[SPheno] Compilation Error in SARAH 4.9.0

[JPEllis]

In version 4.9.0 of SARAH, the SPheno output fails to compile. I have checked this with both SPheno's Makefile and with CMake. The error in both cases is:

Code: Select all

2LPoleFunctions.f90:6:4:

 use model_data_THDMII
    1
Fatal Error: Can't open module file ‘model_data_thdmii.mod’ for reading at (1): No such file or directory
compilation terminated.

Do the changes to do this this mean that I have link against other model's SPheno code? Or is each model self-contained?

[FStaub]

Hi,

I don't think that this is a problem of any of the new features with SARAH 4.9.0 but related to the organisation of the SPheno files in different subdirectoires and the usage of VPATH in the Makefile. However, it's surprising because it worked fine on all machines and with all compilers I tested it. What are you using?
Can you try to move all files back from 3-Body-Decays, Observables and TwoLoopMasses to the main directory and try if compilation works then?

Cheers,
Florian

[JPEllis]

I probably should've mentioned that the model I'm trying to compile is not the THDMII model; in fact, my model makes no references whatsoever to THDMII. The only reference to the THDMII I believe comes from the new matching conditions which default to the THDMII (as I had none).

I haven't tried yet moving everything from the subdirectories to the main directory, but I doubt that'll fix the issue as there is no file "model_data_THDMII" anyway, so the use statement will fail regardless unless I link to an external model. I'll update this post once I've tried. EDIT: Moving everything into the same folder still produces the same error.

The file Package/SPheno/IncludeSPheno/2LPoleFunctions.f90 contains a reference to THDMII, could that be the cause of the issue?

[FStaub]

Yes, there was the wrong version of Package/SPheno/IncludeSPheno/2LPoleFunctionsMSbar.f90 in the tar-ball. You can just remove the line which refers to THDMII.

Cheers,
Florian

[JPEllis]

I have removed the line

Code: Select all

use model_data_THDMII

from Package/SPheno/IncludeSPheno/2LPoleFunctions.f90 which has resolved this initial error.

It now has another error:

Code: Select all

/scratch/gitlab/build/724d280d/0/JP-Ellis/E6Seesaw/SPheno/models/E6Seesaw/TwoLoopMasses/2LPole_E6Seesaw.f90:23688.43:

Pi2P = Matmul(Pi2P,PseudoScalarMixingMatrix)
                                           1
Error: Symbol 'pseudoscalarmixingmatrix' at (1) has no IMPLICIT type
/scratch/gitlab/build/724d280d/0/JP-Ellis/E6Seesaw/SPheno/models/E6Seesaw/TwoLoopMasses/2LPole_E6Seesaw.f90:2789.13:

tad2L=matmul(temptad,SMS)
             1
Error: Different shape on dimension 1 for arguments 'matrix_a' and 'matrix_b' at (1) for intrinsic matmul
/scratch/gitlab/build/724d280d/0/JP-Ellis/E6Seesaw/SPheno/models/E6Seesaw/TwoLoopMasses/2LPole_E6Seesaw.f90:13351.14:

Pi2S = Matmul(Pi2S,SMS)
              1
Error: Different shape on dimension 2 for argument 'matrix_a' and dimension 1 for argument 'matrix_b' at (1) for intrinsic matmul
/scratch/gitlab/build/724d280d/0/JP-Ellis/E6Seesaw/SPheno/models/E6Seesaw/TwoLoopMasses/2LPole_E6Seesaw.f90:13352.14:

Pi2S = Matmul(Transpose(SMS),Pi2S)
              1
Error: Different shape on dimension 2 for argument 'matrix_a' and dimension 1 for argument 'matrix_b' at (1) for intrinsic matmul

As with the initial error, this is independent of the way I build it (i.e. whether I use SPheno's makefile or my CMake script).

[FStaub]

It seems that you defined a particle via partciles.m as "Pseudo-Scalar", but you didn't define in parameters.m the corresponding mass matrix. However, you study a non-SUSY model, or? You shouldn't get the two-loop mass routines at all (where all the problems show up), because they only work fine so far for SUSY models. The reason is in Package/SPheno.m:

Code: Select all

If[OnlyLowEnergySPheno===True,
Include2LoopCorrections=False;,
Include2LoopCorrections=True;
];

The correct check is

Code: Select all

If[NonSUSYModel===True,
Include2LoopCorrections=False;,
Include2LoopCorrections=True;
];

Cheers,
Florian

[JPEllis]

Indeed I do study a non-SUSY model, and I have a set of particles called "Pseudo-Scalar Higgs"; but I did also define the corresponding pseudo-scalar mixing matrix (analogous to the scalar mixing matrix) in parameters.m.

I have applied the change you suggested and I'm currently regenerating everything. I'll let you know when it's done if it resolve the errors.

[JPEllis]

I have applied the change you suggested to Package/SPheno/Spheno.m, and I still get the same error unfortunately.

It appears that when it comes to that change, NonSUSYModel is set to False; and looking through the code, there's only a single location where NonSUSYModel is set (line 138 of variables.m). Does SARAH 4.9.0 require that NonSUSYModel be set explicitly in my model.m? Searching through the wiki, I couldn't find any reference to that variable.

If a set NonSUSYModel to be True before running MakeSPheno, the generated files work fine.