces to interior eigenvalues of Hermitean matrices. These bounds are useful in regions of low spectral density, for example near voids in the spectrum, as is required in many applications. Our bounds are obtained by considering the usual Kaniel-Paige-Saad formalism applied to the shifted and squared matrix. We consider bounds on the convergence of Ritz values from a sequence of Krylov subspaces to interior eigenvalues of Hermitean matrices. These bounds are useful in regions of low spectral density, for example near voids in the spectrum, as is required in many applications. Our bounds are obtained by considering the usual Kaniel-Paige-Saad formalism applied to the shifted and squared matrix. △ Less

Submitted 13 October, 2011; originally announced October 2011.

Comments: 1 figure

MSC Class: 65F15 (Primary) 15A18; 15B57; 65Z05; 65B99 (Secondary)

Improving dynamical lattice QCD simulations through integrator tuning using Poisson brackets and a force-gradient integrator

Authors: M. A. Clark, Bálint Joó, A. D. Kennedy, P. J. Silva

Abstract: We show how the integrators used for the molecular dynamics step of the Hybrid Monte Carlo algorithm can be further improved. These integrators not only approximately conserve some Hamiltonian $H$ but conserve exactly a nearby shadow Hamiltonian $\tilde{H}$. This property allows for a new tuning method of the molecular dynamics integrator and also allows for a new class of integrators (force-gradi… ▽ More We show how the integrators used for the molecular dynamics step of the Hybrid Monte Carlo algorithm can be further improved. These integrators not only approximately conserve some Hamiltonian $H$ but conserve exactly a nearby shadow Hamiltonian $\tilde{H}$. This property allows for a new tuning method of the molecular dynamics integrator and also allows for a new class of integrators (force-gradient integrators) which is expected to reduce significantly the computational cost of future large-scale gauge field ensemble generation. △ Less

Submitted 20 September, 2011; v1 submitted 8 August, 2011; originally announced August 2011.

Comments: 5 pages, 1 figure; minor changes

Better HMC integrators for dynamical simulations

Authors: M. A. Clark, Balint Joo, A. D. Kennedy, P. J. Silva

Abstract: We show how to improve the molecular dynamics step of Hybrid Monte Carlo, both by tuning the integrator using Poisson brackets measurements and by the use of force gradient integrators. We present results for moderate lattice sizes. We show how to improve the molecular dynamics step of Hybrid Monte Carlo, both by tuning the integrator using Poisson brackets measurements and by the use of force gradient integrators. We present results for moderate lattice sizes. △ Less

Submitted 31 October, 2010; originally announced November 2010.

Comments: 6 pages, 1 figure, poster presented at Lattice 2010 (Algorithms and Machines)

Journal ref: PoS Lattice2010:323,2010

Time symmetric Bloch-Nordsieck infrared cancellation

Authors: G. E. C. Burton, A. D. Kennedy

Abstract: It is usually assumed that the infrared divergences that occur for virtual photons in Feynman diagrams cancel against the corresponding divergences in phase space integrals for the emission of real soft photons. If this were the case then why do the corresponding phase space integrals for the absorption of real soft photons not diverge too, as one might expect on grounds of time-reversal symmetry,… ▽ More It is usually assumed that the infrared divergences that occur for virtual photons in Feynman diagrams cancel against the corresponding divergences in phase space integrals for the emission of real soft photons. If this were the case then why do the corresponding phase space integrals for the absorption of real soft photons not diverge too, as one might expect on grounds of time-reversal symmetry, and if so what cancels these divergences? We show that this apparent time asymmetry is an artifact of the way that the diagrams are calculated, and we introduce a procedure for such calculations that is manifestly time symmetric. Moreover, we show that it is not mathematically meaningful to distinguish arbitrarily soft massless particles of positive and negative energy. △ Less

Submitted 26 July, 2010; v1 submitted 14 June, 2010; originally announced June 2010.

Comments: This paper has been withdrawn by the author due to a crucial error in equation 2.6

Report number: Edinburgh 2010/19

Tuning the strange quark mass in lattice simulations

Authors: W. Bietenholz, V. Bornyakov, N. Cundy, M. Göckeler, R. Horsley, A. D. Kennedy, W. G. Lockhart, Y. Nakamura, H. Perlt, D. Pleiter, P. E. L. Rakow, A. Schäfer, G. Schierholz, A. Schiller, H. Stüben, J. M. Zanotti

Abstract: QCD lattice simulations with 2+1 flavours typically start at rather large up-down and strange quark masses and extrapolate first the strange quark mass to its physical value and then the up-down quark mass. An alternative method of tuning the quark masses is discussed here in which the singlet quark mass is kept fixed, which ensures that the kaon always has mass less than the physical kaon mass.… ▽ More QCD lattice simulations with 2+1 flavours typically start at rather large up-down and strange quark masses and extrapolate first the strange quark mass to its physical value and then the up-down quark mass. An alternative method of tuning the quark masses is discussed here in which the singlet quark mass is kept fixed, which ensures that the kaon always has mass less than the physical kaon mass. It can also take into account the different renormalisations (for singlet and non-singlet quark masses) occurring for non-chirally invariant lattice fermions and so allows a smooth extrapolation to the physical quark masses. This procedure enables a wide range of quark masses to be probed, including the case with a heavy up-down quark mass and light strange quark mass. Results show the correct order for the baryon octet and decuplet spectrum and an extrapolation to the physical pion mass gives mass values to within a few percent of their experimental values. △ Less

Submitted 20 March, 2010; v1 submitted 4 March, 2010; originally announced March 2010.

Comments: 13 pages, 4 figures

Report number: DESY 10-029, Edinburgh 2010/06, LTH 867

Journal ref: Phys.Lett.B690:436-441,2010

Results from 2+1 flavours of SLiNC fermions

Authors: W. Bietenholz, V. Bornyakov, N. Cundy, M. Göckeler, R. Horsley, A. D. Kennedy, Y. Nakamura, H. Perlt, D. Pleiter, P. E. L. Rakow, A. Schäfer, G. Schierholz, A. Schiller, H. Stüben, J. M. Zanotti

Abstract: QCD results are presented for a 2+1 flavour fermion clover action (which we call the SLiNC action). A method of tuning the quark masses to their physical values is discussed. In this method the singlet quark mass is kept fixed, which solves the problem of different renormalisations (for singlet and non-singlet quark masses) occuring for non-chirally invariant lattice fermions. This procedure ena… ▽ More QCD results are presented for a 2+1 flavour fermion clover action (which we call the SLiNC action). A method of tuning the quark masses to their physical values is discussed. In this method the singlet quark mass is kept fixed, which solves the problem of different renormalisations (for singlet and non-singlet quark masses) occuring for non-chirally invariant lattice fermions. This procedure enables a wide range of quark masses to be probed, including the case with a heavy up-down quark mass and light strange quark mass. Preliminary results show the correct splittings for the baryon (octet and) decuplet spectrum. △ Less

Submitted 15 October, 2009; originally announced October 2009.

