1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
|
function [V,beta,p,R,q] = cs_qr (A) %#ok
%CS_QR sparse QR factorization (Householder-based).
% [V,beta,p,R] = cs_qr(A) computes the QR factorization of A(p,:).
% [V,beta,p,R,q] = cs_qr(A) computes the QR factorization of A(p,q).
% The V, beta, and p terms represent the Householder vectors and coefficients.
% The fill-reducing ordering q is found via q = cs_amd(A,3).
% The orthogonal factor Q can be obtained via
% Q = cs_qright(V,beta,p,speye(size(V,1))), in which case Q*R=A(:,q) is the
% resulting factorization (the permutation p is folded into Q). A must be
% m-by-n with m >= n. If A is structurally rank deficient, additional empty
% rows may have been added to V and R. Note that V is typically much sparser
% than Q.
%
% Example:
%
% Prob = ssget ('HB/well1033') ; A = Prob.A ; [m n] = size (A) ;
% b = rand (m,1) ;
% [V,beta,p,R,q] = cs_qr (A) ; % QR factorization of A(p,q)
% b1 = cs_qleft (V, beta, p, b) ;
% x1 = R (1:n,1:n) \ b1 (1:n) ;
% x1 (q) = x1 ;
% x2 = A\b ;
% norm (x1-x2)
% Q = cs_qright(V,beta,p,speye(size(V,1))) ; % Note: p accounted for in Q
% norm (Q*R-A(:,q),1)
% fprintf ('nnz(R) %d, nnz(V) %d, nnz(Q) %d\n', nnz(R), nnz(V), nnz(Q)) ;
%
% See also CS_AMD, CS_QRIGHT, CS_QR, CS_DMPERM, QR, COLAMD.
% CXSparse, Copyright (c) 2006-2022, Timothy A. Davis. All Rights Reserved.
% SPDX-License-Identifier: LGPL-2.1+
error ('cs_qr mexFunction not found') ;
|
