Error Analysis of Symplectic Lanczos Method for Hamiltonian Eigenvalue Problem
Received:January 20, 2002  
Key Words: Symplectic Lanczos method   Hamiltonian matrix   eigenvalues   error analysis   Ritz values   Ritz vectors.  
Fund Project:
Author NameAffiliation
YAN Qing-you Dept. of Economics and Statistics
Shandong Finance Institute
Ji'nan
China
School of Business Administration
North China Electric University
Beijing
Chinas 
WEI Xiao-peng Center of Advanced Design Technology
Dalian University
Liaoning
China 
Hits: 2916
Download times: 1128
Abstract:
      A rounding error analysis of the symplectic Lanczos method is given for the Hamil-tonian eigenvalue problem. It is applicable when no break down occurs and shows that the restriction of preserving the Hamiltonian structure does not destroy the characteristic feature of nonsymmetric Lanczos processes. An analog of Paige's theory on the relationship between the loss of orthogonality among the Lanczos vectors and the convergence of Ritz values in the symmetric Lanczos algorithm is discussed. All analysis follows the lines of Bai's analysis of the nonsymmetric Lanczos algorithm and the lines of H.FaBbender's analysis of the symplectic Lanczos algorithm for the symplectic eigenvalue problem. As is expected, it follows that (under certain assumptions) the computed J-orthogonal Lanczos vectors loose J-orthogonality when some Ritz values begin to converge.
Citation:
DOI:10.3770/j.issn:1000-341X.2004.01.017
View Full Text  View/Add Comment