[Flexiblesusy-commits] [FlexibleSUSY/FlexibleSUSY] b44a5e: use Broyden instead of Newton as it is much more s...

GitHub noreply at github.com
Mon Jun 2 08:29:36 BST 2014


  Branch: refs/heads/development
  Home:   https://github.com/FlexibleSUSY/FlexibleSUSY
  Commit: b44a5ec348ddd1d8c3e5faf99d088f05b49fcee0
      https://github.com/FlexibleSUSY/FlexibleSUSY/commit/b44a5ec348ddd1d8c3e5faf99d088f05b49fcee0
  Author: Expander <Hole.destructor at gmx.de>
  Date:   2014-06-02 (Mon, 02 Jun 2014)

  Changed paths:
    M templates/two_scale_model.cpp.in

  Log Message:
  -----------
  use Broyden instead of Newton as it is much more stable

I just did a CMSSM parameter scan over tan(beta) from 1 to 45.
The newton method fails for some points with high tan(beta) and
is much slower than Broyden.




More information about the Flexiblesusy-commits mailing list