ON TWO CLASSES OF TEST MATRICES.

Abstract

Two classes of matrices which are useful in testing numerical algorithms for problems in linear algebra are presented. Their determinants, eigenvalues, eigenvectors and inverses are deduced in analytic forms. These are easily generated by machine or by hand as functions of certain parameters in the original matrices. By appropriately selecting the parameters one is able to generate ill-conditioned matrices without having to deal with high order matrices. (Author)

Document Details

Document Type
Technical Report
Publication Date
Jul 01, 1970
Accession Number
AD0712679

Entities

People

  • Nai-kuan Tsao

Organizations

  • University of HawaiĘ»i System

Tags

DTIC Thesaurus Topics

  • Algebra
  • Algorithms
  • Eigenvalues
  • Eigenvectors
  • Linear Algebra
  • Mathematical Analysis
  • Mathematics
  • Matrices (Mathematics)
  • Vector Spaces

Fields of Study

  • Mathematics

Readers

  • Linear Algebra
  • Software Engineering