fix: SUSY, fix default ktdurham parameter for negative masses
For SUSY generation, the ktdurham parameter is set to 1/4th the mass of the heaviest non-decoupled sparticle, unless it's specified by the user. Currently, the method for deducing this mass doesn't account for e.g. higgsino production where the heaviest non-decoupled sparticle can have a negative mass in the param_card. Instead, this function sets ktdurham according to the largest positive non-decoupled mass, which isn't quite right.
This is a pretty trivial fix to make sure that negative masses are also considered when setting ktdurham.