LPC Computing
Jetmet Tutorial
This Tutorial is obsolete! It was developed with older versions of the
CMS software. In addition the computing system was upgraded several times
since the tutorial was given early in 2004. You might still want to browse
the pages to get an overview but the hands on examples will not work anymore.
We are in the process of developing a new tutorial that works in the new
software and computing environment and that will utilize more up to date CMS
software.This is currently work in progress:
http://www.uscms.org/SoftwareComputing/UserComputing/New_tutorial.html
Jetmet: A physics example
Introduction:
Disclaimer: this is only an example to get going! Remember that OSCAR and ORCA are work in progress! We want to demonstrate how to work within the CMS environment setup on the UAF farm at FNAL.
For Tutorials concerning CMKIN and the CMS OO software have a look at the Tutorials given at CERN :
| The Tutorial main page: | http://cmsdoc.cern.ch/cms/software/tutorials/ |
| Absolute beginner: | http://cmsdoc.cern.ch/orca/Tutorials/20031114-CERN/NewUser/ |
| Full analysis chain: | http://cmsdoc.cern.ch/orca/Tutorials/20031128-CERN/FullChain/Contents.html |
| COBRA: | http://cobra.web.cern.ch/cobra/Tutorial743/index.html |
| Tracking: | http://cmsdoc.cern.ch/orca/Tutorials/20040109-CERN/index.html |
Other good web pages to find information are: http://cobra.web.cern.ch/cobra/Tutorial743/index.html and http://cmsdoc.cern.ch/cmsoo/cmsoo.html
We hope this will still be useful and hopefully this documentation gets better as we get some feedback. All examples were created and tested on Redhat 7 based systems (ssh -t cmsuaf.fnal.gov +r RH7, don't forget the PROC_RESOURCES=RH7 statement in the jdf file) and will not!! work on the RedHat 6 systems. We will use some simple physics analysis to demonstrate how:
- Monte-Carlo events are generated and stored (CMKIN)- The CMS detector is described and its interaction with particles are simulated (OSCAR)
- Simulated events are digitized (ORCA)
In addition we will learn how to:
- setup the CMS software environment here at Fermilab
- use the farm for a small MC production
- generate a MC sample, convert the ntuple with h2root into a root file,...
- how to use root to make a few plots on generator level
- how to run the generated events through OSCAR the GEANT4 base detector
simulation.
- how to run the digitization step using ORCA
- how to run some ntuple makers on the ORCA output
- do some simple analysis
Click the links below to start the exercise. Be aware that you have to do the exercise in sequence:
- General setup
- Event generation on the farm with CMKIN (Pythia), Analysis of the generator level output
- Simulate the events with OSCAR
- Digitize with ORCA
- Run a small ntuple maker ( from Stephan's ORCA tutorial) looking at Calorimeter quantities
- Run the JetMet ntuple maker and make some plots
The physics analysis we will use is a Z' with a mass of 700 GeV that we force to decay into 2 light quark jets. (This actually might be one of the first discoveries at the LHC). The scope here is to learn about the computing environment and to do a very simple analysis. In this exercise we don't mix in pile up!
The figure below shows the simple and robust processing model we will use in all the examples involving the farm to process events in parallel. When the process start all the input data is staged in from mass storage.
Then all the libraries are loaded and all necessary database constants (geometry ...) are read in. During the processing the nodes are completely independent all the output is to local disk. In the end the results are stored in Mass storage. This is very robust all nodes are completely independent. In case a process fails we just have to resubmit that process.
