Recent Changes - Search:

<< Back to ANL ACS

Home Page

Working Examples

PmWiki

edit SideBar

Main /

HomePage

Welcome to the JULY ANL ASC Analysis Workshop Wiki

Introduction

There will be an Analysis Workshop at ANL ASC from 14 July through 16 July. This workshop will use FDR-2 data and will cover similar topics to the BNL Jamboree to be held 24-27 June. FDR-2 data is being processed currently (first week of June) so that we fully expect the data to be ready by the time of the workshop.

As with our previous workshops, there will be a lot of time devoted to working, learning and collaborating, with a small number of talks aimed at providing practical information for working with the FDR data.

The workshop will be designed to help ATLAS physicists at all levels of familiarity with the ATLAS software. Working examples suitable for novices as well as those with expertise will be provided.

You will get the most out of attending all three days--but this is also not necessary in order to profit from this workshop.

Agenda

The tentative agenda is HERE.

Registration

The registration page is HERE.

Participants

The participant list is HERE.

FDR-2 data sets in DQ2

FDR2 data is now available for you to look at. Be aware that there are remnants of earlier production attempts(with some problems) hanging around. The correct data set ends with .o3_f8_m*

To find the data sets use DQ2 tools e.g. 

 > dq2_ls -g -t -r fdr08_run2*AOD*o3_f8_m*

You can use athena version 14.1.0 to read the AODs.

FDR-2 files installed locally

You can find sample FDR data sets (AOD, TAG and DPD) at 

 /data/nas2/users/ryoshida/fdr08_run2/

and FDR-2 ESD data in: 

 /data1/torregrosa/FDR2/ESDs/

Example job running on FDR-2 data locally

A simple example of an athena job that is running on FDR-2 AOD is in 

 /users/ryoshida/testarea/14.1.0/PhysicsAnalysis/...

You can get this to run by following the usual example (see e.g. last workshop) to set up the AnalysisSkeleton job on AOD with 14.1.0, copying the three files

  • /users/ryoshida/../run/AnalysisSkeleton_topOptions.py.mod
    • ../src/AnalysisSkeleton.cxx.mod
    • ../UserAnalysis/AnalysisSkeleton.h.mod

to your area (without the .mod) and compiling as usual. (And change the data set you're accessing)

Basic Coding in Athena

If you are new to writing analysis code in Athena a simple tutorial is provided. A new package is created from scratch that allows one to access a data file and write variables to an ntuple. The code is explained and the user is guided through simple modifications. The hope is that after working through this, the user will be able to make some sense of the more sophisticated packages such as HighETJets or PromptGamma. The tutorial can be accessed here

Example job running on FDR2 data on Grid using pathena

Have the example above running from your ../14.1.0/../run directory 

 > cd ../run/
 > get_files ConfigExtractor.py
 > get_files FakeAppMgr.py
 > source ../cmt/setup.sh
 > source /share/grid/app/asc_app/asc_rel/1.0/setup-script/set_atlas.sh
 > pathena --inDS fdr08_run2.0052293.physics_Egamma.merge.AOD.o3_f8_m10 --outDS user.... AnalysisSkeleton_topOptions.py

Example run locally using TAG files

  • From the same ../14.1.0/../run directory
  • copy /users/ryoshida/../run/AnalysisSkeleton_topOptions_tag.py to your ../run >athena AnalysisSkeleton_topOptions_tag.py

Example job running on FDR2 ESD data locally

The first point is to copy the tar file: 

 /users/torregrosa/testarea/Moseley_Analysis.tar

into your testarea directory. If you have a 14.1.0 directory inside the testarea, copy it inside 14.1.0.

Uncompress it: 

 > tar -zxvf Moseley_Analysis.tar

Go to the cmt directory and make: source setup.sh if you use bash, source setup.csh if you use a c-shell type.

Compile it by typing: 

 > gmake

Then go to the run directory and edit the job options file: 

 Moseley_Analysis_topOptions.py

You can change the input data in: 

 ServiceMgr.EventSelector.InputCollections = [ "some.ESD.file" ]

this defines the input data file, change it to whatever you want to use as input.

