"Higgs from Z to A" Subgroup @ Nikhef


Members: Tristan du Pree (PI) , Hannah Arnold (PostDoc) , Marko Stamenkovic (PhD Student), Brian Moser (PhD Student)

In 2012 the Higgs boson was discovered, at the CERN LHC accelerator. In this research, to be conducted at the ATLAS experiment, physicists will use Higgs bosons to study its properties with precision measurements, to search for other scalar particles, and to shine a light on Dark Matter (NWO Vidi abstract)

Webpage: here

Navigation:


Important News/Notes

This is the prototype of the new subgroup twiki - feel free to add/change/manipulate whatever you consider handy or nice to have.

Talks of group members

Here is a list of talks given by the group members in ATLAS meetings:

Marko's talks: here

Brian's talks: here

Meeting Times

Day Meeting + Time
Monday 09.00 VHcc meeting
13.00 b-tagging calibration
Tuesday 14.00 ATLAS weekly
15.30 b-tagging plenary
16.30 CxAOD Framework
Wednesday 13.00 Hbb group meeting
Thursday 10.00 Nikhef coffee meeting
13.00 VHbb resolved
13.00 b-tagging algorithms
Friday 10.00 Nikhef group meeting
11.00 Nikhef Colloquia
14.30 VHbb boosted

Links

Meeting contributions

Here you can find links to summary websites for several meetings including links to the last contributions. These sites are updated automatically at 12am, 2pm, 4pm and 6pm every day.

Todays ATLAS Meetings

https://atlas.web.cern.ch/Atlas/Collaboration/private/atlas-meetings.html

CxAOD Framework - Tips and Tricks

How to run the framework for new login on lxplus

On lxplus, the description on which variables to set before being able to run after loggin in again seem not to work. Instead of simply typing

ssh -Y lxplus.cern.ch
cd CxAODFramework_dev-R21_1
cd build
setupATLAS
asetup 
source x86_64-slc6-gcc62-opt/setup.sh 

as mentioned in the Twiki it seems to work only if you explicitly set AnalysisBase as

ssh -Y lxplus.cern.ch
cd CxAODFramework_dev-R21_1
cd build
setupATLAS
asetup AnalysisBase,$release,here
source x86_64-slc6-gcc62-opt/setup.sh 

where $release specifies the AnalysisBase release, e.g. 21.2.12 for the dev branch.

Edit: Perhaps someone that uses the Framework as well can confirm?

How to checkout athena code locally and compile/run the framework with it using ACM

This page moved to the official Framework Release 21 TWiki, see https://twiki.cern.ch/twiki/bin/view/AtlasProtected/Release21Migration#How_to_checkout_athena_code_loca

Working with cmake

In order to get the framework to run using cmake the following tipps might be of use. Suppose you are in the following directory

$: ls
source build run

you can follow the first set of commands to setup the ATLAS environment and the ATLAS version of cmake itself

cd source
setupATLAS
asetup AnalysisBase, <your release>, here

In order to compile the code you have to switch to the build directory, tell cmake where to build the binaries from and then you can compile

cd ../build 
cmake ../source
cmake --build .

Afterwards you need to setup the run environment for the code, which you can do by souring the following setup (in the build directory)

source */setup.sh

This only has to be done the first time. If you change something in the code and afterwards recompile, you don't need to source again.

A few last remarks: Every time you want to compile using the above command, make sure that you are in the build directory. If you add packages or change the internal structure of the framework or change the AnalysisBase release (everything that is not only changing code in already existing files), it is safe to also redo (int the build directory) the

cmake ../source

Recent Talks in the Framework Meeting

https://atlas-hbb-framework.web.cern.ch/atlas-hbb-framework/recent_meetings.html

CxAODFramework on Stoomboot

This section briefly sketches how to run the CxAODFramework on the Nikhef batch system Stoomboot. For general information about the system, see https://wiki.nikhef.nl/ct/Stoomboot_cluster. The workflow looks as follows: CxAODs are stored on dCache that is a read only file system. This means you can save things there but not edit them in any way. The VHbb CxAODs are stored at

 /dcache/atlas/higgs/bmoser/VHbbSamples/HIGG*D*_13Tev

where HIGG5D1: 0 lepton
HIGG5D2: 1 lepton
HIGG2D4: 2 lepton

The CxAODFramework should be checked out e.g. on

/project/atlas/Users/<your-directory>

because it is compilable code and therefore the content needs to be modifiable. For a selection of available storage systems, see e.g. https://indico.nikhef.nl/event/997/contribution/3/material/slides/0.pdf. If you run an older version of the framework you need to add the following lines in

