Hello:

Sorry for the top posting but I'm answering directly from the mailing
list and I can not add the pevious mails in the thread with the
context marks >>, so I've copied and pasted below the last post

>From macports you can install the last 1.15 release, but the question
is: has passed the macports gsl package all tests?

On the other hand in the output from svd test can be seen that the
error occurs due to in SVD decomposition the maximun iteration number
was reached. This maximun is fixed in line 147 of svd.c as 100*N. I
don't know in detail the algorithm but I don't know the reason for
this 100*N. Can check someone if the test is passed in mac os x if the
limit is set, for example,  to 200*N?
In svd.c says that the file was written by Gerard Jungman and Brian
Gough. Maybe some of them could examine the problem

Cheers

**********************************************

From:   Sam Mason
Subject:        Re: [Help-gsl] GSL-1.15 installation issue on OSX 10.7.3
Date:   Thu, 5 Apr 2012 10:07:23 +0100

Hi,

If you just want a copy of GSL installed and don't mind about it being
too up to date, you could try MacPorts.  Initial install from here:

  http://www.macports.org/

and then type

  sudo port install gsl-devel

I can't help with building it from source, sorry!  I'd be tempted to
suggest that it's your C compiler's optimiser breaking something, but
that's very wild speculation.


  Sam

On 4 April 2012 21:24, Pierre-Henry Frohring
<address@hidden> wrote:
> Hello,
>
> I tried to install the GSLibrary following the INSTALL directives:
>        $ ./configure
>        $ make
>        $ make check > log 2>&1
>
> It gave me some errors on the svd module. The log file is > 6mb, so i just
> cat the "useful" lines:
>
> Log
> -------------------------------------------------------------------------
> Making check in blas
> make[1]: Nothing to be done for `check'.
> Making check in linalg
> make  test
> make[2]: `test' is up to date.
> make  check-TESTS
> (caught [svd.c:149: SVD decomposition failed to converge (11)])
> returned error code 11 = exceeded max number of iterations
> singular value 0 =                     -1 < 0
> singular value 1 =                      0 vs previous                     -1
> singular value 2 =    17.3205080756888137 vs previous                      0
> (  3,  3)[0,2]:    15.2508026013158453                        0
> (  3,  3)[1,2]:    8.21054322293596961                        1
> (  3,  3)[2,1]:                      0                        1
> FAIL:   SV_decomp (3x3) A=[ 0, 0, 0; 0, 0, 1; 1, 1, 0] [852]
> (caught [svd.c:149: SVD decomposition failed to converge (11)])
> returned error code 11 = exceeded max number of iterations
> singular value 0 =                     -1 < 0
> singular value 1 =                      0 vs previous                     -1
> singular value 2 =    17.3205080756888137 vs previous                      0
> (  3,  3)[0,1]:    15.2508026013158453                        0
> (  3,  3)[1,1]:    8.21054322293596961                        1
> (  3,  3)[2,2]:                      0                        1
> FAIL:   SV_decomp (3x3) A=[ 0, 0, 0; 0, 1, 0; 1, 0, 1] [859]
> (  3,  3)[0,1]:  -0.499999999999999889                        0
> (  3,  3)[1,2]:   -1.00000000000000022                        0
> (  3,  3)[2,1]:   0.499999999999999889                        1
> -------------------------------------------------------------------------
>
> <Stuff>
>
> Log
> -------------------------------------------------------------------------
> FAIL:   SV_decomp_mod (4x4) A=[ 1, 1, 1, 1; 1, 1, 1, 1; 0, 1, 0, 1; 0, 0, 1,
> 1] [233613]
> FAIL: test
> ==================
> 1 of 1 test failed
> ==================
> make[2]: *** [check-TESTS] Error 1
> make[1]: *** [check-am] Error 2
> make: *** [check-recursive] Error 1
> -------------------------------------------------------------------------
>
> I searched is the archive with: "svd make check". The searched shrunk to 4
> results, including this old post (2011 Aug)
> http://lists.gnu.org/archive/html/help-gsl/2011-08/msg00032.html
>
> But no answers…
>
> Anyone has an idea? Thanks for your time and help!
> PH

-- 
*****************************************
José Luis García Pallero
jgpall...@gmail.com
(o<
/ / \
V_/_
Use Debian GNU/Linux and enjoy!
*****************************************

Reply via email to