If you want to run over several files or you want to send the job to the grid, then is preferable that you setup the input files in the following way: 

 import glob
 if not 'PoolESDInput' in dir():
     PoolESDInput = glob.glob("ESD.*.pool.*")
 ServiceMgr.EventSelector.InputCollections = PoolESDInput

you can use the glob module as you use ls, see manual.

We have a whole FDR2 ESD dataset that you can use to play around, just make symbolic links of the files from 

 /data1/torregrosa/FDR2/ESDs/

to your run directory

To run athena just type: 

 > athena.py Moseley_Analysis_topOptions.py | tee Mosquito.out

If everything works fine you'll get an athena aware ntuple as output file: 

 MoseleyAnalysis.aan.root

with the relevant information of jets using ESDs.

Please feel free to adapt it into your needs adding or deleting material. If you have any question or problem, do not hesitate to ask me (Esteban)

Using Athena tools in your analysis

One of the most interesting issues of the package described above is the use of an Athena tool. It is not straight forward to use it inside the code but it is worth to try because it avoids you to retype code already done and makes your analysis more standard and easy to understand. Here we have used a custom made Athena tool to calculate the radius of a jet as an example of a standard Athena tool. The Moseley_Analysis package makes use of this tool and you can incorporate to your analysis this or any other Athena tool available. The basic steps to incorporate an Athena tool into your code analysis are:

  • You should modify your requirements files(inside the cmt directory) in order to include the package that contains the tool that you want to use. We are going to use as an example the one developed in ANL (AnlJetTool): 
 use AnlJetTools		AnlJetTools*
  • Then modify you header file (in the Moseley_Analysis package is the file called MosquitoAlg.h inside the Moseley_Analysis directory). First include the header of the templeate class IAnlJetUtilTool and then create a new member of the type ToolHandle<IAnlJetUtilTool>: 
 #include "AnlJetTools/IAnlJetUtilTool.h"

 ToolHandle<IAnlJetUtilTool> m_AnlJetTool;
  • Finally modify your source file (in the Moseley_Analysis package is the file called MosquitoAlg.cxx inside the src directory):
    • In the class constructor initialize the ToolHandle 
    m_AnlJetTool = ToolHandle<IAnlJetUtilTool>("AnlJetUtilTool");
  • In the initialize method you must retrieve the tool from the ToolService 
    sc = m_AnlJetTool.retrieve();
    if (sc.isFailure()) {
      log << MSG::ERROR << "Unable to find " << m_AnlJetTool.type() << endreq;
      return StatusCode::FAILURE; }
  • Now you can start to make use of the tool, for instance in the execute method you can type: 
    m_AnlJetTool->FindRadius( m_JetContainerName, 75. , m_aan_jet_rad75 );

This method fills the vector m_aan_jet_rad75 with the radius of the jets contained in m_JetContainerName. The 75. means that I want the radius that contain 75% of the energy.

  • In the finalize method make a release the tool: 
   if( !m_AnlJetTool.release().isSuccess() ) return  StatusCode::FAILURE;

The amount of tools available in Athena is endless however their documentation is under development and it is possible that you may email the authors to find out the details.

How to run High-ET photon analysis program for the release 14.1

You can find this information on High-ET photon wiki

How to generate Monte Carlo events using the release 14.1

You can find this information on High-ET photon wiki

Using ELSSI for TAG selection

ELSSI page for FDR2 data is here. You will need to use a browser with your grid certificate.

Marco's Twiki page with an example usage is here.

Calculating luminosity

This was tested with 14.2.0. There have been reports that this does not work with 14.1.0.

> get_files LumiCalc.py

> LumiCalc.py --readoracle --mc --rs=52280 --ru=52304 --trigger=ALL (for example)

Use --help to get options. More information here.

Note --readoracle is needed always from outside CERN.

Note --mc is needed for FDR data. (not for real data)

Useful outside links about FDR2

FDR-2 Jamboree at BNL 24-26 June (Agenda) contain talks on, among other things, DPD making, PROOF, trigger.

Further reports on FDR2 analyses can be found here and here.

Edit - History - Print - Recent Changes - Search
Page last modified on July 25, 2008, at 10:19 AM