/FrameworkExe/util/hsg5frameworkReadCxAOD.cxx
<the-first-part-of-the-file-up-to-the-driver-selection>....} else if (driver=="torque"){
    EL::TorqueDriver* eldriver = new EL::TorqueDriver;
    eldriver->shellInit = "export ATLAS_LOCAL_ROOT_BASE=/cvmfs/atlas.cern.ch/repo/ATLASLocalRootBase && source ${ATLAS_LOCAL_ROOT_BASE}/user/atlasLocalSetup.sh";    
    std::string stbc_queue = "generic";
    config->getif<std::string>("stbc_queue",stbc_queue);
    job.options()->setString (EL::Job::optSubmitFlags, ("-q "+ stbc_queue).c_str()); //generic queue default - 24h walltime
    eldriver -> submitOnly(job,submitDir); //I think this is returning control after the submission
 <up-to-here-things-are-new> }else {
    Error("hsg5framework", "Unknown driver '%s'", driver.c_str());
    return 0;
  }
  return 0;
} <file-ends-here>

In addition, you can now set the following config flags in

/FrameworkExe/data/framework-read.cfg
#torque batch (Nikhef)                                                                                                                                        
string driver                   = torque
string stbc_queue               = generic

How to copy files from /eos/ to Stoomboot /dcache/ in a smart way

An exemplary script that shows how to copy files from /eos/ onto a local file system using xrdcp with multiple cores can be found at

/user/brianm/CopyFiles_dCache/copy_from_eos_1lep_data.sh

The file list needed can be created using the script

/user/brianm/CopyFiles_dCache/scriptForPaths.sh

A few torque commands

To see your currently running jobs, type

qstat -n1 | grep '<my-user-ID>'

To cancel your jobs, type

qdel <the-job-PID>

The PID is displayed as well when you type in the first command...

Boosted SM Vhbb 1L reader development

How to download the development branch reader

The development of the SM boosted Vhbb analysis in the 1 lepton channel within the CxAODReader _VHbb takes place in a special branch until it has been shown that it works as expected and does not alter the resolved analysis. The following section describes how to check out the master branch of the framework in all packages but the CxAODReader _VHbb. The development branch there is called master-SM-boosted-development-1L .

It is easiest to use the script that can be found in

/afs/cern.ch/work/b/bmoser/public/getMaster_BoostedDev_1L.sh

to check out the framework. All credit goes to Adrian, I only changed the checkout branch for the CxAODReader _VHbb to the above mentioned branch. The script needs the following arguments


#1: branch name, in our case origin/master


#2: name of the folder that should be created


#3: do_packages, use 1


#4: do_build, use 1 if you also want to compile the code, 0 instead.

A exemplary command could look as follows:

source getMaster_BoostedDev_1L.sh origin/master CxAODFramework_Boosted1LReaderDev 1 1

Please note that if your USER variable does not match your lxplus user name, you need to set the CERN_USER variable accordingly in order to be able to completely copy all relevant files.

How to develop

The development of the branch should look as follows: The script automatically creates the local branch master-boosted-reader-development-$CERN_USER on which you can develop. If you are satisfied with the changes, push the branch to origin and create a merge request into the boosted development branch (not the master).

See as well https://twiki.cern.ch/twiki/bin/view/AtlasProtected/Release21Migration#An_example_of_a_developing_Frame for further informations.

Storage systems on the grid

Permissions and user quotas

  • Regular users are only permitted to upload data directly to SCRATCHDISK. Data on SCRATCHDISK has a lifetime of 15 days.

  • Using Rucio rules, regular users should be able to transfer data to LOCALGROUPDISK.

  • Quotas determine how much data you can put on an RSE and are also a way of enforcing permissions. Quotas are allocated in the following manner:

    • SCRATCHDISK: every user has a quota of 50% of each SCRATCHDISK RSE. This is to avoid one user filling all the space on an RSE.

    • LOCALGROUPDISK: users who are registered in the VOMS group /atlas/xx have 95% quota on all LOCALGROUPDISK RSEs in country xx.

Miscellaneous

Very nice introduction to SSH and encryption: see here

Edit | Attach | Watch | Print version | History: r23 < r22 < r21 < r20 < r19 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r23 - 2019-11-10 - BrianMoser
 
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    Main All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2019 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback