NAG Library Routine Document

f08zsf  (zggqrf)

 Contents

    1  Purpose
    7  Accuracy

1
Purpose

f08zsf (zggqrf) computes a generalized QR factorization of a complex matrix pair A,B, where A is an n by m matrix and B is an n by p matrix.

2
Specification

Fortran Interface
Subroutine f08zsf ( n, m, p, a, lda, taua, b, ldb, taub, work, lwork, info)
Integer, Intent (In):: n, m, p, lda, ldb, lwork
Integer, Intent (Out):: info
Complex (Kind=nag_wp), Intent (Inout):: a(lda,*), b(ldb,*)
Complex (Kind=nag_wp), Intent (Out):: taua(min(n,m)), taub(min(n,p)), work(max(1,lwork))
C Header Interface
#include nagmk26.h
void  f08zsf_ ( const Integer *n, const Integer *m, const Integer *p, Complex a[], const Integer *lda, Complex taua[], Complex b[], const Integer *ldb, Complex taub[], Complex work[], const Integer *lwork, Integer *info)
The routine may be called by its LAPACK name zggqrf.

3
Description

f08zsf (zggqrf) forms the generalized QR factorization of an n by m matrix A and an n by p matrix B 
A =QR ,   B=QTZ ,  
where Q is an n by n unitary matrix, Z is a p by p unitary matrix and R and T are of the form
R = mmR11n-m0() ,   if ​nm; nm-nnR11R12() ,   if ​n<m,  
with R11 upper triangular,
T = p-nnn0T12() ,   if ​np, pn-pT11pT21() ,   if ​n>p,  
with T12 or T21 upper triangular.
In particular, if B is square and nonsingular, the generalized QR factorization of A and B implicitly gives the QR factorization of B-1A as
B-1A= ZH T-1 R .  

4
References

Anderson E, Bai Z, Bischof C, Blackford S, Demmel J, Dongarra J J, Du Croz J J, Greenbaum A, Hammarling S, McKenney A and Sorensen D (1999) LAPACK Users' Guide (3rd Edition) SIAM, Philadelphia http://www.netlib.org/lapack/lug
Anderson E, Bai Z and Dongarra J (1992) Generalized QR factorization and its applications Linear Algebra Appl. (Volume 162–164) 243–271
Hammarling S (1987) The numerical solution of the general Gauss-Markov linear model Mathematics in Signal Processing (eds T S Durrani, J B Abbiss, J E Hudson, R N Madan, J G McWhirter and T A Moore) 441–456 Oxford University Press
Paige C C (1990) Some aspects of generalized QR factorizations . In Reliable Numerical Computation (eds M G Cox and S Hammarling) 73–91 Oxford University Press

