Re: svds test failures

[Tatsuro MATSUOKA]

Hello
For my case,

The patch
     1.7 -%! randn('state',42);      % Initialize to make normest function 
reproducible
     1.8 -%! rand('state',42)
     1.9 +%! randn ('state', 43);      % Initialize to make normest function 
reproducible
    1.10 +%! rand ('state', 43)
is origin of the fail.

Regards

Tatsuro

--- On Mon, 2011/8/8, Tatsuro MATSUOKA wrote:

> Hello
> 
> I have build the recent dev. source on MinGW.
> and two errors on svds.m are found.
> 
> For the first FAIL is perhaps similar to that pointed out by Marco.
> 
> To see the second fail, I have executed,
>  octave:6> format long
> octave:7>  s = svds (speye (10));
> octave:8> s
> s =
> 
>    1.00000000000000
>    1.00000000000000
>    1.00000000000000
>    1.00000000000000
>    1.00000000000000
>    1.00000000000000
> 
> octave:9> ones (6, 1)
> ans =
> 
>    1
>    1
>    1
>    1
>    1
>    1
> 
> octave:10> assert (s,ones (6, 1),2 * eps)
> octave:11> 
> 
> From the command line the above test does not give fail.   ????
> 
> Regards
> 
> Tatsuro
> 
>   ***** testif HAVE_UMFPACK
>  [u2,s2,v2,flag] = svds(A,k,0,opts);
>  s2 = diag(s2);
>  assert(flag,!1);
>  assert(s2, s(k:-1:1), 1e-10);
> !!!!! test failed
> assert (s2,s (k:-1:1),1e-10) expected
>    38.060
>    38.034
>    38.034
>    38.015
>    38.015
>    38.004
>    38.004
> but got
>    38.034
>    38.034
>    38.015
>    38.015
>    38.004
>    38.004
> Dimensions don't match
> 
> 
> 
> 
>   ***** test
>  s = svds (speye (10));
>  assert (s, ones (6, 1), 2*eps);
> !!!!! test failed
> assert (s,ones (6, 1),2 * eps) expected
>    1
>    1
>    1
>    1
>    1
>    1
> but got
>    1.0000
>    1.0000
>    1.0000
>    1.0000
>    1.0000
>    1.0000
> maximum absolute error 6.66134e-016 exceeds tolerance 4.44089e-016
> shared variables 
>   scalar structure containing the fields:
> 
>     n =  100
>     k =  7
>     A =
> 
> Compressed Column Sparse (rows = 100, cols = 100, nnz = 296 [3%])
> 
>   (1, 1) ->  40
> <snip>
> --- On Mon, 2011/8/8, Rik <address@hidden> wrote:
> 
> > On 08/07/2011 08:25 AM, Marco atzeri wrote:
> > > On 8/7/2011 4:28 PM, Rik wrote:
> > > 
> > >> I'm not opposed to changing the initialization seed and this is probably
> > >> the fastest way to get these tests working.  I tried moving to 43 and it
> > >> works on a Linux system.  In order to be certain you need to run the svds
> > >> code repeatedly.  Here is what I ran
> > >>
> > >>   for i = 1:1000
> > >>     bm(i) = test ("svds");
> > >>   endfor
> > >>   sum(bm)
> > >>
> > >> ans = 1000
> > >>
> > >> If you can repeat this test on your machine then I will just change the
> > >> initialization point.
> > >>
> > >> Regards,
> > >> Rik
> > > 
> > > on cygwin with 43
> > > 
> > > octave:1> for i = 1:1000
> > >>    bm(i) = test ("svds");
> > >>  endfor
> > > octave:2> sum(bm)
> > > ans =  1000
> > > 
> > > eventually we should ask a check also for mingw version.
> > > Just to be sure
> > 
> > I changed the initialization seed in this changeset
> > (http://hg.savannah.gnu.org/hgweb/octave/rev/5d79384e3a1d).  Eventually
> > when someone gets a working 3.4.2 build on mingw then they can verify that
> > 43 works there as well.
> > 
> > --Rik
> >
>