Comments: 7 pages; talk given at the XXVII International Symposium on Lattice Field Theory, July 26-31 2009, Peking University, Bei**g, China

Report number: DESY 09-172, Edinburgh 2009/15, Liverpool LTH 848

Journal ref: PoS LAT2009:102,2009

Force Gradient Integrators

Authors: A. D. Kennedy, M. A. Clark, P. J. Silva

Abstract: We present initial results of the use of Force Gradient integrators for lattice field theories. These promise to give significant performance improvements, especially for light fermions and large lattices. Our results show that this is indeed the case, indicating a speed-up of more than a factor of two, which is expected to increase as the integration step size becomes smaller for larger lattice… ▽ More We present initial results of the use of Force Gradient integrators for lattice field theories. These promise to give significant performance improvements, especially for light fermions and large lattices. Our results show that this is indeed the case, indicating a speed-up of more than a factor of two, which is expected to increase as the integration step size becomes smaller for larger lattices and smaller fermion masses. △ Less

Submitted 15 October, 2009; originally announced October 2009.

Comments: 6 pages, 2 figures, talk presented at Lattice 2009 (Algorithms and Machines)

Journal ref: PoS LAT2009:021,2009

Non-perturbative improvement of stout-smeared three flavour clover fermions

Authors: N. Cundy, M. Göckeler, R. Horsley, T. Kaltenbrunner, A. D. Kennedy, Y. Nakamura, H. Perlt, D. Pleiter, P. E. L. Rakow, A. Schäfer, G. Schierholz, A. Schiller, H. Stüben, J. M. Zanotti

Abstract: We discuss a 3-flavour lattice QCD action with clover improvement in which the fermion matrix has single level stout smearing for the hop** terms together with unsmeared links for the clover term. With the (tree-level) Symanzik improved gluon action this constitutes the Stout Link Non-perturbative Clover or SLiNC action. To cancel O(a) terms the clover term coefficient has to be tuned. We pres… ▽ More We discuss a 3-flavour lattice QCD action with clover improvement in which the fermion matrix has single level stout smearing for the hop** terms together with unsmeared links for the clover term. With the (tree-level) Symanzik improved gluon action this constitutes the Stout Link Non-perturbative Clover or SLiNC action. To cancel O(a) terms the clover term coefficient has to be tuned. We present here results of a non-perturbative determination of this coefficient using the Schroedinger functional and as a by-product a determination of the critical hop** parameter. Comparisons of the results are made with lowest order perturbation theory. △ Less

Submitted 3 June, 2009; v1 submitted 21 January, 2009; originally announced January 2009.

Comments: 30 pages, 13 figures, minor changes, published version

Report number: DESY 08-141, Edinburgh 2009/01, LTH 819

Journal ref: Phys.Rev.D79:094507,2009

Clover improvement for stout-smeared 2+1 flavour SLiNC fermions: non-perturbative results

Authors: N. Cundy, M. Göckeler, R. Horsley, T. Kaltenbrunner, A. D. Kennedy, Y. Nakamura, H. Perlt, D. Pleiter, P. E. L. Rakow, A. Schäfer, G. Schierholz, A. Schiller, H. Stüben, J. M. Zanotti

Abstract: We discuss an action in which the fermion matrix has single level stout smearing for the hop** terms together with unsmeared links for the clover term. With the (tree level) Symanzik improved gluon action this constitutes the Stout Link Non-perturbative Clover or SLiNC action. To cancel O(a) terms the clover coefficient, csw, has to be tuned. We present here preliminary results of a non-pertur… ▽ More We discuss an action in which the fermion matrix has single level stout smearing for the hop** terms together with unsmeared links for the clover term. With the (tree level) Symanzik improved gluon action this constitutes the Stout Link Non-perturbative Clover or SLiNC action. To cancel O(a) terms the clover coefficient, csw, has to be tuned. We present here preliminary results of a non-perturbative determination of csw using the Schrodinger functional and as a by-product also a determination of the critical hop** parameter. A determination of the renormalisation constant for the local vector current is also given. Comparisons of the results are made with lowest order perturbation theory results. △ Less

Submitted 14 November, 2008; originally announced November 2008.

Comments: 7 pages, 5 figures, talk presented at the XXVI International Symposium on Lattice Field Theory, July 14-19, 2008, Williamsburg, Virginia, USA

Report number: DESY 08-164, Edinburgh 2008/31, Liverpool LTH 812

Journal ref: PoS LATTICE2008:132,2008

Tuning HMC using Poisson brackets

Authors: M. A. Clark, A. D. Kennedy, P. J. Silva

Abstract: We discuss how the integrators used for the Hybrid Monte Carlo (HMC) algorithm not only approximately conserve some Hamiltonian $H$ but exactly conserve a nearby shadow Hamiltonian (\tilde H), and how the difference $ΔH \equiv \tilde H - H $ may be expressed as an expansion in Poisson brackets. By measuring average values of these Poisson brackets over the equilibrium distribution… ▽ More We discuss how the integrators used for the Hybrid Monte Carlo (HMC) algorithm not only approximately conserve some Hamiltonian $H$ but exactly conserve a nearby shadow Hamiltonian (\tilde H), and how the difference $ΔH \equiv \tilde H - H $ may be expressed as an expansion in Poisson brackets. By measuring average values of these Poisson brackets over the equilibrium distribution $\propto e^{-H}$ generated by HMC we can find the optimal integrator parameters from a single simulation. We show that a good way of doing this in practice is to minimize the variance of $ΔH$ rather than its magnitude, as has been previously suggested. Some details of how to compute Poisson brackets for gauge and fermion fields, and for nested and force gradient integrators are also presented. △ Less

Submitted 7 October, 2008; originally announced October 2008.

Comments: 7 pages, 4 figures, poster presented at the XXVI International Symposium on Lattice Field Theory, July 14-19, 2008, Williamsburg, Virginia, USA

Journal ref: PoS LATTICE2008:041,2008

Physical Results from 2+1 Flavor Domain Wall QCD and SU(2) Chiral Perturbation Theory

Authors: C. Allton, D. J. Antonio, Y. Aoki, T. Blum, P. A. Boyle, N. H. Christ, S. D. Cohen, M. A. Clark, C. Dawson, M. A. Donnellan, J. M. Flynn, A. Hart, T. Izubuchi, A. Juttner, C. Jung, A. D. Kennedy, R. D. Kenway, M. Li, S. Li, M. F. Lin, R. D. Mawhinney, C. M. Maynard, S. Ohta, B. J. Pendleton, C. T. Sachrajda , et al. (7 additional authors not shown)

Abstract: We have simulated QCD using 2+1 flavors of domain wall quarks on a $(2.74 {\rm fm})^3$ volume with an inverse lattice scale of $a^{-1} = 1.729(28)$ GeV. The up and down (light) quarks are degenerate in our calculations and we have used four values for the ratio of light quark masses to the strange (heavy) quark mass in our simulations: 0.217, 0.350, 0.617 and 0.884. We have measured pseudoscalar… ▽ More We have simulated QCD using 2+1 flavors of domain wall quarks on a $(2.74 {\rm fm})^3$ volume with an inverse lattice scale of $a^{-1} = 1.729(28)$ GeV. The up and down (light) quarks are degenerate in our calculations and we have used four values for the ratio of light quark masses to the strange (heavy) quark mass in our simulations: 0.217, 0.350, 0.617 and 0.884. We have measured pseudoscalar meson masses and decay constants, the kaon bag parameter $B_K$ and vector meson couplings. We have used SU(2) chiral perturbation theory, which assumes only the up and down quark masses are small, and SU(3) chiral perturbation theory to extrapolate to the physical values for the light quark masses. While next-to-leading order formulae from both approaches fit our data for light quarks, we find the higher order corrections for SU(3) very large, making such fits unreliable. We also find that SU(3) does not fit our data when the quark masses are near the physical strange quark mass. Thus, we rely on SU(2) chiral perturbation theory for accurate results. We use the masses of the $Ω$ baryon, and the $π$ and $K$ mesons to set the lattice scale and determine the quark masses. We then find $f_π= 124.1(3.6)_{\rm stat}(6.9)_{\rm syst} {\rm MeV}$, $f_K = 149.6(3.6)_{\rm stat}(6.3)_{\rm syst} {\rm MeV}$ and $f_K/f_π= 1.205(0.018)_{\rm stat}(0.062)_{\rm syst}$. Using non-perturbative renormalization to relate lattice regularized quark masses to RI-MOM masses, and perturbation theory to relate these to $\bar{\rm MS}$ we find $ m_{ud}^{\bar{\rm MS}}(2 {\rm GeV}) = 3.72(0.16)_{\rm stat}(0.33)_{\rm ren}(0.18)_{\rm syst} {\rm MeV}$ and $m_{s}^{\bar{\rm MS}}(2 {\rm GeV}) = 107.3(4.4)_{\rm stat}(9.7)_{\rm ren}(4.9)_{\rm syst} {\rm MeV}$. △ Less

Submitted 3 April, 2008; originally announced April 2008.

Comments: 133 pages, 25 figures

Report number: BNL-HET-08/5, CU-TP-1182, Edinburgh 2008/06, KEK-TH-1232, RBRC-730, SHEP-0812

Journal ref: Phys.Rev.D78:114509,2008

Automating Renormalization of Quantum Field Theories

Authors: A. D. Kennedy, T. Binoth, T. Rippon

Abstract: We give an overview of state-of-the-art multi-loop Feynman diagram computations, and explain how we use symbolic manipulation to generate renormalized integrals that are then evaluated numerically. We explain how we automate BPHZ renormalization using "henges" and "sectors", and give a brief description of the symbolic tensor and Dirac gamma-matrix manipulation that is required. We shall compare… ▽ More We give an overview of state-of-the-art multi-loop Feynman diagram computations, and explain how we use symbolic manipulation to generate renormalized integrals that are then evaluated numerically. We explain how we automate BPHZ renormalization using "henges" and "sectors", and give a brief description of the symbolic tensor and Dirac gamma-matrix manipulation that is required. We shall compare the use of general computer algebra systems such as Maple with domain-specific languages such as FORM, highlighting in particular memory management issues. △ Less

Submitted 6 December, 2007; originally announced December 2007.

Comments: Invited talk given by A.D. Kennedy at the International Workshop on Symbolic-Numeric Computation'07, University of Western Ontario, Canada

Report number: Edinburgh 2007/46

Journal ref: Proceedings of the 2007 international workshop on Symbolic-numeric computation, London, Ontario, Canada, 18-27 (2007). ISBN:978-1-59593-744-5

Speeding up HMC with better integrators

Authors: M. A. Clark, A. D. Kennedy

Abstract: We discuss how dynamical fermion computations may be made yet cheaper by using symplectic integrators that conserve energy much more accurately without decreasing the integration step size. We first explain why symplectic integrators exactly conserve a ``shadow'' Hamiltonian close to the desired one, and how this Hamiltonian may be computed in terms of Poisson brackets. We then discuss how class… ▽ More We discuss how dynamical fermion computations may be made yet cheaper by using symplectic integrators that conserve energy much more accurately without decreasing the integration step size. We first explain why symplectic integrators exactly conserve a ``shadow'' Hamiltonian close to the desired one, and how this Hamiltonian may be computed in terms of Poisson brackets. We then discuss how classical mechanics may be implemented on Lie groups and derive the form of the Poisson brackets and force terms for some interesting integrators such as those making use of second derivatives of the action (Hessian or force gradient integrators). We hope that these will be seen to greatly improve energy conservation for only a small additional cost and that their use will significantly reduce the cost of dynamical fermion computations. △ Less

Submitted 25 October, 2007; v1 submitted 18 October, 2007; originally announced October 2007.

Comments: 6 pages, talk presented by ADK at Lattice 2007 (Algorithms and Machines), Regensburg, 30th July-4th August, 2007

Journal ref: PoSLAT2007:038,2007

Asymptotics of Fixed Point Distributions for Inexact Monte Carlo Algorithms

Authors: M. A. Clark, A. D. Kennedy

Abstract: We introduce a simple general method for finding the equilibrium distribution for a class of widely used inexact Markov Chain Monte Carlo algorithms. The explicit error due to the non-commutivity of the updating operators when numerically integrating Hamilton's equations can be derived using the Baker-Campbell-Hausdorff formula. This error is manifest in the conservation of a ``shadow'' Hamilton… ▽ More We introduce a simple general method for finding the equilibrium distribution for a class of widely used inexact Markov Chain Monte Carlo algorithms. The explicit error due to the non-commutivity of the updating operators when numerically integrating Hamilton's equations can be derived using the Baker-Campbell-Hausdorff formula. This error is manifest in the conservation of a ``shadow'' Hamiltonian that lies close to the desired Hamiltonian. The fixed point distribution of inexact Hybrid algorithms may then be derived taking into account that the fixed point of the momentum heatbath and that of the molecular dynamics do not coincide exactly. We perform this derivation for various inexact algorithms used for lattice QCD calculations. △ Less

Submitted 18 October, 2007; v1 submitted 14 May, 2007; originally announced May 2007.

Comments: 24 pages, accepted for publication in Physics Review D

Journal ref: Phys.Rev.D76:074508,2007

2+1 flavor domain wall QCD on a (2 fm)^3 lattice: light meson spectroscopy with Ls = 16

Authors: C. Allton, D. J. Antonio, T. Blum, K. C. Bowler, P. A. Boyle, N. H. Christ, S. D. Cohen, M. A. Clark, C. Dawson, A. Hart, K. Hashimoto, T. Izubuchi, A. Juttner, C. Jung, A. D. Kennedy, R. D. Kenway, M. Li, S. Li, M. F. Lin, R. D. Mawhinney, C. M. Maynard, J. Noaki, S. Ohta, B. J. Pendleton, S. Sasaki , et al. (5 additional authors not shown)

Abstract: We present results for light meson masses and pseudoscalar decay constants from the first of a series of lattice calculations with 2+1 dynamical flavors of domain wall fermions and the Iwasaki gauge action. The work reported here was done at a fixed lattice spacing of about 0.12 fm on a 16^3\times32 lattice, which amounts to a spatial volume of (2 fm)^3 in physical units. The number of sites in… ▽ More We present results for light meson masses and pseudoscalar decay constants from the first of a series of lattice calculations with 2+1 dynamical flavors of domain wall fermions and the Iwasaki gauge action. The work reported here was done at a fixed lattice spacing of about 0.12 fm on a 16^3\times32 lattice, which amounts to a spatial volume of (2 fm)^3 in physical units. The number of sites in the fifth dimension is 16, which gives m_{res} = 0.00308(4) in these simulations. Three values of input light sea quark masses, m_l^{sea} \approx 0.85 m_s, 0.59 m_s and 0.33 m_s were used to allow for extrapolations to the physical light quark limit, whilst the heavier sea quark mass was fixed to approximately the physical strange quark mass m_s. The exact rational hybrid Monte Carlo algorithm was used to evaluate the fractional powers of the fermion determinants in the ensemble generation. We have found that f_π= 127(4) MeV, f_K = 157(5) MeV and f_K/f_π= 1.24(2), where the errors are statistical only, which are in good agreement with the experimental values. △ Less

Submitted 23 October, 2007; v1 submitted 17 January, 2007; originally announced January 2007.

