ClusterMaker

Makes clusters from CellMaker Cells and store them on the TES

Clusters

Clusters are Atlfast physics output entities which correspond to a summation of Cell energies within a specified cone.

ClusterMaker

The ClusterMaker is an Athena-Algorithm which is created at the start of a job and executed once per event.
The following three public methods are executed by Athena, and are needed to honour the IAlgorithm interface

• initialise() - create helper classes and declare any properties supplied in the jobOptions file
• execute() - perform the algorithm's function once per event
• finalise() - perform any end of job tasks

1. Retrieve reconstructed Cells from the TES
2. Create fully formed Clusters from these Cells
3. Store the Clusters in the TES

ClusterMaker has a number of helper classes which it uses to perform these tasks, all of which are shown in the ClusterMaker class diagram.

• Precluster - used for accumulating Cells during candidate cluster building, prior to acceptance.
• KinematicHelper - used for generic kinematic operations on the kinematic interface which Cells implement. Used in this package for eta-phi separation calculations.
• TesIO - used for input/output to the TES

User Requirements Satisfied

Based on the functionality of CellMaker,
ClusterMaker makes Cluster to enable Jet construction.
It fulfills requirements 3.1.5 of the ARD.

ClusterMaker Detail

jobOptions Properties

jobOptions properties are steering parameters which can be set by the user in the jobOptions file.
They are declared in the constructor and initialised to Default Values.
The options for this algorithm are:

• RConeBarrel - Size of energy cone in barrel region
• RConeForward - Size of energy cone in forward region
• minInitiatorET - Minimum energy value of 'initiator' cell
• minClusterET - Minimum energy for cluster acceptance
• InputLocation - Location on TES of input Cells
• OutputLocation - Location on TES of output Clusters

Initialisation

The initialise method is called once at the beginning of the job.
It is here that the Pre-Cluster, KinematicHelper and Store classes are instantiated.
The initialisation process is shown in the
ClusterMaker initialisation sequence diagram.

Execution

The execute method is performed once per event and carries out the main operation of the algorithm.
The execution process is shown in the ClusterMaker execution sequence diagram.
The strategy for making clusters is as follows

• The collection of Cells is copied into two distinct, mutable, collections (all of pointers:


1. "available initiators"
2. "available cells"


• The algorithms sorts the "available initiators" into descending order of transverse energy.
• The highest energy available initiator is taken.

For accumulating Cells during candidate cluster building, prior to acceptance,
a nested class is used, Precluster.

• All "available cells" within a given cone are found and summed.
• If the resulting sum has a transverse energy exceeding some threshold, then a new Cluster is created using these Cells.
• In this case the cells are removed from the "available cell" collection;
  all cells are also removed from the "available initiator"  collection
• If the threshold cut is failed, then the initiator is simply removed from the "available initiator" collection, but no other action is taken and all cells are still available for further cluster formation around a new initiator.

Class Diagram

The various classes discussed above are shown in the class diagram.
ClusterMaker class diagram