Perturbation d'une matrice hermitienne ou normale

Summary Given a hermitian (normal) matrixA with known eigenelements, we study the behavior of these elements under a hermitian perturbationH. With a symmetric 2×2 matrix, the problem is explicit (algebraic equation of 2nd degree), and we try, in the case of ann×n hermitian matrix, to obtain upper bounds which are as close as possible of exact results forn=2. The results are collected in § I. They state in a unified manner some results of Davis [2], Gavurin [4], Golub [5], Kato-Temple [7], Ortega [3], Wilkinson [8] and others.In § II, we apply this theory to produce error bounds for eigenelements computed by theQR and Jacobi methods. The given error bounds are realistic and easy to compute during the algorithms. When the separation distance between eigenvalues ofA approaches zero, the problem of computing eigenelements ofA+H is ill-conditionned with respect to the eigenvalues, the eigenvectors being orthogonal. The precision then obtained is given.

---

Perturbation d'une matrice hermitienne ou normale

     

Approximation of signals of class Lip(α, p) by linear operators

Mittal, Rhoades [5], [6], [7] and [8] and Mittal et al. [9] and [10] have initiated a study of error estimates E_n(f) through trigonometric-Fourier approximation (tfa) for the situations in which the summability matrix T does not have monotone rows. In this paper we continue the work. Here we extend two theorems of Leindler [4], where he has weakened the conditions on {p_n} given by Chandra [2], to more general classes of triangular matrix methods. Our Theorem also partially generalizes Theorem 4 of Mittal et al. [11] by dropping the monotonicity on the elements of matrix rows, which in turn generalize the results of Quade [15].     

Combinatorial obstructions to the lifting of weaving diagrams

This paper deals with various connections of oriented matroids [3] and weaving diagrams of lines in space [9], [16], [27]. We encode the litability problem of a particular weaving diagramD onn lines by the realizability problem of a partial oriented matroid χ _D with2n elements in rank 4. We prove that the occurrence of a certain substructure inD implies that χD is noneuclidean in the sense of Edmonds, Fukuda, and Mandel [12], [14]. Using this criterion we construct an infinite class of minor-minimal noneuclidean oriented matroids in rank 4. Finally, we give an easy algebraic proof for the nonliftability of the alternating weaving diagram on a bipartite grid of 4×4 lines [16].     

Superconvergence for rectangular mixed finite elements

Summary In this paper we prove superconvergence error estimates for the vector variable for mixed finite element approximations of second order elliptic problems. For the rectangular finite elements of Raviart and Thomas [19] and for those of Brezzi et al. [4] we prove that the distance inL ² between the approximate solution and a projection of the exact one is of higher order than the error itself.This result is exploited to obtain superconvergence at Gaussian points and to construct higher order approximations by a local postprocessing.     

Solution of Maxwell equation in axisymmetric geometry by Fourier series decompostion and by use of H (rot) conforming finite element

Summary. This study deals with the mathematical and numerical solution of time-harmonic Maxwell equation in axisymmetric geometry. Using Fourier decomposition, we define weighted Sobolev spaces of solution and we prove expected regularity results. A practical contribution of this paper is the construction of a class of finite element conforming with the H (rot) space equipped with the weighted measure rdrdz. It appears as an extension of the well-known cartesian mixed finite element of Raviart-Thomas-Nédélec [11]–[15]. These elements are built from classical lagrangian and mixed finite element, therefore no special approximations functions are needed. Finally, following works of Mercier and Raugel [10], we perform an interpolation error estimate for the simplest proposed element. Received March 15, 1996 / Revised version received November 30, 1998 / Published online December 6, 1999     

Ingham-Beurling type theorems with weakened gap conditions

Completing a series of works begun by Wiener [34], Paley and Wiener [28] and Ingham [9], a far-reaching generalization of Parseval"s identity was obtained by Beurling [4] for nonharmonic Fourier series whose exponents satisfy a uniform gap condition. Later this gap condition was weakened by Ullrich [33], Castro and Zuazua [5], Jaffard, Tucsnak and Zuazua [11] and then in [2] in some particular cases. In this paper we prove a general theorem which contains all previous results. Furthermore, applying a different method, we prove a variant of this theorem for nonharmonic Fourier series with vector coefficients. This result, partly motivated by control-theoretical applications, extends several earlier results obtained in [15] and [2]. Finally, applying these results we obtain an optimal simultaneous observability theorem concerning a system of vibrating strings. This revised version was published online in June 2006 with corrections to the Cover Date.     

