Description

Centrality clusterization

Determination of the centrality classes by the clusterization of the forward detector signals (FHCal, hodoscope); the obtained probability density functions are used by CentralityClusterizer (reconstruction/centrality).

File / directory	Description
centrality_clusterization.py	Python script (uproot, constrained k-means from k_means_constrained) clustering the events by the selected features and saving the model and the PDFs, e.g. python3 centrality_clusterization.py -p <tree.root> -f fFHCal,fHodo -b 500 -n 10 -e 0.01 -m results/model.pkl
uproot-bmn-constr.ipynb	Jupyter notebook with the same analysis
results	Obtained models (.pkl), feature PDFs (feature_pdfs.root) and figures

Directory dependency graph for centrality: