A cluster algorithm for the study of jets in high energy physics

A cluster algorithm using angular correlations and leading particle effects is presented which is applicable to the study of jets produced in high energy storage ring collisions. The algorithm uses the concept of a minimal spanning tree and is computationally very efficient. Events are classified by their cluster number and the cluster number frequency distribution can be used for comparison with particle production models. Individual particles are assigned to the clusters and the vector sum of their momenta generate a cluster axis. These cluster properties permit the study of the dynamics of the jet production and fragmentation processes. The example of two and three jet production at PEP and PETRA energies is used to illustrate this technique.