LPDU decomposition

LPDU decompostion.

Description

LPDU decomposition of non-singular square matrix

• MATLAB release: MATLAB 7 (R14)

[Suhas Taneja]

Excellent

[Stefan]

The function fails for "higher" dimension. E.g.

>> A=rand(4,4);
>> det(A)

ans =

    0.0149

>> [L,P,D,U]=lpdu(A)

Warning: Matrix is close to singular or badly scaled.
         Results may be inaccurate. RCOND = 1.421637e-033.
> In LPDU at 83

L =

  1.0e+016 *

    0.0000 0 0 0
   -1.2269 0.0000 0 -0.0000
   -1.2245 0.0000 0.0000 0.0000
    1.1595 0.0000 -0.0000 0.0000

P =

     1 1 0 0
     1 0 1 0
     1 0 0 1
     1 0 0 0

D =

  1.0e+014 *

    1.9878 0 0 0
         0 0.0000 0 0
         0 0 -0.0000 0
         0 0 0 0.0000

U =

    1.0000 3.1587 12.4369 11.8480
         0 1.0000 -4.0000 4.0000
         0 0 1.0000 0.2882
         0 0 0 1.0000

>> norm(L*P*D*U-A)

ans =

  7.2518e+031