F08ZPF (ZGGGLM) (PDF version)
F08 Chapter Contents
F08 Chapter Introduction
NAG Library Manual

NAG Library Routine Document

F08ZPF (ZGGGLM)

Note:  before using this routine, please read the Users' Note for your implementation to check the interpretation of bold italicised terms and other implementation-dependent details.

+ Contents

    1  Purpose
    7  Accuracy

1  Purpose

F08ZPF (ZGGGLM) solves a complex general Gauss–Markov linear (least squares) model problem.

2  Specification

SUBROUTINE F08ZPF ( M, N, P, A, LDA, B, LDB, D, X, Y, WORK, LWORK, INFO)
INTEGER  M, N, P, LDA, LDB, LWORK, INFO
COMPLEX (KIND=nag_wp)  A(LDA,*), B(LDB,*), D(M), X(N), Y(P), WORK(max(1,LWORK))
The routine may be called by its LAPACK name zggglm.

3  Description

F08ZPF (ZGGGLM) solves the complex general Gauss–Markov linear model (GLM) problem
minimize x y2  subject to  d=Ax+By
where A is an m by n matrix, B is an m by p matrix and d is an m element vector. It is assumed that nmn+p, rankA=n and rankE=m, where E= A B . Under these assumptions, the problem has a unique solution x and a minimal 2-norm solution y, which is obtained using a generalized QR factorization of the matrices A and B.
In particular, if the matrix B is square and nonsingular, then the GLM problem is equivalent to the weighted linear least squares problem
minimize x B-1 d-Ax 2 .

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
Anderson E, Bai Z and Dongarra J (1992) Generalized QR factorization and its applications Linear Algebra Appl. (Volume 162–164) 243–271

5  Parameters

1:     M – INTEGERInput
On entry: m, the number of rows of the matrices A and B.
Constraint: M0.
2:     N – INTEGERInput
On entry: n, the number of columns of the matrix A.
Constraint: 0NM.
3:     P – INTEGERInput
On entry: p, the number of columns of the matrix B.
Constraint: PM-N.
4:     A(LDA,*) – COMPLEX (KIND=nag_wp) arrayInput/Output
Note: the second dimension of the array A must be at least max1,N.
On entry: the m by n matrix A.
On exit: A is overwritten.
5:     LDA – INTEGERInput
On entry: the first dimension of the array A as declared in the (sub)program from which F08ZPF (ZGGGLM) is called.
Constraint: LDAmax1,M.
6:     B(LDB,*) – COMPLEX (KIND=nag_wp) arrayInput/Output
Note: the second dimension of the array B must be at least max1,P.
On entry: the m by p matrix B.
On exit: B is overwritten.
7:     LDB – INTEGERInput
On entry: the first dimension of the array B as declared in the (sub)program from which F08ZPF (ZGGGLM) is called.
Constraint: LDBmax1,M.
8:     D(M) – COMPLEX (KIND=nag_wp) arrayInput/Output
On entry: the left-hand side vector d of the GLM equation.
On exit: D is overwritten.
9:     X(N) – COMPLEX (KIND=nag_wp) arrayOutput
On exit: the solution vector x of the GLM problem.
10:   Y(P) – COMPLEX (KIND=nag_wp) arrayOutput
On exit: the solution vector y of the GLM problem.
11:   WORK(max1,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.
12:   LWORK – INTEGERInput
On entry: the dimension of the array WORK as declared in the (sub)program from which F08ZPF (ZGGGLM) 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, LWORKN+minM,P+maxM,P×nb, where nb is the optimal block size.
Constraint: LWORK max1,M+N+P  or LWORK=-1.
13:   INFO – INTEGEROutput
On exit: INFO=0 unless the routine detects an error (see Section 6).

6  Error Indicators and Warnings

Errors or warnings detected by the routine:
INFO<0
If INFO=-i, argument i had an illegal value. An explanatory message is output, and execution of the program is terminated.
INFO=1
The upper triangular factor R associated with A in the generalized RQ factorization of the pair A,B is singular, so that rankA<m; the least squares solution could not be computed.
INFO=2
The bottom N-M by N-M part of the upper trapezoidal factor T associated with B in the generalized QR factorization of the pair A,B is singular, so that rankAB<N; the least squares solutions could not be computed.

7  Accuracy

For an error analysis, see Anderson et al. (1992). See also Section 4.6 of Anderson et al. (1999).

8  Further Comments

When p=mn, the total number of real floating point operations is approximately 832m3-n3+16nm2; when p=m=n, the total number of real floating point operations is approximately 563m3.

9  Example

This example solves the weighted least squares problem
minimize x B-1 d-Ax 2 ,
where
B = 0.5-1.0i .0i+0.0 .0i+0.0 .0i+0.0 .0i+0.0 1.0-2.0i .0i+0.0 .0i+0.0 .0i+0.0 .0i+0.0 2.0-3.0i .0i+0.0 .0i+0.0 .0i+0.0 .0i+0.0 5.0-4.0i ,
d = 6.00-0.40i -5.27+0.90i 2.72-2.13i -1.30-2.80i
and
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 .
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.

9.1  Program Text

Program Text (f08zpfe.f90)

9.2  Program Data

Program Data (f08zpfe.d)

9.3  Program Results

Program Results (f08zpfe.r)


F08ZPF (ZGGGLM) (PDF version)
F08 Chapter Contents
F08 Chapter Introduction
NAG Library Manual

© The Numerical Algorithms Group Ltd, Oxford, UK. 2012