[Flexiblesusy-commits] [FlexibleSUSY/FlexibleSUSY] 0ac525: refactor tachyon check

[GitHub]

  Branch: refs/heads/development
  Home:   https://github.com/FlexibleSUSY/FlexibleSUSY
  Commit: 0ac525733b2617050063a630b786e8a51311978b
      https://github.com/FlexibleSUSY/FlexibleSUSY/commit/0ac525733b2617050063a630b786e8a51311978b
  Author: Alexander Voigt <Alexander.Voigt at desy.de>
  Date:   2016-01-30 (Sat, 30 Jan 2016)

  Changed paths:
    M meta/TreeMasses.m

  Log Message:
  -----------
  refactor tachyon check


  Commit: b1fa9a473bda955656d0dcfb489c9a0f57d876b1
      https://github.com/FlexibleSUSY/FlexibleSUSY/commit/b1fa9a473bda955656d0dcfb489c9a0f57d876b1
  Author: Alexander Voigt <Alexander.Voigt at desy.de>
  Date:   2016-01-30 (Sat, 30 Jan 2016)

  Changed paths:
    M meta/LoopMasses.m
    M meta/TreeMasses.m

  Log Message:
  -----------
  flag bad mass only for massive particles

The criterion for an imprecise mass output will almost always fail for
multiplets which contain at least one massless particle.  The reason
is, that the smallest eigenvalue will be close to zero, but not quite.
The criterion to flag a bad mass,

   eigenvalue_error > precision * Abs(MVPVZ(0))

will certainly be true, because on the r.h.s. you multiply my
approximately zero (the massless eigenvalue).


Compare: https://github.com/FlexibleSUSY/FlexibleSUSY/compare/22466e52c754...b1fa9a473bda