5
Arguments

1:     n – IntegerInput
On entry: n, the number of rows of the matrices A and B.
Constraint: n0.
2:     m – IntegerInput
On entry: m, the number of columns of the matrix A.
Constraint: m0.
3:     p – IntegerInput
On entry: p, the number of columns of the matrix B.
Constraint: p0.
4:     alda* – Complex (Kind=nag_wp) arrayInput/Output
Note: the second dimension of the array a must be at least max1,m.
On entry: the n by m matrix A.
On exit: the elements on and above the diagonal of the array contain the minn,m by m upper trapezoidal matrix R (R is upper triangular if nm); the elements below the diagonal, with the array taua, represent the unitary matrix Q as a product of minn,m elementary reflectors (see Section 3.3.6 in the F08 Chapter Introduction).
5:     lda – IntegerInput
On entry: the first dimension of the array a as declared in the (sub)program from which f08zsf (zggqrf) is called.
Constraint: ldamax1,n.
6:     tauaminn,m – Complex (Kind=nag_wp) arrayOutput
On exit: the scalar factors of the elementary reflectors which represent the unitary matrix Q.
7:     bldb* – Complex (Kind=nag_wp) arrayInput/Output
Note: the second dimension of the array b must be at least max1,p.
On entry: the n by p matrix B.
On exit: if np, the upper triangle of the subarray b1:np-n+1:p  contains the n by n upper triangular matrix T12.
If n>p, the elements on and above the n-pth subdiagonal contain the n by p upper trapezoidal matrix T; the remaining elements, with the array taub, represent the unitary matrix Z as a product of elementary reflectors (see Section 3.3.6 in the F08 Chapter Introduction).
8:     ldb – IntegerInput
On entry: the first dimension of the array b as declared in the (sub)program from which f08zsf (zggqrf) is called.
Constraint: ldbmax1,n.
9:     taubminn,p – Complex (Kind=nag_wp) arrayOutput
On exit: the scalar factors of the elementary reflectors which represent the unitary matrix Z.
10:   workmax1,lwork – Complex (Kind=nag_wp) arrayWorkspace
On exit: if info=0, the real part of work1 contains the minimum value of lwork required for optimal performance.
11:   lwork – IntegerInput
On entry: the dimension of the array work as declared in the (sub)program from which f08zsf (zggqrf) is called.
If lwork=-1, a workspace query is assumed; the routine only calculates the optimal size of the work array, returns this value as the first entry of the work array, and no error message related to lwork is issued.
Suggested value: for optimal performance, lworkmaxn,m,p×maxnb1,nb2,nb3, where nb1 is the optimal block size for the QR factorization of an n by m matrix, nb2 is the optimal block size for the RQ factorization of an n by p matrix, and nb3 is the optimal block size for a call of f08auf (zunmqr).
Constraint: lworkmax1,n,m,p or lwork=-1.
12:   info – IntegerOutput
On exit: info=0 unless the routine detects an error (see Section 6).

6
Error Indicators and Warnings

info<0
If info=-i, argument i had an illegal value. An explanatory message is output, and execution of the program is terminated.

7
Accuracy

The computed generalized QR factorization is the exact factorization for nearby matrices A+E and B+F, where
E2 = Oε A2   and   F2= Oε B2 ,  
and ε is the machine precision.

8
Parallelism and Performance

f08zsf (zggqrf) is threaded by NAG for parallel execution in multithreaded implementations of the NAG Library.
f08zsf (zggqrf) makes calls to BLAS and/or LAPACK routines, which may be threaded within the vendor library used by this implementation. Consult the documentation for the vendor library for further information.
Please consult the X06 Chapter Introduction for information on how to control and interrogate the OpenMP environment used within this routine. Please also consult the Users' Note for your implementation for any additional implementation-specific information.

9
Further Comments

The unitary matrices Q and Z may be formed explicitly by calls to f08atf (zungqr) and f08cwf (zungrq) respectively. f08auf (zunmqr) may be used to multiply Q by another matrix and f08cxf (zunmrq) may be used to multiply Z by another matrix.
The real analogue of this routine is f08zef (dggqrf).

10
Example

This example solves the general Gauss–Markov linear model problem
minx y2   subject to   d=Ax+By  
where
A = 0.96-0.81i -0.03+0.96i -0.91+2.06i -0.98+1.98i -1.20+0.19i -0.66+0.42i 0.62-0.46i 1.01+0.02i 0.63-0.17i 1.08-0.28i 0.20-0.12i -0.07+1.23i ,  
B = 0.5-1.0i 0.0i+0.0 0.0i+0.0 0.0i+0.0 0.0i+0.0 1.0-2.0i 0.0i+0.0 0.0i+0.0 0.0i+0.0 0.0i+0.0 2.0-3.0i 0.0i+0.0 0.0i+0.0 0.0i+0.0 0.0i+0.0 5.0-4.0i  
and
d = 6.00-0.40i -5.27+0.90i 2.72-2.13i -1.30-2.80i .  
The solution is obtained by first computing a generalized QR factorization of the matrix pair A,B. The example illustrates the general solution process, although the above data corresponds to a simple weighted least squares problem.
Note that the block size (NB) of 64 assumed in this example is not realistic for such a small problem, but should be suitable for large problems.

10.1
Program Text

Program Text (f08zsfe.f90)

10.2
Program Data

Program Data (f08zsfe.d)

10.3
Program Results

Program Results (f08zsfe.r)

© The Numerical Algorithms Group Ltd, Oxford, UK. 2017