Skip to content

Latest commit

 

History

History
127 lines (89 loc) · 5.33 KB

README.md

File metadata and controls

127 lines (89 loc) · 5.33 KB
Raw

Jet-hadron analysis

Build Status

This repository contains jet-hadron analysis code, run macros, tools, etc. Most are predominately documented through the docstrings in the code itself.

Quick start

First, setup your environment as you would like (virtualenv, alibuild, etc). This package requires python 3.6 and above. First, a few prerequisites are required which unfortunately cannot be resolved solely by pip because of the packaging details of probfit (which is depended on by reaction_plane_fit).

$ pip install numpy cython

Then the analysis package can be installed as normal with:

$ pip install -e .

Running the tests

Tests are implemented using pytest. They should be run via

$ pytest -l --cov=jetH --cov-report html --cov-branch . --durations=5

Coverage can be measured by adding the argument --cov=module (for example, --cov=JetHConfig). When the Jet-H analysis code is entirely packaged up, the module can just be jetHAnalysis, and it will test everything.

Development

You're now ready to develop.

The code is structured as:

  • Main analysis code in the jet_hadron/analysis package:
    • Correlations analysis in the correlations module.
    • ALICE jet-hadron response matrix in the response_matrix module.
    • STAR jet-hadron response matrix from simulated data in the STAR_response_matrix module.
      • This is based on the ALICE response matrix classes.
    • A number of supporting modules.
    • A comprehensive list of executables can be found in the entry_points of setup.py.
  • Plotting code is predominately in the jet_hadron/plot package. - PlotBase contains shared plotting functions - The other modules are the main plotting modules, split up by functionality.
  • Event generator steering in jet_hadron/event_gen.
    • Includes simulating PYTHIA 6 for STAR at particle and detector level (when provided the appropriate input files).
    • PYTHIA 8 is also supported.
  • Toy models tests are in jet_hadron/toy_models.

Most executables take a uniform set of arguments consisting of a YAML configuration file, along with the parameters of the analysis (collision system, energy, event activity, and leading hadron bias). The flags can be accessed through the --help flag. A YAML config file that stores most of the configuration options is in the config directory.

For example, the jet-hadron correlations analysis can be run with:

$ jetHCorrelations -c config/analysisConfigDev.yaml -e 5.02 -s PbPb -a semi_central -b track

This would analyze semi-central Pb--Pb collisions at 5.02 TeV with a leading track bias as defined in the configuration file. Note that the location of input and output files are specified in the YAML configuration.

In general, modules in the analysis, toy_models, and event_gen packages are executable with python -m ... if they don't have a dedicated executable installed by the package.

The package versions used for my thesis are stored in requirements-thesis.txt.

Running the correlations analysis

When using the command above, the following steps are necessary:

  1. Run the mixed event scale uncertainty with mixedEventSystematics -c .... This requires Z vertex binning to be available in the THnSparse outputs (as it should be for any analyses going forward).
  2. Run correlations analysis. If the mixed event scale uncertainty hasn't been run yet, it will stop when it's needed.

Pre-commit checks

It is strongly recommended to install pre-commit (from pip or elsewhere) and then run

$ pre-commit install

to utilize the git pre-commit checks. They will be run automatically with each commit to help ensure code quality. (They can also be run manually with pre-commit run).

Tools

Highlight Reaction Plane Fit Regions

It is useful to be able to highlight regions of a 2D surface plot to show where the Reaction Plane Fit (RPF) is actually fitting. Code to create this plot is in PlotRPFRegions.py. In short, it modifies the colors the of the plot in the regions that we want to highlight.

To execute it, the user must specify the root file and the histogram name in that root file. It should be something like

$ plotRPFRegions -f output/plotting/PbPb/3360/dev/Track/PbPb_correlations.root -i "jetHDEtaDPhi_jetPt1_trackPt4_corr"

All command line options are available with the -h option. While the Jet-hadron plot is integrated into the plotting code, others should look at the code in PlotRPFRegions. In particularly, look at the standalone functions located at the bottom of the file. These specify a variety of additional options that would be too cumbersome to pass from the command line.

Some options include:

  • Changing highlighted colors and ranges in the function defineHighlightRegions()
    • Can also set them to use transparency, so the underlying plot will show through. NOTE: This is not a perfect effect, but still probably worth taking a look at
  • Changing color scheme (by default it is the ROOT6 kBird color scheme) in the function plotRPFRegions()
  • Modify labels, etc in the function plotRPFRegions()

Note that it is best to execute it in the jetH folder, as a number of other modules are utilized and would also need to be copied!