Volume 28 (2007-2008)


(Incomplete)


Dedicated to Gene Golub on the occasion of his 75th birthday


Welcome to a special volume of ETNA in honor of the 75th birthday of Gene Golub, which fell at the stroke of midnight between February 28 and March 1, 2007.

Everyone in the field of matrix computations has been affected by Gene, who has been famous since the 1960s for his contributions to many subjects. During his prolific career he has worked on least-squares and total least-squares, the singular value decomposition, quadrature, fast Poisson solvers, matrix iterations, preconditioners, factorizations, moment problems, Google-type matrices, and numerous other topics. Gene's contributions to the profession of numerical analysis and applied mathematics have been equally outsized, including being one of the founders of NA-Net, NA Digest, the quadrennial ICIAM Congresses, the Fox Prize, and two highly successful SIAM journals. He has about 30 PhD students and 80 academic grandchildren. And then there is the book Matrix Computations by Golub and Van Loan, with its fourth edition in preparation, which has defined the scope of modern numerical linear algebra. A citation index lists about 16,000 citations to this extraordinary book.

Once you start looking at his citations, it is hard to stop. Besides the book, we find that Gene has around 20 other papers with 100 or more citations, and 100 papers with 20 or more citations! His next hundred papers after that each have between 6 and 20 citations; and on and on it goes. Even counting coauthors is not easy with Gene, but the number seems to be approximately 230, ranging from Alter, Anderssen, Andersson, Arbenz, and Atkins to Zenios, Zha, and Zhang!

Gene has touched us all, whether personally or professionally, and we are happy to commemorate this occasion with a special set of papers.

Martin Gutknecht, Michael Overton, Lothar Reichel, Daniel B. Szyld, Nick Trefethen, Paul Van Dooren, and Andy Wathen.
Special Volume editors

Addendum. Gene Golub died on November 16, 2007. At that time this volume had already six papers published electronically. Papers published afterwards have a different dedication. Possibly no one else has contributed more to numerical linear algebra than Gene Golub. The mathematics community has lost an extraordinary scientist and we have lost a good friend.


1-15 M. R. Osborne
Separable least squares, variable projection, and the Gauss-Newton algorithm
Abstract and links - Postscript document [2843 KBytes] - PDF document [258 KBytes]

16-39 Michele Benzi and Nader Razouk
Decay bounds and O(n) algorithms for approximating functions of sparse matrices
Abstract and links - Postscript document [12574 KBytes] - PDF document [1549 KBytes]

40-64 Melina A. Freitag and Alastair Spence
Convergence theory for inexact inverse iteration applied to the generalised nonsymmetric eigenproblem
Abstract and links - Postscript document [721 KBytes] - PDF document [342 KBytes]

65-77 S. Noschese, L. Pasquini, and L. Reichel
The structured distance to normality of an irreducible real tridiagonal matrix
Abstract and links - Postscript document [319 KBytes] - PDF document [157 KBytes]

78-94 A. Hadjidimos and P. Stratis
Minimization of the spectral norm of the SOR operator in a mixed case
Abstract and links - Postscript document [420 KBytes] - PDF document [248 KBytes]

95-113 Jean Christophe Tremblay and Tucker Carrington Jr.
A refined unsymmetric Lanczos eigensolver for computing accurate eigentriplets of a real unsymmetric matrix
Abstract and links - Postscript document [1559 KBytes] - PDF document [2225 KBytes]

114-135 James V. Lambers
Derivation of high-order spectral methods for time-dependent PDE using modified moments
Abstract and links - Postscript document [447 KBytes] - PDF document [252 KBytes]

136-148 Martin Ludwig, Jochim Koch, and Bernd Fischer
An application of the finite volume method to the bio-heat-transfer-equation in premature infants
Abstract and links - Postscript document [3880 KBytes] - PDF document [3555 KBytes]

149-167 Julianne Chung, James G. Nagy, and Dianne P. O'Leary
A weighted-GCV method for Lanczos-hybrid regularization
Abstract and links - Postscript document [590 KBytes] - PDF document [235 KBytes]

168-173 Dirk Laurie
Variable-precision arithmetic considered perilous - a detective story
Abstract and links - Postscript document [128 KBytes] - PDF document [66 KBytes]