Tutorial - Part 1: Our First Model Point Test

Initialising CheckMATE

Let us go to the directory /home/yeti/Downloads/CheckMATE-1.1.14/bin where you should find the CheckMATE executable. To enter the cross sections, event files and other settings, CheckMATE provides two possible means. For now we will stick to the more intuitive way, which is to provide a single text file with all the options inside (a so called parameter file).

In the same folder you should find an example parameter file testparam.dat with a minimally working example. In general, the structure of such a parameter file looks as follows: ## General Options
[Mandatory Parameters]
Name: (a string that defines the name of the output directory)
Analyses: (which analysis / analyses to be tested)

[Optional Parameters]
(if wanted, set extra parameters here via 'Parameter: Value')

## Process Information (Each new process 'X' must start with [X])
[(Process Name 1)]
XSect: (number) (unit)
XSectErr: (number) (unit)
Events: (location of the event file for process 1)

[(Process Name 2)]
XSect: (number) (unit)
XSectErr: (number) (unit)
Events: (location of the event file for process 2)

...

Using what we have, we want to run the following settings:

• Our run should have the name 'cMSSM'.
• We want to use analysis 'atlas_conf_2013_047' (we will see later why).
• To make our results easier to compare, we should use 'RandomSeed: 42' as an optional parameter.
• We want to use the event files we just created and the NLO+NLL cross sections for the two processes gluinos and stops.
• We don't want to consider the impact of the theory error for now and therefore should set the cross section errors to 0.

If you run CheckMATE with the correct setup described above, it should start with the following crosscheck output:

If your parameter file is ready and working, you can go ahead and start the CheckMATE run by answering y after CheckMATE's startup question. While it is running, let us proceed to the next page where we will briefly discuss what it actually does internally.

Continue