is represented by the elements lying within the envelope of its lower triangular part, that is, between the first nonzero of each row and the diagonal (see Example for an example). The width

nrow (i)

of the

i

th row is the number of elements between the first nonzero element and the element on the diagonal, inclusive. Although, of course, any matrix possesses an envelope as defined, this function is primarily intended for the factorization of symmetric positive definite matrices with an average bandwidth which is small compared with

n

(also see Further Comments).

The method is based on the property that during Cholesky factorization there is no fill-in outside the envelope.

The determination of

L

and

D

is normally the first of two steps in the solution of the system of equations

A x = b

. The remaining step, viz. the solution of

L D L^{T} x = b

, may be carried out using nag_linsys_real_posdef_vband_solve (f04mc).

References

Jennings A (1966) A compact storage scheme for the solution of symmetric linear simultaneous equations Comput. J. 9 281–285

Wilkinson J H and Reinsch C (1971) Handbook for Automatic Computation II, Linear Algebra Springer–Verlag

Parameters

Compulsory Input Parameters

1: $a (lal)$ – double array

The elements within the envelope of the lower triangle of the positive definite symmetric matrix

A

, taken in row by row order. The following code assigns the matrix elements within the envelope to the correct elements of the array:

 k = 0;
 for i = 1:n
   for j = i-nrow(i)+1:i
     k = k + 1;
     a(k) = matrix(i,j);
   end
 end

Optional Input Parameters

1: $n$ – int64int32nag_int scalar: Default: the dimension of the array nrow.
$n$ , the order of the matrix $A$ .

Constraint: $n \geq 1$ .
2: $lal$ – int64int32nag_int scalar: Default: the dimension of the array a.
The dimension of the arrays a and al.

Constraint: $lal \geq nrow (1) + nrow (2) + \dots + nrow (n)$ .

Output Parameters

1: $al (lal)$ – double array: The elements within the envelope of the lower triangular matrix $L$ , taken in row by row order. The envelope of $L$ is identical to that of the lower triangle of $A$ . The unit diagonal elements of $L$ are stored explicitly. See also Further Comments.
2: $d (n)$ – double array: The diagonal elements of the diagonal matrix $D$ . Note that the determinant of $A$ is equal to the product of these diagonal elements. If the value of the determinant is required it should not be determined by forming the product explicitly, because of the possibility of overflow or underflow. The logarithm of the determinant may safely be formed from the sum of the logarithms of the diagonal elements.
3: $ifail$ – int64int32nag_int scalar: $ifail = 0$ unless the function detects an error (see Error Indicators and Warnings).

Error Indicators and Warnings

Errors or warnings detected by the function:

Cases prefixed with W are classified as warnings and do not generate an error of type NAG:error_n. See nag_issue_warnings.

$ifail = 1$

On entry,	$n < 1$ ,
or	for some $i$ , $nrow (i) < 1$ or $nrow (i) > i$ ,
or	$lal < nrow (1) + nrow (2) + \dots + nrow (n)$ .

$ifail = 2$: $A$ is not positive definite, or this property has been destroyed by rounding errors. The factorization has not been completed.

W $ifail = 3$: $A$ is not positive definite, or this property has been destroyed by rounding errors. The factorization has not been completed.

$ifail = - 99$: An unexpected error has been triggered by this routine. Please contact NAG.

$ifail = - 399$: Your licence key may have expired or may not have been installed correctly.

$ifail = - 999$: Dynamic memory allocation failed.

Accuracy

ifail = 0

on exit, then the computed

L

and

D

satisfy the relation

L D L^{T} = A + F

, where

{‖F‖}_{2} \leq k m^{2} ε \times \max_{i} a_{i i}

and

{‖F‖}_{2} \leq k m^{2} ε \times {‖A‖}_{2},

where

k

is a constant of order unity,

m

is the largest value of

nrow (i)

, and

ε

is the machine precision. See pages 25–27 and 54–55 of Wilkinson and Reinsch (1971).

Further Comments

The time taken by nag_matop_real_vband_posdef_fac (f01mc) is approximately proportional to the sum of squares of the values of

nrow (i)

The distribution of row widths may be very non-uniform without undue loss of efficiency. Moreover, the function has been designed to be as competitive as possible in speed with functions designed for full or uniformly banded matrices, when applied to such matrices.

Example

This example obtains the Cholesky factorization of the symmetric matrix, whose lower triangle is:

(\begin{array}{r} 1 \\ 2 & 5 \\ 0 & 3 & 13 \\ 0 & 0 & 0 & 16 \\ 5 & 14 & 18 & 8 & 55 \\ 0 & 0 & 0 & 24 & 17 & 77 \end{array}) .

For this matrix, the elements of nrow must be set to

1

2

2

1

5

3

, and the elements within the envelope must be supplied in row order as:

1, 2, 5, 3, 13, 16, 5, 14, 18, 8, 55, 24, 17, 77 .

Open in the MATLAB editor: f01mc_example

function f01mc_example


fprintf('f01mc example results\n\n');

a = [1;
     2;     5;
            3;     13;
                           16;
     5;    14;     18;      8;     55;
                   24;     17;     77];
nrow = [int64(1);  2; 2; 1; 5; 3];

% Factorize
[L, D, ifail] = f01mc(a, nrow);

% Display
fprintf('    D                    L\n');
k = 0;
for i = 1:6
  fprintf(' %6.3f    ',D(i));
  for j = 1:i-nrow(i)
    fprintf('%8s',' ');
  end
  fprintf('%8.3f',L(k+1:k+nrow(i)));
  fprintf('\n');
  k = k + nrow(i);
end

f01mc example results

    D                    L
  1.000       1.000
  1.000       2.000   1.000
  4.000               3.000   1.000
 16.000                               1.000
  1.000       5.000   4.000   1.500   0.500   1.000
 16.000                               1.500   5.000   1.000

PDF version (NAG web site, 64-bit version, 64-bit version)

Chapter Contents

Chapter Introduction

NAG Toolbox