NAG FL Interface
e02gbf (glinc_​l1sol)

1 Purpose

e02gbf calculates an l1 solution to an over-determined system of linear equations, possibly subject to linear inequality constraints.

2 Specification

Fortran Interface
Subroutine e02gbf ( m, n, mpl, e, lde, f, x, mxs, monit, iprint, k, el1n, indx, w, iw, ifail)
Integer, Intent (In) :: m, n, mpl, lde, mxs, iprint, iw
Integer, Intent (Inout) :: ifail
Integer, Intent (Out) :: k, indx(mpl)
Real (Kind=nag_wp), Intent (In) :: f(mpl)
Real (Kind=nag_wp), Intent (Inout) :: e(lde,mpl), x(n)
Real (Kind=nag_wp), Intent (Out) :: el1n, w(iw)
External :: monit
C Header Interface
#include <nag.h>
void  e02gbf_ (const Integer *m, const Integer *n, const Integer *mpl, double e[], const Integer *lde, const double f[], double x[], const Integer *mxs,
void (NAG_CALL *monit)(const Integer *n, const double x[], const Integer *niter, const Integer *k, const double *el1n),
const Integer *iprint, Integer *k, double *el1n, Integer indx[], double w[], const Integer *iw, Integer *ifail)
The routine may be called by the names e02gbf or nagf_fit_glinc_l1sol.

3 Description

Given a matrix A with m rows and n columns mn and a vector b with m elements, the routine calculates an l1 solution to the over-determined system of equations
Ax=b.  
That is to say, it calculates a vector x, with n elements, which minimizes the l1-norm (the sum of the absolute values) of the residuals
rx=i=1mri,  
where the residuals ri are given by
ri=bi-j=1naijxj,  i=1,2,,m.  
Here aij is the element in row i and column j of A, bi is the ith element of b and xj the jth element of x.
If, in addition, a matrix C with l rows and n columns and a vector d with l elements, are given, the vector x computed by the routine is such as to minimize the l1-norm rx subject to the set of inequality constraints Cxd.
The matrices A and C need not be of full rank.
Typically in applications to data fitting, data consisting of m points with coordinates ti,yi is to be approximated by a linear combination of known functions ϕit,
α1ϕ1t+α2ϕ2t++αnϕnt,  
in the l1-norm, possibly subject to linear inequality constraints on the coefficients αj of the form Cαd where α is the vector of the αj and C and d are as in the previous paragraph. This is equivalent to finding an l1 solution to the over-determined system of equations
j=1nϕjtiαj=yi,  i=1,2,,m,  
subject to Cαd.
Thus if, for each value of i and j, the element aij of the matrix A above is set equal to the value of ϕjti and bi is equal to yi and C and d are also supplied to the routine, the solution vector x will contain the required values of the αj. Note that the independent variable t above can, instead, be a vector of several independent variables (this includes the case where each of ϕi is a function of a different variable, or set of variables).
The algorithm follows the Conn–Pietrzykowski approach (see Bartels et al. (1978) and Conn and Pietrzykowski (1977)), which is via an exact penalty function
gx = γ rx - i=1 l min0, ciT x-di ,  
where γ is a penalty parameter, ciT is the ith row of the matrix C, and di is the ith element of the vector d. It proceeds in a step-by-step manner much like the simplex method for linear programming but does not move from vertex to vertex and does not require the problem to be cast in a form containing only non-negative unknowns. It uses stable procedures to update an orthogonal factorization of the current set of active equations and constraints.

