Application of the Lanczos algorithm for solving the linear systems that occur in continuation problems

We study the Lanczos method for solving symmetric linear systems with multiple right‐hand sides. First, we propose a numerical method of implementing the Lanczos method, which can provide all approximations to the solution vectors of the remaining linear systems. We also seek possible application of this algorithm for solving the linear systems that occur in continuation problems. Sample numerical results are reported. Copyright © 2002 John Wiley & Sons, Ltd.     

Breakdown and near-breakdown control in the CGS algorithm using stochastic arithmetic

In the Conjugate-Gradient-Squared method, a sequence of residualsr _k defined byr _k=P _k ² (A)r₀ is computed. Coefficients of the polynomialsP _k may be computed as a ratio of scalar products from the theory of formal orthogonal polynomials. When a scalar product in a denominator is zero or very affected by round-off errors, situations of breakdown or near-breakdown appear. Using floating point arithmetic on computers, such situations are detected with the use of _i in some ordering relations like |x _i . The user has to choose the _i himself and these choices condition entirely the efficient detection of breakdown or near-breakdown. The subject of this paper is to show how stochastic arithmetic eliminates the problem of the _i with the estimation of the accuracy of some intermediate results.     

Updating the QR decomposition of block tridiagonal and block Hessenberg matrices

We present an efficient block-wise update scheme for the QR decomposition of block tridiagonal and block Hessenberg matrices. For example, such matrices come up in generalizations of the Krylov space solvers MinRes, SymmLQ, GMRes, and QMR to block methods for linear systems of equations with multiple right-hand sides. In the non-block case it is very efficient (and, in fact, standard) to use Givens rotations for these QR decompositions. Normally, the same approach is also used with column-wise updates in the block case. However, we show that, even for small block sizes, block-wise updates using (in general, complex) Householder reflections instead of Givens rotations are far more efficient in this case, in particular if the unitary transformations that incorporate the reflections determined by a whole block are computed explicitly. Naturally, the bigger the block size the bigger the savings. We discuss the somewhat complicated algorithmic details of this block-wise update, and present numerical experiments on accuracy and timing for the various options (Givens vs. Householder, block-wise vs. column-wise update, explicit vs. implicit computation of unitary transformations). Our treatment allows variable block sizes and can be adapted to block Hessenberg matrices that do not have the special structure encountered in the above mentioned block Krylov space solvers.     

Some potentials for the curvature tensor on three-dimensional manifolds

We study equations of Riemann–Lanczos type on three dimensional manifolds. Obstructions to global existence for global Lanczos potentials are pointed out. We check that the imposition of the original Lanczos symmetries on the potential leads to equations which do not have a determined type, leading to problems when trying to prove global existence. We show that elliptic equations can be obtained by relaxing those symmetry requirements in at least two different ways, leading to global existence of potentials under natural conditions. A second order potential for the Ricci tensor is introduced.     

Cornelius Lanczos’s Derivation of the Usual Action Integral of Classical Electrodynamics

The usual action integral of classical electrodynamics is derived starting from Lanczos’s electrodynamics – a pure field theory in which charged particles are identified with singularities of the homogeneous Maxwell’s equations interpreted as a generalization of the Cauchy–Riemann regularity conditions from complex to biquaternion functions of four complex variables. It is shown that contrary to the usual theory based on the inhomogeneous Maxwell’s equations, in which charged particles are identified with the sources, there is no divergence in the self-interaction so that the mass is finite, and that the only approximations made in the derivation are the usual conditions required for the internal consistency of classical electrodynamics. Moreover, it is found that the radius of the boundary surface enclosing a singularity interpreted as an electron is on the same order as that of the hypothetical “bag” confining the quarks in a hadron, so that Lanczos’s electrodynamics is engaging the reconsideration of many fundamental concepts related to the nature of elementary particles.     

An efficient method to set up a Lanczos based preconditioner for discrete ill-posed problems

Rezghi and Hosseini [M. Rezghi, S.M. Hosseini, Lanczos based preconditioner for discrete ill-posed problems, Computing 88 (2010) 79–96] presented a Lanczos based preconditioner for discrete ill-posed problems. Their preconditioner is constructed by using few steps (e.g., k) of the Lanczos bidiagonalization and corresponding computed singular values and right Lanczos vectors. In this article, we propose an efficient method to set up such preconditioner. Some numerical examples are given to show the effectiveness of the method.     

New results on the convergence of the conjugate gradient method

This paper is concerned with proving theoretical results related to the convergence of the conjugate gradient (CG) method for solving positive definite symmetric linear systems. Considering the inverse of the projection of the inverse of the matrix, new relations for ratios of the A‐norm of the error and the norm of the residual are provided, starting from some earlier results of Sadok (Numer. Algorithms 2005; 40 :201–216). The proofs of our results rely on the well‐known correspondence between the CG method and the Lanczos algorithm. Copyright © 2008 John Wiley & Sons, Ltd.     

Comment on Formulating and Generalizing Dirac's, Proca's, and Maxwell's Equations with Biquaternions or Clifford Numbers

Many difficulties of interpretation met by contemporary researchers attempting to recast or generalize Dirac's, Proca's, or Maxwell's theories using biquaternions or Clifford numbers have been encountered long ago by a number of physicists including Lanczos, Proca, and Einstien. In the modern approach initiated by Gürsey, these difficulties are solved by recognizing that most generalizations lead to theories describing superpositions of particles of different intrinsic spin and isospin, so that the correct interpretation emerges from the requirement of full Poincaré covariance, including space and time reversal, as well as reversion and gauge invariance. For instance, the doubling of the number of solutions implied by the simplest generalization of Dirac's equation (i.e., Lanczos's equation) can be interpreted as isospin. In this approach, biquaternions and Clifford numbers become powerful opportunities to formulate the Standard Model of elementary particles, as well as many of its possible generalizations, in very elegant and compact ways.     

Lanczos,Householder transformations,and implicit deflation for fast and reliable dominant singular subspace computation

Many applications, such as subspace‐based models in information retrieval and signal processing, require the computation of singular subspaces associated with the k dominant, or largest, singular values of an m×n data matrix A, where k?min(m,n). Frequently, A is sparse or structured, which usually means matrix–vector multiplications involving A and its transpose can be done with much less than ??(mn) flops, and A and its transpose can be stored with much less than ??(mn) storage locations. Many Lanczos‐based algorithms have been proposed through the years because the underlying Lanczos method only accesses A and its transpose through matrix–vector multiplications. We implement a new algorithm, called KSVD, in the Matlab environment for computing approximations to the singular subspaces associated with the k dominant singular values of a real or complex matrix A. KSVD is based upon the Lanczos tridiagonalization method, the WY representation for storing products of Householder transformations, implicit deflation, and the QR factorization. Our Matlab simulations suggest it is a fast and reliable strategy for handling troublesome singular‐value spectra. Copyright © 2001 John Wiley & Sons, Ltd.     

A Variant of the Inverted Lanczos Method

In this note we study a variant of the inverted Lanczos method which computes eigenvalue approximates of a symmetric matrix A as Ritz values of A from a Krylov space of A ^–1. The method turns out to be slightly faster than the Lanczos method at least as long as reorthogonalization is not required. The method is applied to the problem of determining the smallest eigenvalue of a symmetric Toeplitz matrix. It is accelerated taking advantage of symmetry properties of the correspond ng eigenvector.This revised version was published online in October 2005 with corrections to the Cover Date.