CVFWeightTool

This tool calculates the Cluster Vertex Fraction (CVF) of clusters and applies a weight of 0 to the cluster depending on its CVF. It makes a shallow copy container of the clusters with the pT of the clusters modified appropriately.

Dependencies

This package makes use of UChicago's xAODAnaHelpers package and Giordon Stark's xAODJetReclustering package.

Installing

To install,

mkdir myRootCore && cd $_
rcSetup Base,2.3.41
git clone https://github.com/kratsg/xAODJetReclustering
git clone https://github.com/UCATLAS/xAODAnaHelpers
git clone https://github.com/AvivCukierman/CVFWeightTool
rc checkout_pkg atlasoff/AsgExternal/Asg_FastJet/tags
rc checkout_pkg atlasoff/AsgExternal/Asg_FastJetContrib/tags
rc find_packages
rc compile

(You might have to do a rcSetup -u && rcSetup before the final rc compile just because.)

Configurations

Using

Add this package as a dependency in cmt/Makefile.RootCore.

Add a header

#include <CVFWeightTool/CVFWeightTool.h>

Set up the tool in the initialize() portion of your algorithm as a pointer

  m_CVFTool = new CVFWeightTool(m_name);
  m_CVFTool->setProperty("InputContainer", m_InputContainer);
  m_CVFTool->setProperty("OutputContainer", m_OutputContainer);
  m_CVFTool->setProperty("TrackContainer", m_TrackContainer);
  m_CVFTool->setProperty("doLCWeights", m_doLCWeights);
  m_CVFTool->setProperty("PtThreshold", m_PtThreshold);
  m_CVFTool->setProperty("CVFThreshold", m_CVFThreshold);
  m_CVFTool->initialize();

And then simply call m_CVFTool->execute() in the execute() portion of your algorithm to fill the TStore with the appropriate container(s). Don't forget to delete the pointer when you're done.

if(m_CVFTool) delete m_CVFTool;

You can then do jet clustering as normal on the output cluster container.

N.B. The doLCWeights option indicates whether to apply LC weights (if true) or EM weights (if false) to clusters. The output clusters have their pTs modified in the given state. I.e. with doLCWeights set to true, only the LC pT of the new cluster collection is modified, and with doLCWeights set to false, only the EM pT of the new cluster collection is modified. In order to set the state of the output (or any) cluster collection, the CaloClusterChangeSignalStateList tool should be used in your algorithm, as follows:

#include "xAODCaloEvent/CaloClusterChangeSignalState.h"

 const xAOD::CaloClusterContainer*             new_clusters   (nullptr);
 if(evtStore()->retrieve(new_clusters,m_OutputContainer).isFailure()) Error(APP_NAME,"Could not retrieve the CVF cluster container");

 CaloClusterChangeSignalStateList stateHelperList;
 for(auto clust: *new_clusters){
   if(m_doLCWeights) stateHelperList.add(clust,xAOD::CaloCluster::State(1)); //default is calibrated but we can make it explicit anyway
   else stateHelperList.add(clust,xAOD::CaloCluster::State(0));
   //now clust->pt() will give CVF modified pT, and jet reconstruction will use CVF modified pT
 }

Note that as it behaves like an AsgTool, the functions setProperty() and initialize() have a return type StatusCode.

Track Extrapolation

Track extrapolation is currently only available in Athena. However, the performance of CVF is much better when using extrapolated tracks than when not. Because of this, in the Rootcore version, an estimate for the extrapolated track coordinates at the calorimeter entrance is given, and can be activated with the doExtrapEst option. Validation of the accuracy of this estimate can be found here and here.

Authors

• Aviv Cukierman

Acknowledgements

Giordon Stark, from whom I plagiarized this readme and also whose code I used for inspiration.