MadGraph5_aMC@NLO
MadGraph5_aMC@NLO [AFF+14] [FFH+18] is a general-purpose event generator that is mainly used to compute observables in double hadronic environments such as the LHC.
Pinecard structure
The
output.txtfile (compulsory). This file contains the instructions to generate the source code for the relevant process. For details, please see [AFF+14] and [FFH+18]. The variable@OUTPUT@must be used to generate the directory containing the source files.The
analysis.ffile (compulsory). This Fortran file must fill the histograms from which the HwU files and the PineAPPL grids are generated. Note that a single histogram must not contain more than 100 bins, otherwise MadGraph5_aMC@NLO will crash. However, big histograms can be split up into multiple histograms, for which the runner will merge the PineAPPL grids together.The
*.patchfile(s) (optional). These are one or more.patchfiles that are applied after MadGraph5_aMC@NLO has generated the sources.
launch.txt (compulsory)
This file contains the instructions to run the relevant process, including the relevant physical parameters and cuts.
Theory parameters
To insert the actual values of the theory parameters coming from the theorycard
we provide a special syntax. You can use the names
@GF, @MH@, @MT@, @MW@, @MZ@, @WH@, @WT@, @WW@,
and @WZ@, which will be replaced with their numerical values upon generation.
The names are the same as chosen by mg5_aMC, but written in
uppercase and surrounded with @. For details about more parameters, please
see the Template/NLO/Cards/run_card.dat file in MadGraph5_aMC@NLO.
Cuts
They are implemented in two steps:
cuts relevant variables are defined
cuts code is implemented
A list of available codes and variables can be obtained from the repository.
Patches
For instance, to use a dynamical scale, a patch modifying setscales.f file
should be included in the directory. To create patches use the command diff
-Naurb original new > patch.patch. The patches are applied in an unspecified
order, using patch -p1 ....
A list of available patches can be obtained from the repository.
Additional metadata
output.txt: contains the generatedoutput.txtscript (after all substitutions have been done)launch.txt: contains the generatedlaunch.txtscript (after all substitutions have been done)patch: a list of patches’ names, one per row (corresponding to those described in Patches)tau_min: the minimum \(\tau\) value set by the useruser_cuts: user defined cuts and cuts values, one per row in the formatcut=value(cuts are those defined in Cuts)mg5amc_repoandmg5amc_revno: The repository and revision number of the MadGraph5_aMC@NLO version that was used to generate the grid.
Note
It is guaranteed that the keys listed above are always present in grid’s metadata (even if some of the corresponding values might be empty).
Output
DATASET: The directory created bymg5_aMC. A few interesting files in this subdirectory are:Events/-/MADatNLO.HwU: HwUEvents/-/amcblast_obs_-.pineappl: grids created bymg5_aMC, not yet merged together
output.txt: Run card for the ‘output’ phase, with all variables substituted to their final valuesoutput.log: Output of the external runner during the ‘output’ phaselaunch.txt: Run card for the ‘launch’ phase, with all variables substituted to their final valueslaunch.log: Output of the external runner during the ‘launch’ phaseresults.log: The numerical results of the run, comparing the results of the grid against the results frommg5_aMC. The first column (PineAPPL) are the interpolated results, which should be similar to the Monte Carlo (MC) results in the second column. The third column gives the relative MC uncertainty (sigma). The next column gives the differences in terms of multiples of sigma. The final three columns give the per mille differences of the central, minimum, and maximum scale varied results. Ideally the first two columns are the same and the remaining columns are zero.