ACT_PID

Repository to study particle identification (PID) using ACT charge distributions and tank data from WCTE readout windows.

This repository is targeted at ACT PID taskforce members (Bruno, Marie, Viet, Alie) and provides a lightweight analysis prototype for:

• selecting muons/pions/electrons from WCTE merged ROOT files,
• comparing ACT and tank-based taggers,
• estimating efficiency/purity from ACT charge histograms,
• producing PDF summaries per run.

Repository structure

• python/ACT_PID_analysis.py: driver script for one run; reads a ROOT file, instantiates ACTAnalysis, and prints run info.
• python/ACT_PID_module.py: core ACTAnalysis class; data loading + analysis method stubs + example high-momentum algorithm.
• output/: expected output location for per-run PDF plots (e.g. ACT_PID_run_<run>.pdf).

Requirements

• Python 3.8+
• numpy
• pandas
• matplotlib
• uproot

Install with:

pip install numpy pandas matplotlib uproot

(should not be needed on the CERN clusters)

Quick start

1. Set run_number in python/ACT_PID_analysis.py.
2. Confirm the ROOT path is valid (example path in the script): /eos/experiment/wcte/data/2025_commissioning/processed_offline_data/production_v1_0/<run>/WCTE_merged_production_R<run>.root
3. Create output directory if needed:

mkdir -p output

4. Run:

python3 python/ACT_PID_analysis.py

5. Expected output:
   • printed run info on console
   • currently placeholder for plots and save to output/ACT_PID_run_<run>.pdf from ACTAnalysis.load_data().

What the code does

python/ACT_PID_module.py

ACTAnalysis class:

• __init__(self, input_files): store input file names, init data structures, call load_data().
• load_data(self): for each input ROOT file:
  • read WCTEReadoutWindows tree with branches (temporary):
    • vme_evt_quality_bitmask
    • vme_digi_issues_bitmask
    • vme_act_tagger
    • vme_act_eveto
    • vme_mu_tag_l_charge
  • read vme_analysis_run_info tree into run metadata holding information about the run, the momentum and refractive index of the ACTs
  • open a PdfPages file ../output/ACT_PID_run_<run_number>.pdf per run
• get_run_info(self, run_number): return run metadata DataFrame matching run number.

Placeholder methods to implement:

• apply_ring_finding_algorithm()
• select_muon_sample_tank()
• plot_act35_charge()
• read_cut_lines()
• apply_cuts()
• plot_non_electron_non_proton_distribution()
• scale_and_plot_distributions()
• calculate_efficiency_purity()
• produce_likelihood_distributions()

high_momentum_analysis_example(self) implements an end-to-end histogram-based extraction of muon/pion components and efficiency/purity curves using ACT35 charge and muon tagger cuts. This is very dirty copy-paste from Alie's high momentum analysis, will have to be re-written cleanly.

python/ACT_PID_analysis.py

• sets run_number and file_name for one run
• builds ACTAnalysis(input_files=[file_name])
• prints run metadata
• currently has comments describing desired analysis pipeline steps (ring finding, 2D cuts, ACT02/ACT35 + TOF cuts, distribution scaling, likelihood calculation, comparisons)

notebooks/PID_with_WCTE_tank_Sahar.ipynb

• identifies the different particles based on the shape of the charge deposited inside the tank
• runs over the full dataset
• Applies a 6 ns time cut on all events (cleaner sample)

Analysis workflow (suggested implementation)

1. Add ring finding code into apply_ring_finding_algorithm() using tank charge data.
2. Implement muon selection in select_muon_sample_tank() (2D tank inside/outside cut).
3. Implement ACT35 charge plotting and distribution extraction.
4. Load ACT02 / proton TOF cuts in read_cut_lines() using VME beam analysis lines.
5. Apply filtering cuts in apply_cuts().
6. Compute separate distributions and scale in scale_and_plot_distributions().
7. Calculate efficiency/purity and likelihoods in calculate_efficiency_purity() and produce_likelihood_distributions().
8. Save all figures through PdfPages objects in self.pdfs.

Notes

• Keep output relative path consistent with module code (../output). Running script from ACT_PID/python may create ACT_PID/output via relative path.
• The code is currently a skeleton with key algorithm in high_momentum_analysis_example and should be refactored into modular methods.
• Add unit tests or analysis notebooks as needed for reproducibility.