Gustavson F G, Waśniewski J, Dongarra J J and Langou J (2010) Rectangular full packed format for Cholesky's algorithm: factorization, solution, and inversion ACM Trans. Math. Software 37, 2

Parameters

Compulsory Input Parameters

1: $transr$ – string (length ≥ 1)

Specifies whether the normal RFP representation of

A

or its conjugate transpose is stored.

$transr ='N'$: The matrix $A$ is stored in normal RFP format.
$transr ='C'$: The conjugate transpose of the RFP representation of the matrix $A$ is stored.

Constraint:

transr ='N'

'C'

2: $uplo$ – string (length ≥ 1)

Specifies how

A

has been factorized.

$uplo ='U'$: $A = U^{H} U$ , where $U$ is upper triangular.
$uplo ='L'$: $A = L L^{H}$ , where $L$ is lower triangular.

Constraint:

uplo ='U'

'L'

3: $ar (n \times (n + 1) / 2)$ – complex array

The Cholesky factorization of

A

stored in RFP format, as returned by nag_lapack_zpftrf (f07wr).

4: $b (ldb :)$ – complex array

The first dimension of the array b must be at least

\max (1, n)

The second dimension of the array b must be at least

\max (1, nrhs_p)

The

n

r

right-hand side matrix

B

Optional Input Parameters

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

Constraint: $n \geq 0$ .
2: $nrhs_p$ – int64int32nag_int scalar: Default: the second dimension of the array b.
$r$ , the number of right-hand sides.

Constraint: $nrhs_p \geq 0$ .

Output Parameters

1: $b (ldb :)$ – complex array: The first dimension of the array b will be $\max (1, n)$ .
The second dimension of the array b will be $\max (1, nrhs_p)$ .

The $n$ by $r$ solution matrix $X$ .
2: $info$ – int64int32nag_int scalar: $info = 0$ unless the function detects an error (see Error Indicators and Warnings).

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.

Accuracy

For each right-hand side vector

b

, the computed solution

x

is the exact solution of a perturbed system of equations

(A + E) x = b

, where

if $uplo ='U'$ , $|E| \leq c (n) ε |U^{H}| |U|$ ;
if $uplo ='L'$ , $|E| \leq c (n) ε |L| |L^{H}|$ ,

c (n)

is a modest linear function of

n

, and

ε

is the machine precision

\hat{x}

is the true solution, then the computed solution

x

satisfies a forward error bound of the form

\frac{{‖x - \hat{x}‖}_{\infty}}{{‖x‖}_{\infty}} \leq c (n) cond (A, x) ε

where

cond (A, x) = {‖|A^{- 1}| |A| |x|‖}_{\infty} / {‖x‖}_{\infty} \leq cond (A) = {‖|A^{- 1}| |A|‖}_{\infty} \leq κ_{\infty} (A)

and

κ_{\infty} (A)

is the condition number when using the

\infty

-norm.

Note that

cond (A, x)

can be much smaller than

cond (A)

Further Comments

The total number of real floating-point operations is approximately

8 n^{2} r

The real analogue of this function is nag_lapack_dpftrs (f07we).

Example

This example solves the system of equations

A X = B

, where

A = (\begin{array}{r} 3.23 + 0.00 i & 1.51 - 1.92 i & 1.90 + 0.84 i & 0.42 + 2.50 i \\ 1.51 + 1.92 i & 3.58 + 0.00 i & - 0.23 + 1.11 i & - 1.18 + 1.37 i \\ 1.90 - 0.84 i & - 0.23 - 1.11 i & 4.09 + 0.00 i & 2.33 - 0.14 i \\ 0.42 - 2.50 i & - 1.18 - 1.37 i & 2.33 + 0.14 i & 4.29 + 0.00 i \end{array})

and

B = (\begin{array}{r} 3.93 - 6.14 i & 1.48 + 6.58 i \\ 6.17 + 9.42 i & 4.65 - 4.75 i \\ - 7.17 - 21.83 i & - 4.91 + 2.29 i \\ 1.99 - 14.38 i & 7.64 - 10.79 i \end{array}) .

Here

A

is Hermitian positive definite, stored in RFP format, and must first be factorized by nag_lapack_zpftrf (f07wr).

Open in the MATLAB editor: f07ws_example

function f07ws_example


fprintf('f07ws example results\n\n');

% Symmetric matrix in RFP format
transr = 'n';
uplo   = 'l';
ar = [ 4.09 + 0.00i  2.33 - 0.14i;
       3.23 + 0.00i  4.29 + 0.00i;
       1.51 + 1.92i  3.58 + 0.00i;
       1.90 - 0.84i -0.23 - 1.11i;
       0.42 - 2.50i -1.18 - 1.37i];
n  = int64(4);
n2 = (n*(n+1))/2;
ar  = reshape(ar,[n2,1]);

% RHS
b = [ 3.93 -  6.14i,  1.48 +  6.58i;
      6.17 +  9.42i,  4.65 -  4.75i;
     -7.17 - 21.83i, -4.91 +  2.29i;
      1.99 - 14.38i,  7.64 - 10.79i];

% Factorize a
[ar, info] = f07wr(transr, uplo, n, ar);

if info == 0
  % Compute solution
  [b, info] = f07ws( ...
                     transr, uplo, ar, b);
  fprintf('\n');
  ncols  = int64(80);
  indent = int64(0);
  form   = 'f7.4';
  title  = 'Solutions';

  [ifail] = x04db( ...
                   'g', ' ', b, 'bracket', form, title, ...
                   'int', 'int', ncols, indent);
else
  fprintf('\na is not positive definite.\n');
end

f07ws example results


 Solutions
                    1                 2
 1  ( 1.0000,-1.0000) (-1.0000, 2.0000)
 2  (-0.0000, 3.0000) ( 3.0000,-4.0000)
 3  (-4.0000,-5.0000) (-2.0000, 3.0000)
 4  ( 2.0000, 1.0000) ( 4.0000,-5.0000)

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

Chapter Contents

Chapter Introduction

NAG Toolbox