MadGraph5_aMC@NLO

MC error bin-by-bin

Asked by Keping Xie

Dear everyone,

I am Keping Xie.

I want to estimate the Monte-Carlo errors of a histogram bin-by-bin. I look at the code HwU.f. The following function function can achieve this goal.
      subroutine HwU_accum_iter(inclde,nPSpoints)
However, how I can invoke it in the code in FixedOrderAnalysis, such as analysis_HwU_template.f?

How can I get know the vairable of nPSpoints?
In the code SubProcesses/mint-integrator2.f, there is a variable
         ntotcalls(i)=ncalls*kpoint_iter
return the number of Phase space points. Is it possible to pass this variable to the code analysis_HwU_template.f?

Best regards,
Keping

Question information

Language:
English Edit question
Status:
Solved
For:
MadGraph5_aMC@NLO Edit question
Assignee:
Valentin Hirschi Edit question
Solved by:
Valentin Hirschi
Solved:
Last query:
Last reply:
Revision history for this message
Best Valentin Hirschi (valentin-hirschi) said : 		#1

Monte-Carlo errors on a bin-per-bin basis are already available in MG5aMC FixedOrder analysis using the HwU format.

For instance if you run 'p p > e+ ve [QCD]' at fixed-order NLO using the pre-existing analysis 'analysis_HwU_pp_lvl.f' (that you specify by entering 'analysis_HwU_pp_lvl.o' in the entry 'FO_ANALYSE' of the FO_analyse_card.dat) you will see that the resulting plots in:

<PROC_OUTPUT_DIR>/Events/run_01/MADatNLO.pdf

does show the MC-uncertainty bin by bin.
So by simply following the instructions in the template analysis, you should also obtain these uncertainties.

Let me note that the output of the analysis is the file 'MADatNLO.HwU' which specifies the raw data for the plots.
This is then post-processed automatically by MG5_aMC to produce the pdf shown above.
But you are free to to this post-processing yourself so as to tweak the histrograms generated. You can do this by using our in-house render stored in:

<MG_install_dir>/madgraph/various/histograms.py

You can access help on the functionality of this renderer with the option '--help':

./madgraph/various/histograms.py --help

And you can find more detailed information on its usage here:

https://answers.launchpad.net/mg5amcnlo/+faq/2671

Revision history for this message
Keping Xie (kepingx) said : 		#2

Thanks Valentin Hirschi, that solved my question.