4 References

Bartels R H, Conn A R and Charalambous C (1976) Minimisation techniques for piecewise Differentiable functions – the l solution to an overdetermined linear system Technical Report No. 247, CORR 76/30 Mathematical Sciences Department, The John Hopkins University
Bartels R H, Conn A R and Sinclair J W (1976) A Fortran program for solving overdetermined systems of linear equations in the l1 Sense Technical Report No. 236, CORR 76/7 Mathematical Sciences Department, The John Hopkins University
Bartels R H, Conn A R and Sinclair J W (1978) Minimisation techniques for piecewise differentiable functions – the l1 solution to an overdetermined linear system SIAM J. Numer. Anal. 15 224–241
Conn A R and Pietrzykowski T (1977) A penalty-function method converging directly to a constrained optimum SIAM J. Numer. Anal. 14 348–375

5 Arguments

1: m Integer Input
On entry: the number of equations in the over-determined system, m (i.e., the number of rows of the matrix A).
Constraint: mn.
2: n Integer Input
On entry: the number of unknowns, n (the number of columns of the matrix A).
Constraint: n2.
3: mpl Integer Input
On entry: m+l, where l is the number of constraints (which may be zero).
Constraint: mplm.
4: eldempl Real (Kind=nag_wp) array Input/Output
On entry: the equation and constraint matrices stored in the following manner.
The first m columns contain the m rows of the matrix A; element eij specifying the element aji in the jth row and ith column of A (the coefficient of the ith unknown in the jth equation), for i=1,2,,n and j=1,2,,m. The next l columns contain the l rows of the constraint matrix C; element eij+m containing the element cji in the jth row and ith column of C (the coefficient of the ith unknown in the jth constraint), for i=1,2,,n and j=1,2,,l.
On exit: unchanged, except possibly to the extent of a small multiple of the machine precision. (See Section 9.)
5: lde Integer Input
On entry: the first dimension of the array e as declared in the (sub)program from which e02gbf is called.
Constraint: lden.
6: fmpl Real (Kind=nag_wp) array Input
On entry: fi, for i=1,2,,m, must contain bi (the ith element of the right-hand side vector of the over-determined system of equations) and fm+i, for i=1,2,,l, must contain di (the ith element of the right-hand side vector of the constraints), where l is the number of constraints.
7: xn Real (Kind=nag_wp) array Input/Output
On entry: xi must contain an estimate of the ith unknown, for i=1,2,,n. If no better initial estimate for xi is available, set xi=0.0.
On exit: the latest estimate of the ith unknown, for i=1,2,,n. If ifail=0 on exit, these are the solution values.
8: mxs Integer Input
On entry: the maximum number of steps to be allowed for the solution of the unconstrained problem. Typically this may be a modest multiple of n. If, on entry, mxs is zero or negative, the value returned by x02bbf is used.
9: monit Subroutine, supplied by the user. External Procedure
monit can be used to print out the current values of any selection of its arguments. The frequency with which monit is called in e02gbf is controlled by iprint.
The specification of monit is:
Fortran Interface
Subroutine monit ( n, x, niter, k, el1n)
Integer, Intent (In) :: n, niter, k
Real (Kind=nag_wp), Intent (In) :: x(n), el1n
C Header Interface
void  monit_ (const Integer *n, const double x[], const Integer *niter, const Integer *k, const double *el1n)
1: n Integer Input
On entry: the number n of unknowns (the number of columns of the matrix A).
2: xn Real (Kind=nag_wp) array Input
On entry: the latest estimate of the unknowns.
3: niter Integer Input
On entry: the number of iterations so far carried out.
4: k Integer Input
On entry: the total number of equations and constraints which are currently active (i.e., the number of equations with zero residuals plus the number of constraints which are satisfied as equations).
5: el1n Real (Kind=nag_wp) Input
On entry: the l1-norm of the current residuals of the over-determined system of equations.
monit must either be a module subprogram USEd by, or declared as EXTERNAL in, the (sub)program from which e02gbf is called. Arguments denoted as Input must not be changed by this procedure.
10: iprint Integer Input
On entry: the frequency of iteration print out.
iprint>0
monit is called every iprint iterations and at the solution.
iprint=0
Information is printed out at the solution only. Otherwise monit is not called (but a dummy routine must still be provided).
11: k Integer Output
On exit: the total number of equations and constraints which are then active (i.e., the number of equations with zero residuals plus the number of constraints which are satisfied as equalities).
12: el1n Real (Kind=nag_wp) Output
On exit: the l1-norm (sum of absolute values) of the equation residuals.
13: indxmpl Integer array Output
On exit: specifies which columns of e relate to the inactive equations and constraints. indx1 up to indxk number the active columns and indxk+1 up to indxmpl number the inactive columns.
14: wiw Real (Kind=nag_wp) array Workspace
15: iw Integer Input
On entry: the dimension of the array w as declared in the (sub)program from which e02gbf is called.
Constraint: iw3×mpl+5×n+n2+n+1×n+2/2.
16: ifail Integer Input/Output
On entry: ifail must be set to 0, -1 or 1 to set behaviour on detection of an error; these values have no effect when no error is detected.
A value of 0 causes the printing of an error message and program execution will be halted; otherwise program execution continues. A value of -1 means that an error message is printed while a value of 1 means that it is not.
If halting is not appropriate, the value -1 or 1 is recommended. If message printing is undesirable, then the value 1 is recommended. Otherwise, the value 0 is recommended. When the value -1 or 1 is used it is essential to test the value of ifail on exit.
On exit: ifail=0 unless the routine detects an error or a warning has been flagged (see Section 6).

