> On 20/09/2016, at 19:34, Jean-Pierre Flori <jpfl...@gmail.com> wrote:
> 
> 
> 
> On Tuesday, September 20, 2016 at 5:39:47 AM UTC+2, Steven Trogdon wrote:
> On Gentoo 1 failure:
> 
> sage -t --long src/sage/matrix/matrix_double_dense.pyx
> **********************************************************************
> File "src/sage/matrix/matrix_double_dense.pyx", line 668, in 
> sage.matrix.matrix_double_dense.Matrix_double_dense.condition
> Failed example:
>     A.condition()  # tol 1e-4
> Expected:
>     16332197709146.014
> Got:
>     16333848415071.576
> Tolerance exceeded:
>     16332197709146.014 vs 16333848415071.576, tolerance 1e-04 > 1e-04
> **********************************************************************
> 
> Now usually the tol spec if fairly close. But this way off.
> Maybe due to the switch from ATLAS to OpenBLAS as default BLAS.
> But I'm not sure upgrading makes the switch happens as configure has maybe 
> kept the choice of ATLAS cached.

Possible, but in that case may be hardware dependent. I have been
using openblas in sage-on-gentoo for several years and all the tolerance
problems that I had, have been fixed.

François

-- 
You received this message because you are subscribed to the Google Groups 
"sage-release" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-release+unsubscr...@googlegroups.com.
To post to this group, send email to sage-release@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-release.
For more options, visit https://groups.google.com/d/optout.

Reply via email to