FINITE RINGS WITH COMMUTING NILPOTENT ELEMENTS

As a generalization of Wedderburn's theorem, Herstein [5] proved that a finite ring R is commutative, if all nilpotent elements are contained in the center of R. However a finite ring with commuting nilpotent elements is not necessarily commutative. Recently, in [9] and [10], Simons described the structure of finite rings R with J(R)² = 0 in a variety with definable principal congruences. In this paper, we will consider the difference between the finite commutative rings and the finite rings in which any two nilpotent elements commute with each other. As a consequence, we describe the structure of finite rings R with [J(R), J(R)] = 0 in a variety with definable principal congruences.     

First and second order sufficient conditions for optimal control and the calculus of variations

In this paper we develop first and second order sufficient conditions for optimal control and the calculus of variations problems. Our conditions are derived from the Hamilton-Jacobi approach [15, Thm. 2], which was obtained for the generalized problem of Bolza. We do not require any convexity on the data [7] and [11], or that the control setU is polyhedral [14], or that the control function is in the interior ofU [8]. Instead, we assume a certain inequality which is satisfied in each of the above mentioned cases.The publication of this report has been made possible due to a grant of the Fonds FCAC for the help and support of research.     

Superconvergence and a posteriori error estimation for triangular mixed finite elements

Summary. In this paper,we prove superconvergence results for the vector variable when lowest order triangular mixed finite elements of Raviart-Thomas type [17] on uniform triangulations are used, i.e., that the -distance between the approximate solution and a suitable projection of the real solution is of higher order than the -error. We prove results for both Dirichlet and Neumann boundary conditions. Recently, Duran [9] proved similar results for rectangular mixed finite elements, and superconvergence along the Gauss-lines for rectangular mixed finite elements was considered by Douglas, Ewing, Lazarov and Wang in [11], [8] and [18]. The triangular case however needs some extra effort. Using the superconvergence results, a simple postprocessing of the approximate solution will give an asymptotically exact a posteriori error estimator for the -error in the approximation of the vector variable. Received December 6, 1992 / Revised version received October 2, 1993     

Asymptotic expansions for classical and generalized divided differences including applications

It is a well-known fact that the classical (i.e. polynomial) divided difference of orderm, when applied to a functiong, converges to themth-derivative of this function, if the evaluation points all collapse to a single one.In the first part of this paper we shall sharpen this result in the sense that we prove the existence of an asymptotic expansion with limitg ^(m) /m!. This result allows the application of extrapolation methods for the numerical differentiation of funtions.Moreover, in the second and main part of the paper we study generalized divided differences, which were introduced by Popoviciu [10] and further investigated for example by Karlin [2], Walz [15] and, mainly, Mühlbach [6–8]; we prove the existence of an asymptotic expansion also for these generalized divided differences, if the underlying function space is a Polya space. As a by-product, our results show that the generalized divided difference of orderm converges to the value of a certainmth order differential operator.     

A Systematic “Saddle Point Near a Pole” Asymptotic Method with Application to the Gauss Hypergeometric Function

In recent works [ 1 ] and [ 2 ], we have proposed more systematic versions of the Laplace’s and saddle point methods for asymptotic expansions of integrals. Those variants of the standard methods avoid the classical change of variables and give closed algebraic formulas for the coefficients of the expansions. In this work we apply the ideas introduced in [ 1 ] and [ 2 ] to the uniform method “saddle point near a pole.” We obtain a computationally more systematic version of that uniform asymptotic method for integrals having a saddle point near a pole that, in many interesting examples, gives a closed algebraic formula for the coefficients. The asymptotic sequence is given, in general, in terms of exponential integrals of fractional order (or incomplete gamma functions). In particular, when the order of the saddle point is two, the basic approximant is given in terms of the error function (as in the standard method). As an application, we obtain new asymptotic expansions of the Gauss Hypergeometric function ₂F₁(a, b, c; z) for large b and c with c > b + 1 .     

Relations among homogeneity,collapsibility and nonconfounding in distribution effects

In this paper, the concept of distribution effect is proposed without the causal diagram. Following the notation of Stone [11], we assume that the exposure treatment X is an unknown deterministic function of the confounder set Pa(X) and a random error ε. We discuss sufficient and necessary conditions for homogeneity, collapsibility and nonconfounding for distribution effects and discuss relations among them.     

Absolutely continuous functions of several variables and diffeomorphisms

In [4], a class of absolutely continuous functions of d-variables, motivated by applications to change of variables in an integral, has been introduced. The main result of this paper states that absolutely continuous functions in the sense of [4] are not stable under diffeomorphisms. We also show an example of a function which is absolutely continuous with respect cubes but not with respect to balls.     

