hide long namesshow long names
hide short namesshow short names
Integer type:  int32  int64  nag_int  show int32  show int32  show int64  show int64  show nag_int  show nag_int

PDF version (NAG web site, 64-bit version, 64-bit version)
Chapter Contents
Chapter Introduction
NAG Toolbox

NAG Toolbox: nag_lapack_zcposv (f07fq)

Purpose

nag_lapack_zcposv (f07fq) uses the Cholesky factorization
A = UHU   or   A = LLH
A=UHU   or   A=LLH
to compute the solution to a complex system of linear equations
AX = B ,
AX=B ,
where AA is an nn by nn Hermitian positive definite matrix and XX and BB are nn by rr matrices.

Syntax

[a, x, iter, info] = f07fq(uplo, a, b, 'n', n, 'nrhs_p', nrhs_p)
[a, x, iter, info] = nag_lapack_zcposv(uplo, a, b, 'n', n, 'nrhs_p', nrhs_p)

Description

nag_lapack_zcposv (f07fq) first attempts to factorize the matrix in reduced precision and use this factorization within an iterative refinement procedure to produce a solution with full precision normwise backward error quality (see below). If the approach fails the method switches to a full precision factorization and solve.
The iterative refinement can be more efficient than the corresponding direct full precision algorithm. Since the strategy implemented by nag_lapack_zcposv (f07fq) must perform iterative refinement on each right-hand side, any efficiency gains will reduce as the number of right-hand sides increases. Conversely, as the matrix size increases the cost of these iterative refinements become less significant relative to the cost of factorization. Thus, any efficiency gains will be greatest for a very small number of right-hand sides and for large matrix sizes. The cut-off values for the number of right-hand sides and matrix size, for which the iterative refinement strategy performs better, depends on the relative performance of the reduced and full precision factorization and back-substitution. nag_lapack_zcposv (f07fq) always attempts the iterative refinement strategy first; you are advised to compare the performance of nag_lapack_zcposv (f07fq) with that of its full precision counterpart nag_lapack_zposv (f07fn) to determine whether this strategy is worthwhile for your particular problem dimensions.
The iterative refinement process is stopped if iter > 30iter>30 where iter is the number of iterations carried out thus far. The process is also stopped if for all right-hand sides we have
resid < sqrt(n) x A ε ,
resid < n x A ε ,
where residresid is the -norm of the residual, xx is the -norm of the solution, AA is the -norm of the matrix AA and εε is the machine precision returned by nag_machine_precision (x02aj).

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 http://www.netlib.org/lapack/lug
Golub G H and Van Loan C F (1996) Matrix Computations (3rd Edition) Johns Hopkins University Press, Baltimore
Higham N J (2002) Accuracy and Stability of Numerical Algorithms (2nd Edition) SIAM, Philadelphia

Parameters

Compulsory Input Parameters

1:     uplo – string (length ≥ 1)
Specifies whether the upper or lower triangular part of AA is stored.
uplo = 'U'uplo='U'
The upper triangular part of AA is stored.
uplo = 'L'uplo='L'
The lower triangular part of AA is stored.
Constraint: uplo = 'U'uplo='U' or 'L''L'.
2:     a(lda, : :) – complex array
The first dimension of the array a must be at least max (1,n)max(1,n)
The second dimension of the array must be at least max (1,n)max(1,n)
The nn by nn Hermitian positive definite matrix AA.
  • If uplo = 'U'uplo='U', the upper triangular part of aa must be stored and the elements of the array below the diagonal are not referenced.
  • If uplo = 'L'uplo='L', the lower triangular part of aa must be stored and the elements of the array above the diagonal are not referenced.
3:     b(ldb, : :) – complex array
The first dimension of the array b must be at least max (1,n)max(1,n)
The second dimension of the array must be at least max (1,nrhs)max(1,nrhs)
The right-hand side matrix BB.

Optional Input Parameters

1:     n – int64int32nag_int scalar
Default: The first dimension of the array n.
nn, the number of linear equations, i.e., the order of the matrix AA.
Constraint: n0n0.
2:     nrhs_p – int64int32nag_int scalar
Default: The second dimension of the array b.
rr, the number of right-hand sides, i.e., the number of columns of the matrix BB.
Constraint: nrhs0nrhs0.

