Cross sections and luminosities

CMS Luminosity
#1

Hi all

I am trying to create an exercise for one of my classes using CMS Open Data. I tried to look at some data and MC samples, but I cannot find the information about cross sections and luminosities. For instance, for this data sample [1] and this MC sample [2]. Is there an easy way to know this?

As a suggestion, for data maybe it would be easy to know the luminosity per txt file. In that sense I can only process a few txt files and normalized my MC samples to that subset of the total dataset.

Thanks a lot for the help
cheers,

[1] CERN Open Data Portal
[2] CERN Open Data Portal

#2

for the first one the cross sections and luminosities are in the json file - in the link CERN Open Data Portal.

#3

Thanks for the answer @susmita-sbhaduri but in the json file I see the list of good runs, not how much luminosity I get.

The question is about, if I only process RunB for instance, how much luminosity do I have? Or, if I dont want to process the whole RunB and just a subset, how can I calculate the amount of luminosity?
In current CMSSW we can use brilcalc, not sure if it is the same here.

thanks again

#4

Dear @algomez,
In the latest CMSSW, you must have noticed in the script “demoanalyzer_cfg.py” that FWCore.PythonUtilities.LumiList utility is imported and Lumisection for each event is extracted with “myLumis = LumiList.LumiList(filename = goodJSON).getCMSSWString().split(’,’)”, where goodJSON is the very json file with good run numbers and corresponding luminosity sections.
Further you may check other functions of FWCore.PythonUtilities.LumiList for getting luminosity values in the thread FWCore.PythonUtilities.LumiList broken in CMSSW_10_2_0_pre6? · Issue #23774 · cms-sw/cmssw · GitHub.

Hope this would help you.
Best
Susmita

#5

thanks again for the reply.
But I think I am still confused. With the LumiList tool I pass to the code the lumisections to be processed but, as far as I know, that does not compute the amount of integrated luminosity I get with these lumisections. Am I getting this right?

#6

There is information here CERN Open Data Portal on how to calculate luminosity for various options.

There is also this tutorial Luminosity available. Chapter 3 is probably the most relevant to your queries.

brilcalc is also available as a docker container (this is not yet documented on the portal). If you have docker installed it’s as simple as

docker run -it --name brilws gitlab-registry.cern.ch/cms-cloud/brilws-docker

which should bring you to the command-line prompt and brilcalc should be available:

bril@48c3625738ff:~$ brilcalc --version
3.6.2

#7

Thanks @mccauley, this is exactly what I need for the luminosity.
I’ll take a look.

I guess that the comment about the cross sections is still open.

#8

Good to hear @algomez. Yes, cross sections are still an open issue.

#9

You are absolutely correct. I think by now you have run brilcalc in docker from the instructions given in the link CERN Open Data Portal.
I was not aware on the tutorial though. Thanks to @mccauley for that.

#10

Hi @algomez, I am afraid the most precise information regarding cross-sections is still a work in progress for our open data effort. We are making our best efforts to provide this information in a centralized location.

Now, is there a particular reason why you would want to use this particular MC dataset? Note that the DY1JetsToLL has a “at least one jet” restriction. You could use a more general dataset (for which there is even a nanoAOD-like derived dataset for outreach). The cross section for this process was well cataloged in a CMS internal report (CMS-SMP-15-008), which was the base for one of CMS published analyses. The number quoted there at NLO is 3503.71 pb (which is also corroborated by its usage in an outreach H->tau tau analysis). I believe that this would be enough it is for education purposes. Please let us know if this suits you.

#11

Hi @ecarrera

thanks for the information. Yes I can use the more general DY sample that you pointed out. But since I was thinking on some other exercises I was trying to see if any other MC sample listed in opendata has the info about cross sections.

Thanks again for all the help.

cheers,