LU FACTORIZATION WITH PANEL RANK REVEALING PIVOTING AND ITS COMMUNICATION AVOIDING VERSION

Khabou, Amal and Demmel, James W. and Grigori, Laura and Gu, Ming (2013) LU FACTORIZATION WITH PANEL RANK REVEALING PIVOTING AND ITS COMMUNICATION AVOIDING VERSION. SIAM Journal on Matrix Analysis and Applications, Volume (Issue ). pp. 1401-1429. ISSN 0895-4798

This is the latest version of this item.

[img] PDF
120863691.pdf

Download (551kB)
Official URL: http://epubs.siam.org/doi/abs/10.1137/120863691

Abstract

We present block LU factorization with panel rank revealing pivoting (block LU_PRRP), a decomposition algorithm based on strong rank revealing QR panel factorization. Block LU_PRRP is more stable than Gaussian elimination with partial pivoting (GEPP), with a theoretical upper bound of the growth factor of $(1+ \tau b)^{(n/ b)-1}$, where $b$ is the size of the panel used during the block factorization, $\tau$ is a parameter of the strong rank revealing QR factorization, $n$ is the number of columns of the matrix, and for simplicity we assume that n is a multiple of b. We also assume throughout the paper that $2\leq b \ll n$. For example, if the size of the panel is $b = 64$, and $\tau = 2$, then $(1+2b)^{(n/b)-1} = (1.079)^{n-64} \ll 2^{n-1}$, where $2^{n-1}$ is the upper bound of the growth factor of GEPP. Our extensive numerical experiments show that the new factorization scheme is as numerically stable as GEPP in practice, but it is more resistant to pathological cases. The block LU_PRRP factorization does only $O(n^2 b)$ additional floating point operations compared to GEPP. We also present block CALU_PRRP, a version of block LU_PRRP that minimizes communication and is based on tournament pivoting, with the selection of the pivots at each step of the tournament being performed via strong rank revealing QR factorization. Block CALU_PRRP is more stable than CALU, the communication avoiding version of GEPP, with a theoretical upper bound of the growth factor of $(1+ \tau b)^{{n\over b}(H+1)-1}$, where $H$ is the height of the reduction tree used during tournament pivoting. The upper bound of the growth factor of CALU is $2^{n(H+1)-1}$. Block CALU_PRRP is also more stable in practice and is resistant to pathological cases on which GEPP and CALU fail.

Item Type: Article
Uncontrolled Keywords: LU factorization, growth factor, numerical stability, communication avoiding, strong rank revealing QR factorization
Subjects: MSC 2010, the AMS's Mathematics Subject Classification > 65 Numerical analysis
Depositing User: Amal Khabou
Date Deposited: 28 Oct 2013
Last Modified: 20 Oct 2017 14:13
URI: http://eprints.maths.manchester.ac.uk/id/eprint/2033

Available Versions of this Item

Actions (login required)

View Item View Item