This page describes the original cluster finder algorithm. In order to use this algorithm you should set with the algorithm kEmcClOld.

Salient features of the method implemented in the program are,

• Clustering have been performed for each sub detector separately.
• Currently clusters are found for each module in the sub detectors. There are some specific reasons for adopting this approach especially for Shower Max Detectors (SMD's). For towers, we are still discussing how it should be implemented properly. We have tried to give some evaluation results for this cluster finder.
• There are some parameters used in the code with their default values. These default values are obtained after preliminary evaluation, but for very specific study it might be necessary to change these parameters. 
• The output is written in StEvent format.

Cluster algorithm

• Performs clustering module by module
• Loops over hits for each sub detector module
• Looks for local maximums

Cluster parameters

• mEnergySeed – minimum hit energy to start looking for a cluster
• mEnergyAdd -- minimum hit energy to consider the hit part of a cluster
• mSizeMax – maximum size of a cluster
• mEnergyThresholdAll – minimum hit energy a cluster should have in order to be saved

Neighborhood criterion

Because of the difference in dimension and of readout pattern in different sub detectors, we need to adopt different criterion for obtaining the members of the clusters.

• BEMC: Tower gets added to the existing cluster if it is adjacent to the seed. Maximum number of towers in a cluster is governed by the parameter mSizeMax. It should be noted that BEMC, which takes tower as unit is 2-dimensional detector and by adjacent tower, it includes the immediate neighbors in eta and in phi.
• BSMD: As SMDs are basically one-dimensional detector, so the neighborhood criterion is given by the condition that for a strip to be a neighbor, it has to be adjacent to any of the existing members of the clusters. Here also maximum number of strips in a cluster is governed by mSizeMax parameter.

Cluster Object

After obtaining the clusters, following properties are obtained for each cluster and they are used as the members of the cluster object.

• Cluster Energy (Total energy of the cluster member).
• Eta cluster (Mean eta position of the cluster).
• Phi cluster (Mean phi position of the cluster).
• Sigma eta, Sigma phi (widths of the cluster in eta nd in phi).
• Hits (Members of the cluster).

Some Plots

BSMDE clusters for single photon events