Input Parameters Omitted from the MATLAB Interface

lda ldb ldx work swork rwork

Output Parameters

1:     a(lda, : :) – complex array
The first dimension of the array a will be max (1,n)max(1,n)
The second dimension of the array will be max (1,n)max(1,n)
ldamax (1,n)ldamax(1,n).
If iterative refinement has been successfully used (INFO = 0INFO=0 and iter0iter0, see description below), then a is unchanged. If full precision factorization has been used (INFO = 0INFO=0 and iter < 0iter<0, see description below), then the array AA contains the factor UU or LL from the Cholesky factorization A = UHUA=UHU or A = LLHA=LLH.
2:     x(ldx, : :) – complex array
The first dimension of the array x will be max (1,n)max(1,n)
The second dimension of the array will be max (1,nrhs)max(1,nrhs)
ldxmax (1,n)ldxmax(1,n).
If INFO = 0INFO=0, the nn by rr solution matrix XX.
3:     iter – int64int32nag_int scalar
Information on the progress of the interative refinement process.
iter < 0iter<0
Iterative refinement has failed for one of the reasons given below, full precision factorization has been performed instead.
1-1 The function fell back to full precision for implementation- or machine-specific reasons.
2-2 Narrowing the precision induced an overflow, the function fell back to full precision.
3-3 An intermediate reduced precision factorization failed.
31-31 The maximum permitted number of iterations was exceeded.
iter > 0iter>0
Iterative refinement has been sucessfully used. iter returns the number of iterations.
4:     info – int64int32nag_int scalar
info = 0info=0 unless the function detects an error (see Section [Error Indicators and Warnings]).

Error Indicators and Warnings

  INFO > 0andINFONINFO>0andINFON
The leading minor of order __ of AA is not positive definite, so the factorization could not be completed, and the solution has not been computed.

Accuracy

For each right-hand side vector bb, the computed solution xx is the exact solution of a perturbed system of equations (A + E)x = b(A+E)x=b, where c(n)c(n) is a modest linear function of nn, and εε is the machine precision. See Section 10.1 of Higham (2002) for further details.
An approximate error bound for the computed solution is given by
(x1)/(x1) κ(A) (E1)/(A1)
x^ - x 1 x 1 κ(A) E 1 A 1
where κ(A) = A11 A1 κ(A) = A-1 1 A 1 , the condition number of A A  with respect to the solution of the linear equations. See Section 4.4 of Anderson et al. (1999) for further details.

Further Comments

The real analogue of this function is nag_lapack_dsposv (f07fc).

Example

function nag_lapack_zcposv_example
a = [3.23,        1.51-1.92i,  1.90+0.84i,  0.42+2.50i;
     1.51+1.92i,  3.58,       -0.23+1.11i, -1.18+1.37i;
     1.90-0.84i, -0.23-1.11i,  4.09,        2.33-0.14i;
     0.42-2.50i, -1.18-1.37i,  2.33+0.14i,  4.29];
b = [3.93-06.14i; 6.17+09.42i; -7.17-21.83i; 1.99-14.38i];
% Solve the equations Ax = b for x
[a, x, iter, info] = nag_lapack_zcposv('u', a, b);

if info == 0
  % Print Solution
  fprintf('\nSolution:\n');
  disp(x);
else
  fprintf('\nThe leading minor of order %d is not positive definite.\n', info);
end
 

Solution:
   1.0000 - 1.0000i
  -0.0000 + 3.0000i
  -4.0000 - 5.0000i
   2.0000 + 1.0000i


function f07fq_example
a = [3.23,        1.51-1.92i,  1.90+0.84i,  0.42+2.50i;
     1.51+1.92i,  3.58,       -0.23+1.11i, -1.18+1.37i;
     1.90-0.84i, -0.23-1.11i,  4.09,        2.33-0.14i;
     0.42-2.50i, -1.18-1.37i,  2.33+0.14i,  4.29];
b = [3.93-06.14i; 6.17+09.42i; -7.17-21.83i; 1.99-14.38i];
% Solve the equations Ax = b for x
[a, x, iter, info] = f07fq('u', a, b);

if info == 0
  % Print Solution
  fprintf('\nSolution:\n');
  disp(x);
else
  fprintf('\nThe leading minor of order %d is not positive definite.\n', info);
end
 

Solution:
   1.0000 - 1.0000i
  -0.0000 + 3.0000i
  -4.0000 - 5.0000i
   2.0000 + 1.0000i



PDF version (NAG web site, 64-bit version, 64-bit version)
Chapter Contents
Chapter Introduction
NAG Toolbox

© The Numerical Algorithms Group Ltd, Oxford, UK. 2009–2013