Comments: RBC and UKQCD Collaborations. 17 pages, 14 figures. Typeset with ReVTEX4. v2: replaced with the version published in PRD with improved introduction

Report number: BNL-HET-07/1, CU-TP-1175, Edinburgh 2007/1, KEK-TH-1132, RBRC-637, SHEP-0701, SWAT/07/502

Journal ref: Phys.Rev.D76:014504,2007

First results from 2+1-Flavor Domain Wall QCD: Mass Spectrum, Topology Change and Chiral Symmetry with $L_s=8$

Authors: D. J. Antonio, T. Blum, K. C. Bowler, P. A. Boyle, N. H. Christ, S. D. Cohen, M. A. Clark, C. Dawson, A. Hart, K. Hashimoto, T. Izubuchi, B. Joó, C. Jung, A. D. Kennedy, R. D. Kenway, S. Li, H. W. Lin, M. F. Lin, R. D. Mawhinney, C. M. Maynard, J. Noaki, S. Ohta, S. Sasaki, A. Soni, R. J. Tweedie , et al. (1 additional authors not shown)

Abstract: We present results for the static interquark potential, light meson and baryon masses, and light pseudoscalar meson decay constants obtained from simulations of domain wall QCD with one dynamical flavour approximating the $s$ quark, and two degenerate dynamical flavours with input bare masses ranging from $m_s$ to $m_s/4$ approximating the $u$ and $d$ quarks. We compare these quantities obtained… ▽ More We present results for the static interquark potential, light meson and baryon masses, and light pseudoscalar meson decay constants obtained from simulations of domain wall QCD with one dynamical flavour approximating the $s$ quark, and two degenerate dynamical flavours with input bare masses ranging from $m_s$ to $m_s/4$ approximating the $u$ and $d$ quarks. We compare these quantities obtained using the Iwasaki and DBW2 improved gauge actions, and actions with larger rectangle coefficients, on $16^3\times32$ lattices. We seek parameter values at which both the chiral symmetry breaking residual mass due to the finite lattice extent in the fifth dimension and the Monte Carlo time history for topological charge are acceptable for this set of quark masses at lattice spacings above 0.1 fm. We find that the Iwasaki gauge action is best, demonstrating the feasibility of using QCDOC to generate ensembles which are good representations of the QCD path integral on lattices of up to 3 fm in spatial extent with lattice spacings in the range 0.09-0.13 fm. Despite large residual masses and a limited number of sea quark mass values with which to perform chiral extrapolations, our results for light hadronic physics scale and agree with experimental measurements within our statistical uncertainties. △ Less

Submitted 21 December, 2006; v1 submitted 5 December, 2006; originally announced December 2006.

Comments: RBC and UKQCD Collaborations. 82 pages, 34 figures Typos corrected

Report number: BNL-HET-06/17, CU-TP-1159, Edinburgh 2006/39, KEK-TH-1115, RBRC-631

Journal ref: Phys.Rev.D75:114501,2007

Accelerating Staggered Fermion Dynamics with the Rational Hybrid Monte Carlo (RHMC) Algorithm

Authors: M. A. Clark, A. D. Kennedy

Abstract: Improved staggered fermion formulations are a popular choice for lattice QCD calculations. Historically, the algorithm used for such calculations has been the inexact R algorithm, which has systematic errors that only vanish as the square of the integration step-size. We describe how the exact Rational Hybrid Monte Carlo (RHMC) algorithm may be used in this context, and show that for parameters… ▽ More Improved staggered fermion formulations are a popular choice for lattice QCD calculations. Historically, the algorithm used for such calculations has been the inexact R algorithm, which has systematic errors that only vanish as the square of the integration step-size. We describe how the exact Rational Hybrid Monte Carlo (RHMC) algorithm may be used in this context, and show that for parameters corresponding to current state-of-the-art computations it leads to a factor of approximately seven decrease in cost as well as having no step-size errors. △ Less

Submitted 18 October, 2007; v1 submitted 6 October, 2006; originally announced October 2006.

Comments: 4 pages, 2 figures, 1 table

Journal ref: Phys.Rev.D75:011502,2007

Accelerating Dynamical Fermion Computations using the Rational Hybrid Monte Carlo (RHMC) Algorithm with Multiple Pseudofermion Fields

Authors: M. A. Clark, A. D. Kennedy

Abstract: There has been much recent progress in the understanding and reduction of the computational cost of the Hybrid Monte Carlo algorithm for Lattice QCD as the quark mass parameter is reduced. In this letter we present a new solution to this problem, where we represent the fermionic determinant using (n) pseudofermion fields, each with an (\nth) root kernel. We implement this within the framework of… ▽ More There has been much recent progress in the understanding and reduction of the computational cost of the Hybrid Monte Carlo algorithm for Lattice QCD as the quark mass parameter is reduced. In this letter we present a new solution to this problem, where we represent the fermionic determinant using (n) pseudofermion fields, each with an (\nth) root kernel. We implement this within the framework of the Rational Hybrid Monte Carlo (RHMC) algorithm. We compare this algorithm with other recent methods in this area and find it is competitive with them. △ Less

Submitted 22 August, 2006; originally announced August 2006.

Comments: 4 pages, 2 figures. Submitted to PRL

Journal ref: Phys.Rev.Lett.98:051601,2007

Algorithms for Dynamical Fermions

Authors: A. D. Kennedy

Abstract: This is the write-up of three lectures on algorithms for dynamical fermions that were given at the ILFTN workshop 'Perspectives in Lattice QCD' in Nara during November 2005. The first lecture is on the fundamentals of Markov Chain Monte Carlo methods and introduces the Hybrid Monte Carlo (HMC) algorithm and symplectic integrators; the second lecture covers topics in approximation theory and thereb… ▽ More This is the write-up of three lectures on algorithms for dynamical fermions that were given at the ILFTN workshop 'Perspectives in Lattice QCD' in Nara during November 2005. The first lecture is on the fundamentals of Markov Chain Monte Carlo methods and introduces the Hybrid Monte Carlo (HMC) algorithm and symplectic integrators; the second lecture covers topics in approximation theory and thereby introduces the Rational Hybrid Monte Carlo (RHMC) algorithm and ways of evading integrator instabilities by means of multiple pseudofermion fields; the third lecture introduces on-shell chiral (Ginsparg-Wilson) lattice fermions and discusses five-dimensional formulations for computing fermion propagators for such fermions. △ Less

Submitted 21 February, 2012; v1 submitted 31 July, 2006; originally announced July 2006.

Comments: 62 pages, 16 figures. Published by World Scientific in proceedings of the ILFTN workshop "Perspectives in Lattice QCD". Revised version corrects (or at least changes) some sign errors, and makes a few other improvements

Light meson masses and decay constants in 2+1 flavour domain wall QCD

Authors: D. J. Antonio, K. C. Bowler, P. A. Boyle, M. A. Clark, B. Joó, A. D. Kennedy, R. D. Kenway, C. M. Maynard, R. J. Tweedie, A. Yamaguchi, RBC Collaboration, UKQCD Collaboration

Abstract: We present results for light meson masses and psedoscalar meson decay constants in 2+1 flavour domain wall QCD with the DBW2 and Iwasaki gauge actions, using lattices with linear sizes in the range 1.6 to 2.2fm and $u$ and $d$ quark masses as low as one quarter of the strange quark mass. All data were generated on the QCDOC machines at the University of Edinburgh and Brookhaven National Laborato… ▽ More We present results for light meson masses and psedoscalar meson decay constants in 2+1 flavour domain wall QCD with the DBW2 and Iwasaki gauge actions, using lattices with linear sizes in the range 1.6 to 2.2fm and $u$ and $d$ quark masses as low as one quarter of the strange quark mass. All data were generated on the QCDOC machines at the University of Edinburgh and Brookhaven National Laboratory. Despite large residual masses and a limited number of sea quark mass values with which to perform chiral extrapolations, our results agree with experiment and scale within errors. △ Less

Submitted 8 December, 2005; originally announced December 2005.

Comments: 6 pages 4 figures, PoS style, contribution to lattice 2005

Journal ref: PoS LAT2005 (2005) 080

Baryons in 2+1 flavour domain wall QCD

Authors: D. J. Antonio, K. C. Bowler, P. A. Boyle, M. A. Clark, B. Joó, A. D. Kennedy, R. D. Kenway, C. M. Maynard, R. J. Tweedie, A. Yamaguchi, RBC, UKQCD Collaborations

Abstract: We present results for some of the light baryon masses and their excited states in 2+1 flavour domain wall QCD. We considered several lattice spacings, with the DBW2 and Iwasaki gauge actions and different sea quark masses on a volume of $16^3\times32$ and a fifth dimension of size 8. All data were generated on the QCDOC machines. Despite large residual massses and a limited number of sea quark… ▽ More We present results for some of the light baryon masses and their excited states in 2+1 flavour domain wall QCD. We considered several lattice spacings, with the DBW2 and Iwasaki gauge actions and different sea quark masses on a volume of $16^3\times32$ and a fifth dimension of size 8. All data were generated on the QCDOC machines. Despite large residual massses and a limited number of sea quark mass values with which to perform chiral extrapolations, our results are in reasonable agreement with experiment and scale within errors. Finite size effects on most ensembles appear to be small. △ Less

Submitted 4 November, 2005; originally announced November 2005.

Comments: 6 pages 3 figures, PoS style, contribution to lattice 2005

Report number: PoS(LAT2005)098

Journal ref: PoSLAT2005:098,2006

Comparison of Chiral Fermion Methods

Authors: Robert G. Edwards, Balint Joo, Anthony D. Kennedy, Kostas Orginos, Urs Wenger

Abstract: We present a comparison of various five-dimensional representations of chiral fermions considering their cost and residual chiral symmetry breaking. We present a comparison of various five-dimensional representations of chiral fermions considering their cost and residual chiral symmetry breaking. △ Less

Submitted 21 October, 2005; v1 submitted 18 October, 2005; originally announced October 2005.

Comments: LATTICE 2005(Chiral Fermions), 12 pages (joint proceedings for R.G. Edwards and B. Joo)

Journal ref: PoS LAT2005:146,2006

Algorithm Shootout: R versus RHMC

Authors: M. A. Clark, Ph. de Forcrand, A. D. Kennedy

Abstract: We present initial results comparing the RHMC and R algorithms on large lattices with small quark masses using chiral fermions. We also present results concerning staggered fermions near the deconfinement/chiral phase transition. We find that the RHMC algorithm not only eliminates the step-size error of the R algorithm, but is also considerably more efficient. We discuss several possibilities fo… ▽ More We present initial results comparing the RHMC and R algorithms on large lattices with small quark masses using chiral fermions. We also present results concerning staggered fermions near the deconfinement/chiral phase transition. We find that the RHMC algorithm not only eliminates the step-size error of the R algorithm, but is also considerably more efficient. We discuss several possibilities for further improvement to the RHMC algorithm. △ Less

Submitted 30 September, 2005; originally announced October 2005.

Comments: Proceedings from Lattice 2005 (Dublin)

Journal ref: PoS LAT2005 (2005) 115

Algorithms for Lattice QCD with Dynamical Fermions

Authors: A. D. Kennedy

Abstract: We consider recent progress in algorithms for generating gauge field configurations that include the dynamical effects of light fermions. We survey what has been achieved in recent state-of-the-art computations, and examine the trade-offs between performance and control of systematic errors. We briefly review the use of polynomial and rational approximations in Hybrid Monte Carlo algorithms, and… ▽ More We consider recent progress in algorithms for generating gauge field configurations that include the dynamical effects of light fermions. We survey what has been achieved in recent state-of-the-art computations, and examine the trade-offs between performance and control of systematic errors. We briefly review the use of polynomial and rational approximations in Hybrid Monte Carlo algorithms, and some of the theory of on-shell chiral fermions on the lattice. This provides a theoretical framework within which we compare algorithmic alternatives for their implementation; and again we examine the trade-offs between speed and error control. △ Less

Submitted 30 September, 2004; originally announced September 2004.

Comments: Review presented at Lattice2004(plenary), Fermilab, June 21-26, 2004. 14 pages, 8 figures

Accelerating Fermionic Molecular Dynamics

Authors: M. A. Clark, A. D. Kennedy

Abstract: We consider how to accelerate fermionic molecular dynamics algorithms by introducing n pseudofermion fields coupled with the nth root of the fermionic kernel. This reduces the maximum pseudofermionic force, and thus allows a larger molecular dynamics integration step size without hitting an instability in the integrator. We consider how to accelerate fermionic molecular dynamics algorithms by introducing n pseudofermion fields coupled with the nth root of the fermionic kernel. This reduces the maximum pseudofermionic force, and thus allows a larger molecular dynamics integration step size without hitting an instability in the integrator. △ Less

Submitted 22 September, 2004; v1 submitted 21 September, 2004; originally announced September 2004.

Comments: 3 pages, 1 figure

Journal ref: Nucl.Phys.Proc.Suppl.140:838-840,2005

Exact 2+1 flavour RHMC simulations

Authors: M. A. Clark, A. D. Kennedy, Z. Sroczynski

Abstract: We consider the Rational Hybrid Monte Carlo algorithm for performing exact 2+1 flavour fermion simulations. The specific cases of ASQTAD and domain wall fermions are considered. We find that in both cases the naive performance is similar to conventional hybrid algorithms. We consider the Rational Hybrid Monte Carlo algorithm for performing exact 2+1 flavour fermion simulations. The specific cases of ASQTAD and domain wall fermions are considered. We find that in both cases the naive performance is similar to conventional hybrid algorithms. △ Less

Submitted 22 September, 2004; v1 submitted 21 September, 2004; originally announced September 2004.

Comments: 3 pages, no figures

Journal ref: Nucl.Phys.Proc.Suppl.140:835-837,2005

Fast Evaluation of Zolotarev Coefficients

Authors: A. D. Kennedy

Abstract: We review the theory of elliptic functions leading to Zolotarev's formula for the sign function over the range (ε\leq|x| \leq1). We show how Gauss' arithmetico-geometric mean allows us to evaluate elliptic functions cheaply, and thus to compute Zolotarev coefficients ``on the fly'' as a function of (ε). This in turn allows us to calculate the matrix functions (\sgn H), (\sqrt H), and (1/\sqrt H)… ▽ More We review the theory of elliptic functions leading to Zolotarev's formula for the sign function over the range (ε\leq|x| \leq1). We show how Gauss' arithmetico-geometric mean allows us to evaluate elliptic functions cheaply, and thus to compute Zolotarev coefficients ``on the fly'' as a function of (ε). This in turn allows us to calculate the matrix functions (\sgn H), (\sqrt H), and (1/\sqrt H) both quickly and accurately for any Hermitian matrix (H) whose spectrum lies in the specified range. △ Less

Submitted 14 December, 2004; v1 submitted 27 February, 2004; originally announced February 2004.

Comments: 21 pages, 2 figures. Made various minor corrections

Approximation Theory for Matrices

Authors: A. D. Kennedy

Abstract: We review the theory of optimal polynomial and rational Chebyshev approximations, and Zolotarev's formula for the sign function over the range (ε\leq |z| \leq1). We explain how rational approximations can be applied to large sparse matrices efficiently by making use of partial fraction expansions and multi-shift Krylov space solvers. We review the theory of optimal polynomial and rational Chebyshev approximations, and Zolotarev's formula for the sign function over the range (ε\leq |z| \leq1). We explain how rational approximations can be applied to large sparse matrices efficiently by making use of partial fraction expansions and multi-shift Krylov space solvers. △ Less

Submitted 27 February, 2004; originally announced February 2004.

Comments: 10 pages, 7 figures

Journal ref: Nucl.Phys.Proc.Suppl. 128C (2004) 107-116

Structure functions and form factors close to the chiral limit from lattice QCD

Authors: T. Bakeyev, D. Galletly, M. Göckeler, M. Gürtler, R. Horsley, B. Joó, A. D. Kennedy, B. Pendleton, H. Perlt, D. Pleiter, P. E. L. Rakow, G. Schierholz, A. Schiller, T. Streuer, H. Stüben

Abstract: Results for nucleon matrix elements (arising from moments of structure functions) and form factors from a mixture of runs using Wilson, clover and overlap fermions (both quenched and unquenched) are presented and compared in an effort to explore the size of the chiral `regime', lattice spacing errors and quenching artefacts. While no run covers this whole range of effects the partial results ind… ▽ More Results for nucleon matrix elements (arising from moments of structure functions) and form factors from a mixture of runs using Wilson, clover and overlap fermions (both quenched and unquenched) are presented and compared in an effort to explore the size of the chiral `regime', lattice spacing errors and quenching artefacts. While no run covers this whole range of effects the partial results indicate a picture of small lattice spacing errors, small quenching effects and only reaching the chiral regime at rather light quark masses. △ Less

Submitted 8 March, 2004; v1 submitted 11 November, 2003; originally announced November 2003.

Comments: 7 pages, 7 figures; contribution to the 2003 Workshop on Lattice Hadron Physics, Cairns, Australia; error in Fig. 4 corrected; minor text changes

Report number: DESY 03-183, Edinburgh 2003/19, LTH 610, LU-ITP 2003/26

Journal ref: Nucl.Phys.Proc.Suppl.128:82-88,2004

Quark spectra and light hadron phenomenology from overlap fermions with improved gauge field action

Authors: D. Galletly, M. Gürtler, R. Horsley, B. Joo, A. D. Kennedy, H. Perlt, B. J. Pendleton, P. E. L. Rakow, G. Schierholz, A. Schiller, T. Streuer

Abstract: We present first results from a simulation of quenched overlap fermions with improved gauge field action. Among the quantities we study are the spectral properties of the overlap operator, the chiral condensate and topological charge, quark and hadron masses, and selected nucleon matrix elements. To make contact with continuum physics, we compute the renormalization constants of quark bilinear o… ▽ More We present first results from a simulation of quenched overlap fermions with improved gauge field action. Among the quantities we study are the spectral properties of the overlap operator, the chiral condensate and topological charge, quark and hadron masses, and selected nucleon matrix elements. To make contact with continuum physics, we compute the renormalization constants of quark bilinear operators in perturbation theory and beyond. △ Less

Submitted 8 January, 2004; v1 submitted 8 October, 2003; originally announced October 2003.

Comments: Lattice2003(chiral); 12 pages (espcrc2.sty), based on talks given by M.G., R.H., H.P. and T.S. at Lattice 2003; error in the first moment <x> of the nucleon structure function corrected

Report number: DESY 03-161, Edinburgh 2003/15, LU-ITP 2003/017, LTH 593

Journal ref: Nucl.Phys.Proc.Suppl.129:453-455,2004

The RHMC Algorithm for 2 Flavours of Dynamical Staggered Fermions

Authors: M. A. Clark, A. D. Kennedy

Abstract: We describe an implementation of the Rational Hybrid Monte Carlo (RHMC) algorithm for dynamical computations with two flavours of staggered quarks. We discuss several variants of the method, the performance and possible sources of error for each of them, and we compare the performance and results to the inexact R algorithm. We describe an implementation of the Rational Hybrid Monte Carlo (RHMC) algorithm for dynamical computations with two flavours of staggered quarks. We discuss several variants of the method, the performance and possible sources of error for each of them, and we compare the performance and results to the inexact R algorithm. △ Less

Submitted 28 September, 2003; v1 submitted 15 September, 2003; originally announced September 2003.

Comments: Lattice2003(machine) 3 pages, 1 figure. Added reference

Diagrammatic Young Projection Operators for U(n)

Authors: Henriette Elvang, Predrag Cvitanović, Anthony D. Kennedy

Abstract: We utilize a diagrammatic notation for invariant tensors to construct the Young projection operators for the irreducible representations of the unitary group U(n), prove their uniqueness, idempotency, and orthogonality, and rederive the formula for their dimensions. We show that all U(n) invariant scalars (3n-j coefficients) can be constructed and evaluated diagrammatically from these U(n) Young… ▽ More We utilize a diagrammatic notation for invariant tensors to construct the Young projection operators for the irreducible representations of the unitary group U(n), prove their uniqueness, idempotency, and orthogonality, and rederive the formula for their dimensions. We show that all U(n) invariant scalars (3n-j coefficients) can be constructed and evaluated diagrammatically from these U(n) Young projection operators. We prove that the values of all U(n) 3n-j coefficients are proportional to the dimension of the maximal representation in the coefficient, with the proportionality factor fully determined by its S[k] symmetric group value. We also derive a family of new sum rules for the 3-j and 6-j coefficients, and discuss relations that follow from the negative dimensionality theorem. △ Less

Submitted 19 July, 2003; originally announced July 2003.

