MpdRoot_KFParticle/physics/ebye/analysis_with_identity_method

Intro

Precise determination of the moments of multiplicity distributions of identified particles could be challenging due to the misidentification in detectors. The so-called Identity Method [1-3] allows one to solve this problem. Using the code placed in this folder, performance of the Identity Method was tested on the A–A events simulated in the conditions of the MPD experiment at NICA (using TPC dE/dx vs mometum histograms). Namely, moments within a single kinematic window as well as coefficients of forward-backward pseudorapidity correlations are extracted.

[1] M. Gazdzicki et al., Phys.Rev. C 83, 054907 (2011)

[2] M. Gorenstein, Phys.Rev. C 84, 024902 (2011)

[3] A. Rustamov, M. Gorenstein, Phys.Rev. C 86, 044906 (2012)

Table of content:

1. Analysis of MPD simulated data
2. Extraction of the special tree with tracks for identity method
3. Calculation of "true" moments, extraction of dE/dx distributions
4. Analysis with Identity Method
5. Plotting final results

The analysis consists of several steps. Note that at each step one needs to properly set paths, filenames, and other parameters inside the scripts.

1. Analysis of MPD simulated data

As the first step, trees with event- and track-level information are extracted from the DST root-files (that contain MC and reconstructed data) using the task get_tree_MPD_task.cc from the folder analysis_task. For that, DST files from a chosen directory (e.g. on NICA cluster) are merged into a chain (file names are taken from inputFilesByHand.txt). Execution of the analysis can be started using the bashRunJobs.sh script.

2. Extraction of the special tree with tracks for identity method

Obtained trees can be then processed using a script get_info_and_identity_tree.cxx in the analysis_of_output_tree folder. The script does several things: 1) it extracts a special tree with tracks for Identity Method (called "treeIM") - to be used below, 2) it also performs calculations for Forward-Backward correlations analysis at SIM, REC_PRIMARY and REC_ALL levels, 3) a lot of supplementary information is extracted as well.

3. Calculation of "true" moments, extraction of dE/dx distributions

After that, the script get_dEdx_vs_mom_and_true_pi_K_p_moments_from_tree.C runs over obtained an_results.root file in order to: 1) extract the dEdx vs pTPC distributions and put them in file_dEdx_in_bins.root (IMPORTANT: this is done with utilizing the PID info from simulations! With real data, this is not possible to do, and sophisticated fits of dE/dx distributions in narrow momentum slices will be required.) 2) calculate the true (i.e. primary-particle level) moments (first, second, cross-moments) and put them in file_truth_moments_IM.root.

4. Analysis with Identity Method

In the identity_analysis folder, the script fit_dEdx_pt_slices.C takes the file_dEdx_in_bins.root file and fit narrow momentum slices for each particle species with a function (i.e. Gaus). The results for all slices are then stored into fitted_dEdx_in_pt_bins.root.

Now everything is ready to run the analysis of the moments with the Identity Method itself. The main script is analyse_identity_dEdx.C, which requires the fitted_dEdx_in_pt_bins.root and file_dEdx_in_bins.root files, as well as the root-file an_results.root with the identity tree "treeIM" (which has a structure shown in the cartoon structure_tree_for_Identity_Method_analysis_cartoon.png). This code is based on the Identity module created by A. Rustamov and M. Arslandok, NIM A 946 (2019) 162622, https://arxiv.org/abs/1807.06370, and also a more advanced script provided by A.Rustamov. In order to load necessary libraries from the identity_analysis/src directory, the analysis should be started via the run.C script. The main output is a file_extracted_moments_IM.root file.

5. Plotting final results

Finally, truth moments from file_truth_moments_IM.root can be compared with those reconstructed with the Identity Method (file_extracted_moments_IM.root) by running a script draw_comparison_truth_IM.C in the draw_output folder.

A set of results obtained with this code can be found in Altsybeev, Petrov, J. Phys.: Conf. Ser. 1690 012124 (2020), https://arxiv.org/abs/2010.08742.