Error during calculation of beta functions

[AlexanderVoigt]

Dear SARAH developers,

a student and I are writing a SARAH model which consists of the SM with extra charginos, neutralinos and sleptons (see attachment).
We ran CheckModel[] to check that there are no obvious mistakes. However, when starting the model and calling CalcRGEs[] like this:

Code: Select all

Needs["SARAH`"];
Start["LSplitMSSM"];
CheckModel[];
CalcRGEs[];

then during the calculation Mathematica (version 9) reports some internal errors:

Code: Select all

Calculate non-supersymmetric RGEs
Initializing Invariants... 

Part::partw: Part 1 of {} does not exist.

Part::partw: Part 1 of {} does not exist.

Part::partw: Part 1 of {} does not exist.

General::stop: Further output of Part::partw
     will be suppressed during this calculation.

Part::pspec: Part specification {}[[1,1]]
     is neither a machine-sized integer nor a list of machine-sized integers.

Part::pspec: Part specification {}[[1,1]]
     is neither a machine-sized integer nor a list of machine-sized integers.

RGEs::StillEntangled: Can't disentangle the contributions to the running of λ
Check: -6 g1hhhh caa[1] - 6 g2hhhh caa[1] + 3 λ caa[1] - 2 g1hhhh caa[4] - 
...

Is there a mistake in our model file, or maybe some bug in SARAH or incompatibility with Mathematica 9?

Many thanks in advance,
Alex

[FStaub]

Hi,

first of all: this model is a mess! Are you sure you want to study it?

Okay, more seriously:
There are problems because SARAH can't disentangle the running of different parameters. (See for instance http://stauby.de/sarah_userforum/viewto ... angle#p699 what that means).
In your model, this appears very often. That means you have a redundant set of couplings which are linear dependent. For instance, I can't see any difference between g1hhhh and Lambda concerning the involved particles and their quantum numbers. The same is most likely true for many other couplings.

Thus, when you have something like

Code: Select all

c1 phi1.phi2.phi3.phi4 + c2 phi1.phi2.phi3.phi4 + ..  =  (c1 + c2 + c3) phi1.phi2.phi3.phi4

in your Lagrangian, you need to introduce a new parameter ('effective coupling')

Code: Select all

C phi1.phi2.phi3.phi4

to be able to get the RGEs. There is no way to let c1, c2, c3 run separately because they are now part of the same coupling...

Cheers,
Florian

PS: Another benefit would be, that your model becomes more readible.
PPS: Since I'm not sure that this is a bug, I'll move the topic

[AlexanderVoigt]

Hi Florian,

many thanks for the hint, I did in deed overlook this redundancy!
The model becomes a bit neater if one removes the redundant couplings, but as usual with EFTs: A full SUSY model is always nicer!

Best,
Alex