Comments: 14 pages, 225 figures

The Paraldor Project

Authors: T. Ashby, D. Galletly, B. Joo, A. D. Kennedy, G. Lacagnina

Abstract: Paraldor is an experiment in bringing the power of categorical languages to lattice QCD computations. Our target language is Aldor, which allows the capture of the mathematical structure of physics directly in the structure of the code using the concepts of categories, domains and their inter-relationships in a way which is not otherwise possible with current popular languages such as Fortran, C… ▽ More Paraldor is an experiment in bringing the power of categorical languages to lattice QCD computations. Our target language is Aldor, which allows the capture of the mathematical structure of physics directly in the structure of the code using the concepts of categories, domains and their inter-relationships in a way which is not otherwise possible with current popular languages such as Fortran, C, C++ or Java. By writing high level physics code portably in Aldor, and implementing switchable machine dependent high performance back-ends in C or assembler, we gain all the power of categorical languages such as modularity, portability, readability and efficiency. △ Less

Submitted 3 September, 2002; originally announced September 2002.

Comments: 4 pages, 2 figures, Lattice 2002 conference proceedings

Report number: Edinburgh {2002/14}

Comparing the R algorithm and RHMC for staggered fermions

Authors: M. A. Clark, B. Joo, A. D. Kennedy

Abstract: The R algorithm is widely used for simulating two flavours of dynamical staggered fermions. We give a simple proof that the algorithm converges to the desired probability distribution to within O(dt^2) errors, but show that the relevant expansion parameter is (dt/m)^2, m being the quark mass. The Rational Hybrid Monte Carlo (RHMC) algorithm provides an exact (i.e., has no step size errors) alter… ▽ More The R algorithm is widely used for simulating two flavours of dynamical staggered fermions. We give a simple proof that the algorithm converges to the desired probability distribution to within O(dt^2) errors, but show that the relevant expansion parameter is (dt/m)^2, m being the quark mass. The Rational Hybrid Monte Carlo (RHMC) algorithm provides an exact (i.e., has no step size errors) alternative for simulating the square root of the staggered Dirac operator. We propose using it to test the validity of the R algorithm for simulations carried out with dt m. △ Less

Submitted 3 September, 2002; originally announced September 2002.

Comments: 3 pages, proceedings from Lattice 2002 poster presentation

The overlap operator as a continued fraction

Authors: A. Borici, A. D. Kennedy, B. J. Pendleton, U. Wenger

Abstract: We use a continued fraction expansion of the sign-function in order to obtain a five dimensional formulation of the overlap lattice Dirac operator. Within this formulation the inverse of the overlap operator can be calculated by a single Krylov space method where nested conjugate gradient procedures are avoided. We show that the five dimensional linear system can be made well conditioned using e… ▽ More We use a continued fraction expansion of the sign-function in order to obtain a five dimensional formulation of the overlap lattice Dirac operator. Within this formulation the inverse of the overlap operator can be calculated by a single Krylov space method where nested conjugate gradient procedures are avoided. We show that the five dimensional linear system can be made well conditioned using equivalence transformations on the continued fractions. This is of significant importance when dynamical overlap fermions are simulated. △ Less

Submitted 12 October, 2001; originally announced October 2001.

Comments: 3 pages, 1 figure, talk presented by U. Wenger at Lattice2001(chiral)

Journal ref: Nucl.Phys.Proc.Suppl. 106 (2002) 757-759

Cost of the Generalised Hybrid Monte Carlo Algorithm for Free Field Theory

Authors: A. D. Kennedy, Brian Pendleton

Abstract: We study analytically the computational cost of the Generalised Hybrid Monte Carlo (GHMC) algorithm for free field theory. We calculate the Metropolis acceptance probability for leapfrog and higher-order discretisations of the Molecular Dynamics (MD) equations of motion. We show how to calculate autocorrelation functions of arbitrary polynomial operators, and use these to optimise the GHMC momen… ▽ More We study analytically the computational cost of the Generalised Hybrid Monte Carlo (GHMC) algorithm for free field theory. We calculate the Metropolis acceptance probability for leapfrog and higher-order discretisations of the Molecular Dynamics (MD) equations of motion. We show how to calculate autocorrelation functions of arbitrary polynomial operators, and use these to optimise the GHMC momentum mixing angle, the trajectory length, and the integration stepsize for the special cases of linear and quadratic operators. We show that long trajectories are optimal for GHMC, and that standard HMC is more efficient than algorithms based on Second Order Langevin Monte Carlo (L2MC), sometimes known as Kramers Equation. We show that contrary to naive expectations HMC and L2MC have the same volume dependence, but their dynamical critical exponents are z = 1 and z = 3/2 respectively. △ Less

Submitted 21 August, 2000; originally announced August 2000.

Comments: 54 pages, 3 figures

Report number: Edinburgh 2000/16

Journal ref: Nucl.Phys. B607 (2001) 456-510

Instability in the Molecular Dynamics Step of Hybrid Monte Carlo in Dynamical Fermion Lattice QCD Simulations

Authors: Balint Joo, Brian Pendleton, Anthony D. Kennedy, Alan C. Irving, James C. Sexton, Stephen M. Pickles, Stephen P. Booth, UKQCD Collaboration

Abstract: We investigate instability and reversibility within Hybrid Monte Carlo simulations using a non-perturbatively improved Wilson action. We demonstrate the onset of instability as tolerance parameters and molecular dynamics step sizes are varied. We compare these findings with theoretical expectations and present limits on simulation parameters within which a stable and reversible algorithm is obta… ▽ More We investigate instability and reversibility within Hybrid Monte Carlo simulations using a non-perturbatively improved Wilson action. We demonstrate the onset of instability as tolerance parameters and molecular dynamics step sizes are varied. We compare these findings with theoretical expectations and present limits on simulation parameters within which a stable and reversible algorithm is obtained for physically relevant simulations. Results of optimisation experiments with respect to tolerance prarameters are also presented. △ Less

Submitted 26 May, 2000; originally announced May 2000.

Comments: RevTeX, Some results here were presented at Vielat 99, Vienna, Austria, Sept 1999 22 Pages, 10 figures, to be submitted to PRD

Report number: Edinburgh 2000/12, LTH 455

Journal ref: Phys.Rev. D62 (2000) 114501

Cost of Generalised HMC Algorithms for Free Field Theory

Authors: A. D. Kennedy, Brian Pendleton

Abstract: We study analytically the computational cost of the Generalised Hybrid Monte Carlo (GHMC) algorithm for free field theory. We calculate the autocorrelation functions of operators quadratic in the fields, and optimise the GHMC momentum mixing angle, the trajectory length, and the integration stepsize. We show that long trajectories are optimal for GHMC, and that standard HMC is much more efficien… ▽ More We study analytically the computational cost of the Generalised Hybrid Monte Carlo (GHMC) algorithm for free field theory. We calculate the autocorrelation functions of operators quadratic in the fields, and optimise the GHMC momentum mixing angle, the trajectory length, and the integration stepsize. We show that long trajectories are optimal for GHMC, and that standard HMC is much more efficient than algorithms based on the Second Order Langevin (L2MC) or Kramers Equation. We show that contrary to naive expectations HMC and L2MC have the same volume dependence, but their dynamical critical exponents are z=1 and z=3/2 respectively. △ Less

Submitted 28 January, 2000; originally announced January 2000.

