#21119: Fix numerical noise on Haswell (single doctest)
-------------------------------------+-------------------------------------
       Reporter:  leif               |         Type:  defect
         Status:  new                |     Priority:  trivial
      Milestone:  sage-7.3           |    Component:  porting
       Keywords:                     |    Merged in:
        Authors:  Leif Leonhardy     |    Reviewers:
Report Upstream:  N/A                |  Work issues:
         Branch:                     |       Commit:
  u/leif/haswell_noise               |  34a1ddc84850ace439d3fe4cec7923347ba69780
   Dependencies:                     |     Stopgaps:
-------------------------------------+-------------------------------------
 Fixes numerical noise in a single doctest in
 `sage/matrix/matrix_double_dense.pyx` on Intel Haswell.

 The doctest fails when doing a native build with any GCC version that
 supports Haswell:
 {{{
 **********************************************************************
 File "src/sage/matrix/matrix_double_dense.pyx", line 3761, in
 sage.matrix.matrix_double_dense.Matrix_double_dense.exp
 Failed example:
     A.exp()  # tol 1e-14
 Expected:
     [-19.614602953804912 + 12.517743846762578*I   3.7949636449582176 +
 28.88379930658099*I]
     [ -32.383580980922254 + 21.88423595789845*I   2.269633004093535 +
 44.901324827684824*I]
 Got:
     [-19.614602953804923 + 12.517743846762574*I  3.7949636449582007 +
 28.883799306580993*I]
     [-32.383580980922275 + 21.884235957898447*I  2.2696330040935115 +
 44.901324827684846*I]
 Tolerance exceeded in 1 of 8:
     2.269633004093535 vs 2.2696330040935115, tolerance 1e-14 > 1e-14
 **********************************************************************
 1 item had failures:
    1 of  11 in sage.matrix.matrix_double_dense.Matrix_double_dense.exp
     [636 tests, 1 failure, 3.14 s]
 }}}

 I've simply increased the tolerance from 1e-14 to 1.5e-14.

--
Ticket URL: <https://trac.sagemath.org/ticket/21119>
Sage <http://www.sagemath.org>
Sage: Creating a Viable Open Source Alternative to Magma, Maple, Mathematica, 
and MATLAB

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

Reply via email to