Issue with maddump script

Dear Dipyaman,

1) "InvalidParamCard : Parameter yukawa: 1 is not at zero"

this error is due to a restriction in the model file. If you look inside the DMsimp model folder (under the directory models) you will find some files named restrict-*.dat. As a first approach, you can remove them and try to re-run from scratch the script. This should fix your issue (anyway, give a look if some of the restrictions may be useful for you)

2) I realized that there is a typo in the script from the paper. Be careful to make the replacement (towards the last lines):

set wy1 auto -> set wy0 auto

(y0 is the scalar mediator here)

3) I've just uploaded on the trunk branch of maddump, all the required files to run the examples in the paper ( except for the leptophobic case, the relative UFO file is not public) in the example_SHiP folder. Be aware, of course, that in order to reproduce the final sensitivity plots, some extra work is needed starting from the maddump results. For example, the electron recoil cuts are imposing afterwards on the generated events and the phase space factor associated to the decay pi > a y1 has been added a-posteriori (we are working to improve the book-keeping of all the possible factors).

Cheers,
Luca

Dear Luca,

Thank you very much for the clarification. It is now running after I removed the restriction files.

with best regards,
Dipyaman

Another way to avoid the restriction file is to do
import MODELNAME-full
instead of
import MODELNAME

Cheers,

Olivier

Thanks Luca, that solved my question.

Hi,

There are some more questions. I am a little confused about the events spectrum in maddump. If I am not wrong, the event information is stored in ehist.dat. Can you kindly tell me what are the columns of that file? One more thing. Can you kindly tell me how do you fix the normalization? For e.g. If I take SHIP, how do you get the number 19663072216.4 in the flux norm?

 I will be highly obliged to you if you kindly answer the queries.

Thanks in advance,
Dipyaman

Dear Dipyaman,

I do not understand completely your first question. Let me explain maddump outputs. When you perform a scan, the main results of the scan are summarized in text file named scan_run[xxx].txt, while the events file are located in the Events directory of your production/interaction/decay process as in the usual mg5 output (in lhe format). I would suggest to activate the scan mode also when you want to run only one specific point in the parameters space in order to get the summary file.
When you turn on the testplots options ( as in the input card in the appendix of the paper ) other output file are produced and stored in the same run_xxx folder as for the output events file for the interaction process. They allow to check the intermediate steps of the simulation. In particular:
1) in_DM.dat: it is the list of the particles which actually cross the detectors and contribute to the 2dfit (each point is given in the format E, theta, cos(phi), sin(phi) with theta and phi as defined in the paper )
2) ehist.dat: it is the histogram of the effective energy DM pdf (the first two columns are the edges of the bins, the third column the height and the fourth the error )
3) cell_fortran.dat: it is the collection of the squares of the 2d mesh (the result of the 2 fit) and mesh2D.png is the corresponding plot in the 2D plane.
I hope this will guide to you to fix your issue.

For what concerns the normalization factor, we normalize the flux in order to take into account the following prefactor related to the production mechanism

Number of POT * xsec(pN > DM DM~ ) / xsec (pN > anything )

We assumed that the cross section (xsec) proton - Nucleous ( in our case Molybdenum, A = 96 ) scales with respect to the proton proton collision as follows

a) xsec(p N > DM DM~) = A * xsec(p p > DM DM~)

b) xsec(p N > anything) = A^(0.7) * xsec(p p > anything)

Since the xsec(p N > DM DM~) is computed by maddump (and add to the result setting to True the flag prod_xsec_in_norm ) and it is given in pb, we set the following normalization factor

flux_norm = number of POT * A^(0.3) / xsec (p p > anything ) pb^-1
                     = 2*10^20 * 96^(0.3) / (40 *10^9) pb ^-1
                     = 19663072216.4 pb^-1

(where 40 mb is the pp xsec )

Cheers,
Luca

Dear Luca,

Thank you for the clarification. Just one more doubt. Did you use total events only to calculate the sensitivity or shape of the distribution also used?

With best regards,
Dipyaman

Dear Dipyaman,

please could you clarify your question? What are the
distributions you are referring to?

In the computation of the sensitivity, we take into account
the properties of the incoming beam of DM produced at the
production point and the geometrical acceptance of the
detector, keeping the main effects of the correlations between
them.

Cheers,
Luca

Dear Luca,

I want to know how you define chi^2. Is it like,

chi^2 = (total signal events)^2/total_back_ground_events

?

Regards,
Dipyaman

Thanks Luca, that solved my question.