The open archive for STFC research publications

ePubs will be undergoing scheduled maintenance on 19/06/2024 between 10:00-12:00 and the service will be unavailable during this time.

Full Record Details

Persistent URL http://purl.org/net/epubs/work/29535
Record Status Checked
Record Id 29535
Title Incremental norm estimation for dense and sparse matrices
Abstract We discuss the benefits of an incremental approach to norm estimation for triangular matrices. Our investigation covers both dense and sparse matrices. If we use our incremental norm estimation on explicitly generated entries of the inverse of the triangular matrix, we can relate our results to incremental condition estimation (ICE). We show that our algorithm extends more naturally to the sparse case than ICE. Our scheme can be used to develop a robust pivot selection criterion for QR factorization or as the basis of a rank-revealing factorization.
Organisation CCLRC
Funding Information
Related Research Object(s): 33215 , 33213
Licence Information:
Language English (EN)
Type Details URI(s) Local file(s) Year
Report RAL Technical Reports RAL-TR-2001-005. 2001. raltr-2001005.pdf 2001