Re: QR updating changeset

[Jaroslav Hajek]

On Wed, Mar 5, 2008 at 3:52 AM, John W. Eaton <address@hidden> wrote:
>
> Finally, I'm seeing this error now (even without my changes, so I
> don't think it is a problem that I introduced):
>
>   ====>>>>> processing /tmp/jwe/octave/src/DLD-FUNCTIONS/qrinsert.cc
>    ***** test
>   A = [0.620405 + 0.956953i  0.480013 + 0.048806i  0.402627 + 0.338171i;
>        0.589077 + 0.658457i  0.013205 + 0.279323i  0.229284 + 0.721929i;
>        0.092758 + 0.345687i  0.928679 + 0.241052i  0.764536 + 0.832406i;
>        0.912098 + 0.721024i  0.049018 + 0.269452i  0.730029 + 0.796517i;
>        0.112849 + 0.603871i  0.486352 + 0.142337i  0.355646 + 0.151496i ];
>
>   x = [0.20351 + 0.05401i;
>        0.13141 + 0.43708i;
>        0.29808 + 0.08789i;
>        0.69821 + 0.38844i;
>        0.74871 + 0.25821i ];
>
>   [Q,R] = qr(A);
>   [Q,R] = qrinsert(Q,R,3,x);
>   assert(norm(vec(Q'*Q - eye(5)),Inf) < 1e1*eps)
>   assert(norm(vec(triu(R)-R),Inf) == 0)
>   assert(norm(vec(Q*R - [A(:,1:2) x A(:,3)]),Inf) < norm(A)*1e1*eps)
>
>  !!!!! test failed
>  error: assert (norm (vec (Q * R - [A(:, 1:2), x, A(:, 3)]), Inf) < norm (A) 
> * 1e1 * eps) failed
>
> Your patch and my changes are all in my public archive now, so will
> you please update and take a look?
>
> Thanks,
>
> jwe
>
>
>

John,
I can't reproduce this failure, all tests are passing when I compile.
Would you please check again with the newest changesets? (Submitted in
the more recent thread).
In case it persists, please run the attached script (the failing test
but more verbosely) so that I can see whether this is a numerical
issue. The difference 1e1*eps is an ad hoc value though it seems more
than enough on my machine.

regards,


-- 
RNDr. Jaroslav Hajek
computing expert
Aeronautical Research and Test Institute (VZLU)
Prague, Czech Republic
url: www.highegg.matfyz.cz

test_qrinsert_failed.m
Description: Text document