Comments: LATTICE99(Algorithms and Machines) - 3 pages, 1 PostScript figure

Report number: Edinburgh 2000/04

Journal ref: Nucl.Phys.Proc.Suppl.83:816-818,2000

Status of the QCDSP project

Authors: Dong Chen, ** Chen, Norman H. Christ, Robert G. Edwards, George R. Fleming, Alan Gara, Sten Hansen, Chulwoo Jung, Adrian L. Kaehler, Anthony D. Kennedy, Gregory W. Kilcup, Yubing Luo, Catalin I. Malureanu, Robert D. Mawhinney, John Parsons, ChengZhong Sui, Pavlos M. Vranas, Yuri Zhestkov

Abstract: We describe the completed 8,192-node, 0.4Tflops machine at Columbia as well as the 12,288-node, 0.6Tflops machine assembled at the RIKEN Brookhaven Research Center. Present performance as well as our experience in commissioning these large machines is presented. We outline our on-going physics program and explain how the configuration of the machine is varied to support a wide range of lattice Q… ▽ More We describe the completed 8,192-node, 0.4Tflops machine at Columbia as well as the 12,288-node, 0.6Tflops machine assembled at the RIKEN Brookhaven Research Center. Present performance as well as our experience in commissioning these large machines is presented. We outline our on-going physics program and explain how the configuration of the machine is varied to support a wide range of lattice QCD problems, requiring a variety of machine sizes. Finally a brief discussion is given of future prospects for large-scale lattice QCD machines. △ Less

Submitted 1 October, 1998; originally announced October 1998.

Comments: LATTICE98(machines), 3 pages, 1 picture, 1 figure

Report number: CU-TP-911

Journal ref: Nucl.Phys.Proc.Suppl. 73 (1999) 898-900

A New Exact Method for Dynamical Fermion Computations with Non-Local Actions

Authors: Ivan Horvath, A. D. Kennedy, Stefan Sint

Abstract: We introduce a new algorithm which we call the {Rational Hybrid Monte Carlo} Algorithm (RHMC). This method uses a rational approximation to the fermionic kernel together with a noisy Kennedy-Kuti acceptance step to give an efficient algorithm with no molecular dynamics integration step-size errors. We introduce a new algorithm which we call the {Rational Hybrid Monte Carlo} Algorithm (RHMC). This method uses a rational approximation to the fermionic kernel together with a noisy Kennedy-Kuti acceptance step to give an efficient algorithm with no molecular dynamics integration step-size errors. △ Less

Submitted 14 September, 1998; originally announced September 1998.

Comments: LATTICE98(algorithms)

Journal ref: Nucl.Phys.Proc.Suppl. 73 (1999) 834-836

QCDSP - A status report

Authors: QCDSP Collaboration, Dong Chen, ** Chen, Norman H. Christ, Robert G. Edwards, George R. Fleming, Alan Gara, Sten Hansen, Chulwoo Jung, Adrian L. Kaehler, Anthony D. Kennedy, Gregory W. Kilcup, Yubing Luo, Catalin I. Malureanu, Robert D. Mawhinney, John Parsons, James C. Sexton, ChengZhong Sui, Pavlos M. Vranas

Abstract: The QCDSP machine at Columbia University has grown to 2,048 nodes achieving a peak speed of 100 Gigaflops. Software for quenched and Hybrid Monte Carlo (HMC) evolution schemes has been developed for staggered fermions, with support for Wilson and clover fermions under development. We provide an overview of the runtime environment, the current status of the QCDSP construction program and prelimin… ▽ More The QCDSP machine at Columbia University has grown to 2,048 nodes achieving a peak speed of 100 Gigaflops. Software for quenched and Hybrid Monte Carlo (HMC) evolution schemes has been developed for staggered fermions, with support for Wilson and clover fermions under development. We provide an overview of the runtime environment, the current status of the QCDSP construction program and preliminary results not presented elsewhere in these proceedings. △ Less

Submitted 28 September, 1997; originally announced September 1997.

Comments: 3 pages, 1 figure, LaTeX with espcrc2.sty, talk presented at LATTICE97(Poster) by G. R. Fleming

Report number: CU-TP-860

Journal ref: Nucl.Phys.Proc.Suppl. 63 (1998) 997-999

The LHMC Algorithm for Free Field Theory: Reexamining Overrelaxation

Authors: Ivan Horvath, A. D. Kennedy

Abstract: We analyze the autocorrelations for the LHMC algorithm in the context of free field theory. In this case this is just Adler's overrelaxation algorithm. We consider the algorithm with even/odd, lexicographic, and random updates, and show that its efficiency depends crucially on this ordering of sites when optimized for a given class of operators. In particular, we show that, contrary to previous… ▽ More We analyze the autocorrelations for the LHMC algorithm in the context of free field theory. In this case this is just Adler's overrelaxation algorithm. We consider the algorithm with even/odd, lexicographic, and random updates, and show that its efficiency depends crucially on this ordering of sites when optimized for a given class of operators. In particular, we show that, contrary to previous expectations, it is possible to eliminate critical slowing down (z[int]=0) for a class of interesting observables, including the magnetic susceptibility: this can be done with lexicographic updates but is not possible with even/odd (z[int]=1) or random (z[int]=2) updates. We are considering the dynamical critical exponent z[int] for integrated autocorrelations rather than for the exponential autocorrelation time; this is reasonable because it is the integrated autocorrelation which determines the cost of a Monte Carlo computation. △ Less

Submitted 27 August, 1997; originally announced August 1997.

Comments: LaTeX, 33 pages, 3 postscript figures

Report number: FSU-SCRI-97-94

Journal ref: Nucl.Phys. B510 (1998) 367-400

A Simple Proof of the BPH Theorem

Authors: A. D. Kennedy

Abstract: A new formalism is given for the renormalization of quantum field theories to all orders of perturbation theory, in which there are manifestly no overlap** divergences. We prove the BPH theorem in this formalism, and show how the local subtractions add up to counterterms in the action. Applications include the renormalization of lattice perturbation theory, the decoupling theorem, Zimmermann o… ▽ More A new formalism is given for the renormalization of quantum field theories to all orders of perturbation theory, in which there are manifestly no overlap** divergences. We prove the BPH theorem in this formalism, and show how the local subtractions add up to counterterms in the action. Applications include the renormalization of lattice perturbation theory, the decoupling theorem, Zimmermann oversubtraction, the renormalization of operator insertions, and the operator product expansion. △ Less

Submitted 10 December, 1996; originally announced December 1996.

Comments: 8 pages

Report number: FSU-SCRI-96-132