6 Error Indicators and Warnings

If on entry ifail=0 or -1, explanatory error messages are output on the current error message unit (as defined by x04aaf).
Errors or warnings detected by the routine:
ifail=1
The constraints cannot all be satisfied simultaneously: they are not compatible with one another. Hence no solution is possible.
ifail=2
The limit imposed by mxs has been reached without finding a solution. Consider restarting from the current point by simply calling e02gbf again without changing the arguments.
ifail=3
The routine has failed because of numerical difficulties; the problem is too ill-conditioned. Consider rescaling the unknowns.
ifail=4
Elements 1 to m of one of the first mpl columns of the array e are all zero – this corresponds to a zero row in either of the matrices A or C.
On entry, iw is too small. iw=value. Minimum possible dimension: value.
On entry, lde=value and n=value.
Constraint: lden.
On entry, m=value and n=value.
Constraint: mn.
On entry, mpl=value and m=value.
Constraint: mplm.
On entry, n=value.
Constraint: n2.
ifail=-99
An unexpected error has been triggered by this routine. Please contact NAG.
See Section 7 in the Introduction to the NAG Library FL Interface for further information.
ifail=-399
Your licence key may have expired or may not have been installed correctly.
See Section 8 in the Introduction to the NAG Library FL Interface for further information.
ifail=-999
Dynamic memory allocation failed.
See Section 9 in the Introduction to the NAG Library FL Interface for further information.

7 Accuracy

The method is stable.

8 Parallelism and Performance

e02gbf is threaded by NAG for parallel execution in multithreaded implementations of the NAG Library.
e02gbf 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 effect of m and n on the time and on the number of iterations varies from problem to problem, but typically the number of iterations is a small multiple of n and the total time taken is approximately proportional to mn2.
Linear dependencies among the rows or columns of A and C are not necessarily a problem to the algorithm. Solutions can be obtained from rank-deficient A and C. However, the algorithm requires that at every step the currently active columns of e form a linearly independent set. If this is not the case at any step, small, random perturbations of the order of rounding error are added to the appropriate columns of e. Normally this perturbation process will not affect the solution significantly. It does mean, however, that results may not be exactly reproducible.

10 Example

Suppose we wish to approximate in 0,1 a set of data by a curve of the form
y=ax3+bx2+cx+d  
which has non-negative slope at the data points. Given points ti,yi we may form the equations
yi=ati3+bti2+cti+d  
for i=1,2,,6, for the 6 data points. The requirement of a non-negative slope at the data points demands
3ati2+2bti+c0  
for each ti and these form the constraints.
(Note that, for fitting with polynomials, it would usually be advisable to work with the polynomial expressed in Chebyshev series form (see the E02 Chapter Introduction). The power series form is used here for simplicity of exposition.)

10.1 Program Text

Program Text (e02gbfe.f90)

10.2 Program Data

Program Data (e02gbfe.d)

10.3 Program Results

Program Results (e02gbfe.r)