Problem with SSP - plotting and No valid parameters found in SSP scan

[FStaub]

Hi,

you are correct. It's Perl script not a shell script, i.e. the ending is 'pl'.

Cheers
Florian

[felipeagf]

Hello Florian

I am running SSP using Example2_MicrOmegas_ContourScan.m of SSP/Input directory as a starting point in a Low Scale version of UMSSM which is in good agreement with the SSP template written for low scale input by SARAH. But when I run SSP it starts working just for few minutes and after that it stops without showing no error messages. Even when I start Mathematica from where I run SSP via the Terminal, before stopping the scanning operation, it just says that the parameter scan operation was killed without indicating what caused that problem.

Attached here you will find an image of my screen that shows how this problem appears in the Terminal and also the important input files I used in this parameter scan operation called ScanTANBvSlam in order to identify what was the cause of this problem.

In this case of problem where the operation starts running well but suddenly stops in one of the intermediate steps with no ERROR MESSAGES what do you think that should be done to avoid this?



Thank you very much for your attention.




Cheers
Felipe

[FStaub]

Hi,

you can try mathematica 8. The kernel is much more stable than the one from version 11.

Cheers,
Florian

[felipeagf]

Hello Florian

I am making some linear distribution scans over the main parameters of UMSSM (tanB, \lam, vS, Tlam) in a wide range of values. This, of course, causes huge modifications in the masses of the particles of this model and, consequently, this should cause significant variations in the relic density observable calculated by micrOmegas. But, after making these scans using the files that are attached here, when I open the plots of (Relic Density) vs (Masses) related to these operations, all the plots show a constant value for the relic density for all points. This is weird because when I run micrOmegas with the same spectrum file and make a simple variation in one of the parameters tanB, \lam, vS or Tlam I obtain a very different value for the relic density.

I attached here some of these weird plots obtained from SSP which do not show variations of the relic density. You can run the files attached here in your computer to see why the relic density calculated during the intermediate steps of the scan operation does not vary.



Best wishes
Cheers
Felipe

[felipeagf]

Hi Florian

I attached here the DefaultSettings file that I used in my scan analysis.



Cheers
Felipe

[FStaub]

Yes, there is obviously something going wrong. Maybe, it helps to take a look at the SpectrumFiles.spc in the output directory. There you can see all information calculated by micromegas., Maybe, something looks fishy. Otherwise, try to run some points "by hand" without SSP, i.e. copy the spc file to the micromegas directory and run CalcOmega. Maybe, any error message shows up there.

Cheers,
Florian

[felipeagf]

Hi Florian

Thanks for your tips.


Cheers
Felipe

[felipeagf]

Hi Florian

I looked at the SpectrumFiles.spc generated by SSP after making some Scans and I noted that the problem is that for each parameter point SSP writes in the Block DARKMATTER the same results for all Dark matter properties. Is there a way to modify this operation in order to make SSP operate correctly and write for each parameter point of the scan its correct dark matter properties? In other words, when I run SSP, it generates a file called SpectrumFiles.spc which contains all the correct results of Spheno associated with each parameter point and a block called DARKMATTER which shows the results from MicrOmegas, but this block does not show the correct results of MicrOmegas that should be written in this spectrum file (SpectrumFiles.spc) generated by SSP. Why is this happening?


I attached here the spectrum file generated by SSP which shows this problem in the block called DARKMATTER and I also attached here my input scan file together with the DefaultSetting file that I used.
How can I solve this problem and make SSP show in its SpectrumFiles.spc the correct values of Dark Matter properties in DARKMATTER block in the same way it writes the correct values obtained from SPheno in all the other blocks.





Cheers,
Felipe

[FStaub]

Hi,

this looks strange and shouldn't happen:
SSP includes the content of omg.out at the end of the spectrum files. And before MicrOmegas is called, it deletes the omg.out from the previous point to make sure that not the same information is by accident stored for different points. So, I can only think of two problems:
1) There is a problem with copying the SPheno.spc.UMSSM file into the micromegas directory. That you can easily check. Is this file updated for each point?
2) There is a problem with deleting and writing the omg. out. This might have to do with user privileges for instance which forbid to change that file.
Did you run SSP via the terminal? In that case you should get warnings if some file operations are not allowed.

Cheers,
Florian

[felipeagf]

Hi Florian

Thank you for your reply.

When I run SSP via the Terminal it shows the following warning messages

rm: cannot remove '/afs/atlass01.physik.uni-bonn.de/user/felipeag/Computerprograms2/SSP-1.2.3/Output/RANDOMScanvS/SpectrumFiles.spc': No such file or directory
rm: cannot remove 'omg.out': No such file or directory
rm: cannot remove 'omg.out': No such file or directory
rm: cannot remove 'omg.out': No such file or directory
rm: cannot remove 'omg.out': No such file or directory



So, maybe it is related to the second problem you mentioned. When I run micromegas in my model directory UMSSM it uses as input a file called "SPheno.spc.UMSSM_low" since this is a low energy scenario and writes in the same directory an output file called omg.out which I do not know why SSP cannot locate it and write it in the SpectrumFiles.spc it produces. I thought the problem could be solved writing the correct input file name in the DefaultSettingsUMSSM file used by SSP but since it asks the "DEFAULT[SPhenoInputFile]", "DEFAULT[SPhenoSpectrumFile]", "DEFAULT[MicroOmegasInputFile]" and "DEFAULT[MicroOmegasOutputFile]" so I wrote respectively "LesHouches.in.UMSSM", "SPheno.spc.UMSSM", "SPheno.spc.UMSSM" and "omg.out" as the DEFAULT names for these input and output files.

So how can I make SSP start using the correct first omg.out file produced by micrOmegas and delete it before going to the next parameter point in order to write it's correct DARK MATTER predictions?

Sorry for this question.



Cheers and Best wishes
Felipe