Eventdisplay Release Tests

Introduction

Releases of Eventdisplay are required to pass a list of testing procedures before being tagged for science analysis.

Testing a release needs time and should cover large parts of the parameter space (e.g., zenith angle, epochs and atmospheres, NSB range) Note that it is impossible to test every single IRF function and every single possible science case.

This repository contains code and macros for release testing. Results of release tests are saved in separate repositories:

• v492
• v491
• v490
• before v490

Simplified overview of the testing (somewhat outdated):

Abbreviations

• SZE = small zenith angles (0-40deg)
• MZE = medium zenith angles (40-50 deg)
• LZE = large zenith angles (50-70 deg)

Directory structure

• macros and scripts: ./release_tests/
• most important scripts area:
  • data-MC comparison in ./release_tests/montecarlo/mc_data_comparison
  • Crab analysis in release_tests/sources/Crab
• steering of macros and scripts is done using run parameter files (see example v490/V6.runparameter.dat)

Testing the release of a new Eventdisplay version

The following points are required for a full release tests, including consistency tests for the instrument response functions

Checklist for release

• check that version number is correct: ./bin/evndisp --version (needs to be adjusted in Makefile and src/VGlobalRunParameter.cpp)
• compilation of all steps with warnings and errors
  • linux
  • OS X
• for releases with many changes:
  • use valgrind to check for memory leaks
    • e.g., valgrind --log-file=valgrind.evndisp.log --suppressions=$ROOTSYS/etc/valgrind-root.supp ./bin/evndisp <command line parameters>
• use display to look through a number (>10) of data and MC events
  • do images look reasonable (e.g. ellipses?, islands?, etc)
  • location of FADC integration window (check high and low-gain channels)
  • pedestal level correct?
  • for MC: check direction and core reconstruction (the true position is plotted also in the display)
  • check other tabs: pedestal, pedestal variations, gains, etc.
• MC / data comparison (see below)
• Instrument response functions tests (see below)
• Tests with known sources
  • Crab Nebula
  • other sources for soft spectra, hard spectra, ...
• prepare wiki page with new release information

MC / data comparison

Comparison of distributions obtained from Crab observations (on-off) and simulated gamma rays. Verifies that MC model describes sufficiently well the data.

Input:

• mscw_energy files from Crab observations
• mscw_energy files from MC simulations

Code:

• release_tests/montecarlo/mc_data_comparison

Completeness of IRFs

Check for successful lookup table merging:

cd release_tests/lookuptable-testing
./test_tables.sh

Correction factors and NSB

Checks that correction factors are correctly applied.

Input:

• mscw_energy files from MC simulations (note: usually these files are not on disk)
• correction factors from $VERITAS_EVNDISP_AUX

Code:

• release_tests/nsb

Instrument response functions

Compare effective areas at fixed energies and energy thresholds. Code can be found in release_tests/montecarlo/energythresholds.

Examples e.g., in ./v483b/energytresholds/

• energy thresholds
• effective areas at 300 GeV, 500 GeV, 1 TeV

Compare effective areas between different epochs. (not completely implemented yet)

Crab Nebula

• light curves (flux vs MJD)
• SZE and LZE spectra
• spectra per major epoch (V4, V5, V6)
• spectra per V6 minor epoch (V6_2013, V6_2014, ...)

Production scripts

Radial acceptances

Calculate and test radial acceptance files.