Homotopic residual correction processes

We present and analyze homotopic (continuation) residual correction algorithms for the computation of matrix inverses. For complex indefinite Hermitian input matrices, our homotopic methods substantially accelerate the known nonhomotopic algorithms. Unlike the nonhomotopic case our algorithms require no pre-estimation of the smallest singular value of an input matrix. Furthermore, we guarantee rapid convergence to the inverses of well-conditioned structured matrices even where no good initial approximation is available. In particular we yield the inverse of a well-conditioned matrix with a structure of Toeplitz/Hankel type in flops. For a large class of input matrices, our methods can be extended to computing numerically the generalized inverses. Our numerical experiments confirm the validity of our analysis and the efficiency of the presented algorithms for well-conditioned input matrices and furnished us with the proper values of the parameters that define our algorithms.

     

Inversion of circulant matrices over

In this paper we consider the problem of inverting an circulant matrix with entries over . We show that the algorithm for inverting circulants, based on the reduction to diagonal form by means of FFT, has some drawbacks when working over . We present three different algorithms which do not use this approach. Our algorithms require different degrees of knowledge of and , and their costs range, roughly, from to operations over . Moreover, for each algorithm we give the cost in terms of bit operations. We also present an algorithm for the inversion of finitely generated bi-infinite Toeplitz matrices. The problems considered in this paper have applications to the theory of linear cellular automata.

     

Linear functionals of eigenvalues of random matrices

Let be a random unitary matrix with distribution given by Haar measure on the unitary group. Using explicit moment calculations, a general criterion is given for linear combinations of traces of powers of to converge to a Gaussian limit as . By Fourier analysis, this result leads to central limit theorems for the measure on the circle that places a unit mass at each of the eigenvalues of . For example, the integral of this measure against a function with suitably decaying Fourier coefficients converges to a Gaussian limit without any normalisation. Known central limit theorems for the number of eigenvalues in a circular arc and the logarithm of the characteristic polynomial of are also derived from the criterion. Similar results are sketched for Haar distributed orthogonal and symplectic matrices.

     

Decay of positive waves for hyperbolic systems of balance laws

We prove Olenik-type decay estimates for entropy solutions of strictly hyperbolic systems of balance laws built out of a wave-front tracking procedure inside which the source term is treated as a nonconservative product localized on a discrete lattice.

     

Involutions and characters of upper triangular matrix groups

We study the realizability over of representations of the group of upper-triangular matrices over . We prove that all the representations of are realizable over if , but that if , has representations not realizable over . This theorem is a variation on a result that can be obtained by combining work of J. Arregi and A. Vera-López and of the authors, but the proof of the theorem in this paper is much more natural.

     

Matrices over orders in algebraic number fields as sums of -th powers

David R. Richman proved that for every integral matrix is a sum of seven -th powers. In this paper, in light of a question proposed earlier by M. Newman for the ring of integers of an algebraic number field, we obtain a discriminant criterion for every matrix over an order of an algebraic number field to be a sum of (seven) -th powers.

     

A noncommutative moment problem

We prove a noncommutative moment theorem and relate it to Connes' problem of embedding finite factor von Neumann algebras into an ultraproduct of the hyperfinite factor. We include a linear-algebraic equivalent of Connes' problem, which asks for a characterization of all noncommutative polynomials which have positive trace when the variables are replaced by contractive hermitian matrices.

     

Sequences of consecutive smooth polynomials over a finite field

Given , we show that there are infinitely many sequences of consecutive -smooth polynomials over a finite field. The number of polynomials in each sequence is approximately .

     

Polynomial detection of matrix subalgebras

The double Capelli polynomial of total degree is

It was proved by Giambruno-Sehgal and Chang that the double Capelli polynomial of total degree is a polynomial identity for . (Here, is a field and is the algebra of matrices over .) Using a strengthened version of this result obtained by Domokos, we show that the double Capelli polynomial of total degree is a polynomial identity for any proper -subalgebra of . Subsequently, we present a similar result for nonsplit inequivalent extensions of full matrix algebras.

     

Chern numbers of almost complex manifolds

It is shown that any system of numbers that can be realised as the system of Chern numbers of an almost complex manifold of dimension , , can also be realised in this way by a connected almost complex manifold. This answers an old question posed by Hirzebruch.

     

Cube-approximating bounded wavelet sets in

We prove that for any real expansive matrix , there exists a bounded -dilation wavelet set in the frequency domain (the inverse Fourier transform of whose characteristic function is a band-limited single wavelet in the time domain ). Moreover these wavelet sets can approximate a cube in arbitrarily. This result improves Dai, Larson and Speegle's result about the existence of (basically unbounded) wavelet sets for real expansive matrices.

     

On searching for solutions of the Diophantine equation

We propose an efficient search algorithm to solve the equation for a fixed value of 0$">. By parametrizing , this algorithm obtains and (if they exist) by solving a quadratic equation derived from divisors of . Thanks to the use of several efficient number-theoretic sieves, the new algorithm is much faster on average than previous straightforward algorithms. We performed a computer search for six values of below 1000 for which no solution had previously been found. We found three new integer solutions for and 931 in the range of .

     

A proof of Weinberg's conjecture on lattice-ordered matrix algebras

Let be a subfield of the field of real numbers and let () be the matrix algebra over . It is shown that if is a lattice-ordered algebra over in which the identity matrix 1 is positive, then is isomorphic to the lattice-ordered algebra with the usual lattice order. In particular, Weinberg's conjecture is true.

     

Maps on the -dimensional subspaces of a Hilbert space preserving principal angles

In a former paper we studied transformations on the set of all -dimensional subspaces of a Hilbert space which preserve the principal angles. In the case when , we could determine the general form of all such maps. The aim of this paper is to complete our result by considering the problem in the remaining case .

     

Local automorphisms and derivations on

The aim of this note is to give a short proof that 2-local derivations on , the matrix algebra over the complex numbers are derivations and to give a shorter proof that 2-local *-automorphisms on are *-automorphisms.

     

BV estimates of Lax-Friedrichs' scheme for a class of nonlinear hyperbolic conservation laws

We give uniform BV estimates and -stability of Lax-Friedrichs' scheme for a class of systems of strictly hyperbolic conservation laws whose integral curves of the eigenvector fields are straight lines, i.e., Temple class, under the assumption of small total variation. This implies that the approximate solutions generated via the Lax-Friedrichs' scheme converge to the solution given by the method of vanishing viscosity or the Godunov scheme, and then the Glimm scheme or the wave front tracking method.

     

On the local structure of rank-one convex hulls

In this note we prove that if is a compact set of matrices containing an isolated point with no rank-one connection into the convex hull of , then the rank-one convex hull separates as

This is an extension of a result of P. Pedregal, which holds for matrices.

     

On the maximal difference between an element and its inverse in residue rings

We investigate the distribution of where

Exponential sums provide a natural tool for obtaining upper bounds on this quantity. Here we use results about the distribution of integers with a divisor in a given interval to obtain lower bounds on . We also present some heuristic arguments showing that these lower bounds are probably tight, and thus our technique can be a more appropriate tool to study than a more traditional way using exponential sums.

     

Two new Weyl-type bounds for the Dirichlet Laplacian

In this paper, we prove two new Weyl-type upper estimates for the eigenvalues of the Dirichlet Laplacian. As a consequence, we obtain the following lower bounds for its counting function. For , one has

and

where

is a constant which depends on , the dimension of the underlying space, and Bessel functions and their zeros.