NAG FL Interface
e04rnf (handle_​set_​linmatineq)

1 Purpose

e04rnf is a part of the NAG optimization modelling suite and defines one or more linear matrix constraints of the problem.

2 Specification

Fortran Interface
Subroutine e04rnf ( handle, nvar, dima, nnza, nnzasum, irowa, icola, a, nblk, blksizea, idblk, ifail)
Integer, Intent (In) :: nvar, dima, nnza(nvar+1), nnzasum, irowa(nnzasum), icola(nnzasum), nblk, blksizea(nblk)
Integer, Intent (Inout) :: idblk, ifail
Real (Kind=nag_wp), Intent (In) :: a(nnzasum)
Type (c_ptr), Intent (In) :: handle
C Header Interface
#include <nag.h>
void  e04rnf_ (void **handle, const Integer *nvar, const Integer *dima, const Integer nnza[], const Integer *nnzasum, const Integer irowa[], const Integer icola[], const double a[], const Integer *nblk, const Integer blksizea[], Integer *idblk, Integer *ifail)
The routine may be called by the names e04rnf or nagf_opt_handle_set_linmatineq.

3 Description

After the initialization routine e04raf has been called, e04rnf may be used to add one or more linear matrix inequalities
i=1 n xi Ai - A0 0 (1)
to the problem definition. Here Ai are d by d symmetric matrices. The expression S0 stands for a constraint on eigenvalues of a symmetric matrix S, namely, all the eigenvalues should be non-negative, i.e., the matrix S should be positive semidefinite.
Typically, this will be used in linear semidefinite programming problems (SDP)
minimize xn cTx   (a) subject to   i=1 n xi Aik - A0k 0 ,  k=1,,mA   (b) lBBxuB   (c) lxxux   (d) (2)
or to define the linear part of bilinear matrix inequalities (3)(b) in (BMI-SDP)
minimize xn 12 xTHx + cTx   (a) subject to   i,j=1 n xi xj Qijk + i=1 n xi Aik - A0k 0 ,  k=1,,mA   (b) lBBxuB   (c) lxxux   (d) (3)
e04rnf can be called repeatedly to accumulate more matrix inequalities. See Section 3.1 in the E04 Chapter Introduction for more details about the NAG optimization modelling suite.

3.1 Input data organization

All the matrices Ai, for i=0,1,,n, are symmetric and thus only their upper triangles are passed to the routine. They are stored in sparse coordinate storage format (see Section 2.1.1 in the F11 Chapter Introduction), i.e., every nonzero from the upper triangles is coded as a triplet of row index, column index and the numeric value. These triplets of all (upper triangle) nonzeros from all Ai matrices are passed to the routine in three arrays: irowa for row indices, icola for column indices and a for the values. No particular order of nonzeros within one matrix is enforced but all nonzeros from A0 must be stored first, followed by all nonzero from A1, followed by A2, etc.
The number of stored nonzeros from each Ai matrix is given in nnzai+1, thus this array indicates which section of arrays irowa, icola and a belongs to which Ai matrix. See Table 1 and the example in Section 9. See also e04rdf which uses the same data organization.
irowa upper triangle upper triangle   upper triangle
icola nonzeros nonzeros nonzeros
a from ​ A0 from ​ A1   from ​ An
  nnza1 nnza2   nnzan+1
Table 1
Coordinate storage format of matrices A0, A1,, An in input arrays
There are two possibilities for defining more matrix inequality constraints
i=1 n xi A i k - A 0 k 0 ,  k=1,2,,mA (4)
to the problem. The first is to call e04rnf mA times and define a single matrix inequality at a time. This might be more straightforward and therefore it is recommended. Alternatively, it is possible to merge all mA constraints into one inequality and pass them in a single call to e04rnf. It is easy to see that (4) can be equivalently expressed as one bigger matrix inequality with the following block diagonal structure
i=1 n xi Ai1 Ai2 AimA - A01 A02 A0mA 0 .  
If dk denotes the dimension of inequality k, the new merged inequality has dimension d= k=1 mA dk and each of the Ai matrices is formed by A i 1 , A i 2 ,, A i mA stored as mA diagonal blocks. In such a case, nblk is set to mA and blksizeak to dk, the size of the kth diagonal blocks. This might be useful in connection with e04rdf.
On the other hand, if there is no block structure and just one matrix inequality is provided, nblk should be set to 1 and blksizea is not referenced.

3.2 Definition of Bilinear Matrix Inequalities (BMI)

e04rnf is designed to be used together with e04rpf to define bilinear matrix inequalities (3)(b). e04rnf sets the linear part of the constraint and e04rpf expands it by higher order terms. To distinquish which linear matrix inequality (or more precisely, which block) is to be expanded, e04rpf needs the number of the block, idblk. The blocks are numbered as they are added, starting from 1.
Whenever a matrix inequality (or a set of them expressed as diagonal blocks) is stored, the routine returns idblk of the last inequality added. idblk is just the order of the inequality amongst all matrix inequalities accumulated through the calls. The first inequality has idblk=1, the second one idblk=2, etc. Therefore if you call e04rnf for the very first time with nblk=42, it adds 42 inequalities with idblk from 1 to 42 and the routine returns idblk=42 (the number of the last one). A subsequent call with nblk=1 would add only one inequality, this time with idblk=43 which would be returned.

4 References

None.

5 Arguments

1: handle Type (c_ptr) Input
On entry: the handle to the problem. It needs to be initialized by e04raf and must not be changed before the call to e04rnf.
2: nvar Integer Input
On entry: n, the number of decision variables x in the problem. It must be unchanged from the value set during the initialization of the handle by e04raf.
3: dima Integer Input
On entry: d, the dimension of the matrices Ai, for i=0,1,,nvar.
Constraint: dima>0.
4: nnzanvar+1 Integer array Input
On entry: nnzai+1, for i=0,1,,nvar, gives the number of nonzero elements in the upper triangle of matrix Ai. To define Ai as a zero matrix, set nnzai+1=0. However, there must be at least one matrix with at least one nonzero.
Constraints:
  • nnzai0;
  • i=1 n+1 nnzai1.
5: nnzasum Integer Input
On entry: the dimension of the arrays irowa, icola and a, at least the total number of all nonzeros in all matrices Ai.
Constraints:
  • nnzasum>0;
  • i=1 n+1 nnzai nnzasum.
6: irowannzasum Integer array Input
7: icolannzasum Integer array Input
8: annzasum Real (Kind=nag_wp) array Input
On entry: nonzero elements in upper triangle of matrices Ai stored in coordinate storage. The first nnza1 elements belong to A0, the following nnza2 elements belong to A1, etc. See explanation above.
Constraints:
  • 1irowaidima, irowaiicolaidima;
  • irowa and icola match the block diagonal pattern set by blksizea.
9: nblk Integer Input
On entry: mA, number of diagonal blocks in Ai matrices. As explained above it is equivalent to the number of matrix inequalities supplied in this call.
Constraint: nblk1.
10: blksizeanblk Integer array Input
On entry: if nblk>1, sizes dk of the diagonal blocks.
If nblk=1, blksizea is not referenced.
Constraints:
  • blksizeai1;
  • i=1 mA blksizeai=dima.
11: idblk Integer Input/Output
On entry: if idblk=0, new matrix inequalities are created. This is the only value allowed at the moment; nonzero values are reserved for future releases of the NAG Library.
Constraint: idblk=0.
On exit: the number of the last matrix inequality added. By definition, it is the number of the matrix inequalities already defined plus nblk.
12: 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 -1 is recommended since useful values can be provided in some output arguments even when ifail0 on exit. 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 supplied handle does not define a valid handle to the data structure for the NAG optimization modelling suite. It has not been initialized by e04raf or it has been corrupted.
ifail=2
The problem cannot be modified in this phase any more, the solver has already been called.
ifail=4
On entry, idblk=value.
Constraint: idblk=0.
On entry, nvar=value, expected value=value.
Constraint: nvar must match the value given during initialization of handle.
ifail=6
On entry, dima=value.
Constraint: dima>0.
On entry, i=value and nnzai=value.
Constraint: nnzai0.
On entry, nnzasum=value and sumnnza=value.
Constraint: nnzasumsumnnza.
On entry, sumnnza=value.
Constraint: sumnnza1.
ifail=7
On entry, dima=value and sumblksizea=value.
Constraint: sumblksizea=dima.
On entry, i=value and blksizeai=value.
Constraint: blksizeai1.
On entry, nblk=value.
Constraint: nblk>0.
ifail=8
An error occurred in matrix Ai, i=value (counting indices 1nvar+1).
On entry, j=value, icolaj=value and dima=value.
Constraint: 1icolajdima.
An error occurred in matrix Ai, i=value (counting indices 1nvar+1).
On entry, j=value, irowaj=value and dima=value.
Constraint: 1irowajdima.
An error occurred in matrix Ai, i=value (counting indices 1nvar+1).
On entry, j=value, irowaj=value and icolaj=value.
Constraint: irowajicolaj (elements within the upper triangle).
An error occurred in matrix Ai, i=value (counting indices 1nvar+1).
On entry, j=value, irowaj=value and icolaj=value. Maximum column index in this row given by the block structure defined by blksizea is value.
Constraint: all elements of Ai must respect the block structure given by blksizea.
An error occurred in matrix Ai, i=value (counting indices 1nvar+1).
On entry, more than one element of Ai has row index value and column index value.
Constraint: each element of Ai must have a unique row and column index.
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

Not applicable.

8 Parallelism and Performance

e04rnf is not threaded in any implementation.

9 Further Comments

The following example demonstrates how the elements of the A i k matrices are organized within the input arrays. Let us assume that there are two blocks defined (nblk=2). The first has dimension 3 by 3 (blksizea1=3) and the second 2 by 2 (blksizea2=2). For simplicity, the number of variables is 2. Please note that the values were chosen to ease orientation rather than to define a valid problem.
A 0 1 = 0.1 0 0.3 0 0.2 0.4 0.3 0.4 0 , A 1 1 ​ empty ​ A 2 1 = 2.1 0 0 0 2.2 0 0 0 2.3 , A 0 2 = 0 -0.1 -0.1 0 , A 1 2 = -1.1 0 0 -1.2 , A 2 2 = -2.1 -2.2 -2.2 -2.3 .  
Both inequalities will be passed in a single call to e04rnf, therefore the matrices are merged into the following block diagonal form:
A0 = 0.1 0 0.3 0 0.2 0.4 0.3 0.4 0 0 -0.1 -0.1 0 ,  
A1 = 0 0 0 0 0 0 0 0 0 -1.1 0 0 -1.2 ,  
A2 = 2.1 0 0 0 2.2 0 0 0 2.3 -2.1 -2.2 -2.2 -2.3 .  
All matrices are symmetric and therefore only the upper triangles are passed to the routine. The coordinate storage format is used. Note that elements within the same matrix do not need to be in any specific order. The table below shows one of the ways the arrays could be populated.
irowa 0.2 0.2 -0.4 0.1 0.1 -0.4 -0.5 0.1 0.2 0.3 -0.4 -0.4 -0.5
icola 0.2 0.3 -0.5 0.1 0.3 -0.4 -0.5 0.1 0.2 0.3 -0.4 -0.5 -0.5
a 0.2 0.4 -0.1 0.1 0.3 -1.1 -1.2 2.1 2.2 2.3 -2.1 -2.2 -2.3
  A0 A1 A2
nnza 5 2 6

10 Example

There are various problems which can be successfully reformulated and solved as an SDP problem. The following example shows how a maximization of the minimal eigenvalue of a matrix depending on certain parameters can be utilized in statistics.
For further examples, please refer to Section 10 in e04raf.
Given a series of M vectors of length p, vi : i=1,2,,M this example solves the SDP problem:
maximize λ1,,λM,t t subject to   i=1 M λi vi viT tI i=1 M λi =1 λi0 ,  k=1,,M .  
This formulation comes from an area of statistics called experimental design and corresponds to finding an approximate E optimal design for a linear regression.
A linear regression model has the form:
y=Xβ+ε  
where y is a vector of observed values, X is a design matrix of (known) independent variables and ε is a vector of errors. In experimental design it is assumed that each row of X is chosen from a set of M possible vectors, vi : i=1,2,,M . The goal of experimental design is to choose the rows of X so that the error covariance is ‘small’. For an E optimal design this is defined as the X that maximizes the minimum eigenvalue of XTX.
In this example we construct the E optimal design for a polynomial regression model of the form:
y = β0+ β1 x+ β2 x2+ β3 x3+ β4 x4+ε  
where x 1-j×0.05 : j=0,1,,40 .

10.1 Program Text

Program Text (e04rnfe.f90)

10.2 Program Data

Program Data (e04rnfe.d)

10.3 Program Results

Program Results (e04rnfe.r)