A unified theory of weak continuity for functions

In this paper we introduce a new notion of weaklyM-continuous functions as functions from a set satisfying some minimal conditions into a set satisfying some minimal conditions. We obtain some characterizations and several properties of such functions. This function leads to the formulation of a unified theory of weak continuity [27], almosts-continuity [43],p(θ)-continuity [10] andp-continuity [59].     

The direct product of a hopfian group with a group with cyclic ccnter

In this paper we continue our study of hopficity begun in [1], [2], [3], [4] and [5]. LetA be hopfian and letB have a cyclic center of prime power order. We improve Theorem 4 of [2] by showing that ifB has finitely many normal subgroups which form a chain (we sayB isn-normal), thenAxB is hopfian. We then consider the case whenB is ap-group of nilpotency class 2 and show that in certain casesAxB is hopfian.     

Semilattice decompositions of semigroups

The purpose of this paper is to develop a general theory of semilattice decompositions of semigroups from the point of view of obtaining theorems of the type: A semigroup S has propertyD if and only if S is a semilattice of semigroups having property β. As such we are able to extend the theories of Clifford [3], Andersen [1], Croisot [5], Tamura and Kimura [14], Petrich [9], Chrislock [2], Tamura and Shafer [15], Iyengar [7] and Weissglass and the author [10]. The root of our whole theory is Tamura's semilattice decomposition theorem [12, 13]. Of this, we give a new proof. The results of this paper were obtained by the author between January and July of 1971, while an undergraduate at the University of California, Santa Barbara.     

Generalized Heteroclinic Cycles in Spherically Invariant Systems and Their Perturbations

Summary. In this paper we want to investigate the effects of forced symmetry-breaking perturbations—see Lauterbach & Roberts [29], as well as [28], [31]—on the heteroclinic cycle which was found in the l = 1 , l = 2 mode interaction by Armbruster and Chossat [1], [12] and generalized by Chossat and Guyard [25], [14]. We show that this cycle is embedded in a larger class of cycles, which we call a generalized heteroclinic cycle (GHC). After describing the structure of this set, we discuss its stability. Then the persistence under symmetry-breaking perturbations is investigated. We will discuss also the application to the spherical Bénard problem, which was the initial motivation for this work. Received March 11, 1997; first revision received October 10, 1997; second revision received April 13, 1998; accepted July 16, 1998     

Reconstruction of Wideband Reflectivity Densities by Wavelet Transforms

This paper is concerned with reconstruction problems arising in the context of radar signal analysis. The goal in radar is to obtain information about objects by emitting certain signals and analyzing the reflected echoes. In this paper, we shall focus on the general wideband model for radar echoes and on the case of continuously distributed objects D (reflectivity density). In this case, the echo is given by an inverse wavelet transform of the density D where the role of the analyzing wavelet is played by the transmitted signal. However, the null space of an inverse wavelet transform is nontrivial, it is described by the corresponding reproducing kernel. Following the approach of Naparst [14] and Rebolla-Neira et al. [16], we suggest to treat this problem by transmitting not just one signal but a family of signals. Indeed, a reconstruction formula for one- and 2-dimensional reflectivity densities can be derived, provided that the set of outgoing signals forms an orthogonal basis or – more general – a frame. We also present some rigorous error estimates for these reconstruction formulas. The theoretical results are confirmed by some numerical examples.     

On weakly (τ, m)-continuous functions

In this paper we introduce a new notion of weakly (τ, m)-continuous functions as functions from a topological space into a set satisfying some minimal conditions. We obtain some characterizations and several properties of such functions. This function leads to the formulation of a unified theory of weak continuity [20], almosts-continuity [33],p(θ)-continuity [10] andp-continuity [41].     

A generalization of results of P. Erdös,G. Katona,and D. J. Kleitman concerning Sperner's theorem

A result of Sperner [1] determining the maximal number of subsets of a given set, such that no one is included in the other, has been generalized and strengthened modifying the restrictive condition by Erdös [2], by Katona [3], and Kleitman [7], and by others [4, 5]. A different kind of generalization has been obtained by de Bruijn, Tenbergen, and Kruywijk [6] using the same restrictive condition but for more general entities than sets, namely, for systems in which the elements may occur more than once. That approach is fundamental for number-theoretical considerations since the totality of prime divisors of a given number (each considered with its multiplicity) is of that nature. In this paper we will generalize the results of [2] and [3, 7] in the sense of [6].