FDR-I and TAG update 03/20

Analysis selecting events using TAG DB

Get the event selection

Load your certificate in the Web browser. Open ELSSI Web page: https://atldbdev01.cern.ch/tagservices/tutorial/index.htm Follow these steps to obtain the selection, but feel free to have your own selection. Select the following query in the Create query tab:

1. In Temporal cut choose "7-Feb-2008: All runs"
2. In Streams check "fdr08_run1_StreamEgamma_o1_r12_t1 (119620 events)"
3. In Data quality choose "All data - ALL luminosity blocks"
4. In Trigger make no selection (leave it blank)
5. In Physics attributes chose "NLooseElectrons" from ELECTRONS, the option "greater than" and enter "1". Then click "Add Constraint"

Now select the Review query tab. Here you have to change the temporal constraint in the query to obtain

RunNumber=3070 and NLooseElectron>1

In the Perform query tab you can count, display and retrieve the results

1. In Count count the resulting events (1874?), then click "Back to Selection"
2. In Retrieve results control that stream and query are correct and click "Retrieve"

In the same tab you'll receive the Result, something like:

Results

Executing extraction...

Your results can be retrieved by right-clicking this link and selecting Save link as... 

If you have a CERN account, your results are available as /afs/cern.ch/atlas/maxidisk/d39/fdr08_run1_StreamEgamma_o1_r12_t1_1205945712_user.root
This file is in a scratch area that may be purged periodically without warning.

You can download the ROOT collection with your event selection:

• using the URL in link
• copying it from AFS

Run the skim locally

You will have to setup ATLAS rel 13.0.40 and have in the run directory the TAG file (root collection with the events) downloaded from ELSSI and a PoolFileCatalog.xml with all the AODs you will need.

Run the skim with Pathena

Setup

 > . /osg/app/atlas_app/atlas_rel/13.0.40/cmtsite/setup.sh -tag=13.0.40.3,AtlasProduction
 > . /osg/app/atlas_app/atlas_rel/13.0.40/AtlasProduction/13.0.40.3/AtlasProductionRunTime/cmt/setup.sh
 > export CMTPATH=`pwd`:${CMTPATH}
 > export CVSROOT=:ext:mambelli@atlas-sw.cern.ch:/atlascvs;export CVS_RSH=ssh;export PATHENA_GRID_SETUP_SH=/share/osg-client/setup.sh 
 > cmt co PhysicsAnalysis/DistributedAnalysis/PandaTools;
...
 > pushd PhysicsAnalysis/DistributedAnalysis/PandaTools/cmt/
/ecache/marco/wdir_pathena/PhysicsAnalysis/DistributedAnalysis/PandaTools/cmt /ecache/marco/wdir_pathena
 > . ./setup.sh 
 > make
...
 > popd
/ecache/marco/wdir_pathena

Run with unpatched Pathena for reference. Until Pathena is modified you can skip this.

 > cd tag_pathena/
 > ls
AnalysisSkeleton_topOptions.py  fdr08_run1_StreamEgamma_o1_r12_t1_1205945712_user.root  mytags.pool.root
 > echo "mytags.pool.root" > filelist.txt
 > pathena -v --inputFileList=filelist.txt --shipInput --site=ANALY_MWT2 --outDS=user.MarcoMambelli.test.tag.080319._01 AnalysisSkeleton_topOptions.py
...
(0, "(lp1\n(L8607991L\nL24L\nS'800f9261-d021-4a4d-8c4c-3514fe08a151'\np2\ntp3\na(L8607992L\nL24L\nS'133472d0-660a-4998-8dc7-1e0f4389c7b6'\np4\ntp5\na.")
===================
 JobID  : 24
 Status : 0
  > build
    PandaID=8607991
  > run
    PandaID=8607992

Get the patched extPoolRefs.C file and compile patched Pathena.

The jobOption file is attached as well if you like to run the same analysis: AnalysisSkeleton_topOptions.py.txt - remove .txt from the name.

 > ls
AnalysisSkeleton_topOptions.py  extPoolRefs.C  fdr08_run1_StreamEgamma_o1_r12_t1_1205945712_user.root  filelist.txt  mytags.pool.root
> pwd
/ecache/marco/wdir_pathena/tag_pathena
> pushd ../PhysicsAnalysis/DistributedAnalysis/PandaTools/share/
/ecache/marco/wdir_pathena/PhysicsAnalysis/DistributedAnalysis/PandaTools/share /ecache/marco/wdir_pathena/tag_pathena
> mv extPoolRefs.C extPoolRefs.C.orig
> cp /ecache/marco/wdir_pathena/tag_pathena/extPoolRefs.C ./
> popd
/ecache/marco/wdir_pathena/tag_pathena
> pushd ../PhysicsAnalysis/DistributedAnalysis/PandaTools/cmt/
/ecache/marco/wdir_pathena/PhysicsAnalysis/DistributedAnalysis/PandaTools/cmt /ecache/marco/wdir_pathena/tag_pathena
> make
...
 all ok.
> popd
/ecache/marco/wdir_pathena/tag_pathena

Pathena submission. For redundancy I sent the same job both to UChicago and BNL.

> pathena -v --inputFileList=filelist.txt --shipInput --site=ANALY_MWT2 --outDS=user.MarcoMambelli.test.tag.080319._03 AnalysisSkeleton_topOptions.py
...
===================
 JobID  : 26
 Status : 0
  > build
    PandaID=8608013
  > run
    PandaID=8608014
> pathena -v --inputFileList=filelist.txt --shipInput --outDS=user.MarcoMambelli.test.tag.080319._04 AnalysisSkeleton_topOptions.py
...
===================
 JobID  : 27
 Status : 0
  > build
    PandaID=8608015
  > run
    PandaID=8608016

Update on ELSSI and Pathena

ELSSI

• use the tutorial version: https://atldbdev01.cern.ch/tagservices/tutorial/index.htm
• stream selection has been cleaned up: there are 4 valid streams (and only 4 are displayed)
  • triggers show in the trigger tab
  • the missing stream table has ben added

Pathena

• file name problem (removal of .root from file name, file was not found) has been fixed
• patch to select StreamAOD instead of default RefName is attached ( extPoolRefs.C). A more robust solution will be included in Pathena

FAQ/Help/Caveats

ELSSI

• support email is hn-atlas-physicsMetadata@cern.ch
• if you can select, count the events and the extraction works but you cannt download/find the root collection send n email to hn-atlas-physicsMetadata@cern.ch (we may need to restert the selection server)
• if you get an Oracle error, send your stream+query to hn-atlas-physicsMetadata@cern.ch. There may be error with the query or DB problems

-- MarcoMambelli - 19 Mar 2008