Introduction

This page contains instructions for generating GCD files for use in simulation production. We no longer support MJD-based GCD files and switched to run-based GCDs when I3OmDb was decommissioned in favor of I3Live (MongoDB).

  • 2020 - RunID = 134142

  • 2019 - RunID = 132868

Generation

  • Run the script sim-services/resources/gcd_validation/generate_gcd.py –season=2020 .

This script pulls from the DB, makes all in the corrections needed, tests, and validates the resulting GCD file as well, all in one go.

The output :

  • GeoCalibDetectorStatus_{YEAR}.Run{DAQ Run from GCD server}.{Template version}.{Pass2, Pass1)}_V{version ID}.i3.gz - This is the GCD file.

  • GeoCalibDetectorStatus_{YEAR}.Run{DAQ Run from GCD server}.{Template version}.{Pass2, Pass1)}_V{version ID}.log - The logfile

Details

The generation script runs several scripts, all of which can be found in sim-services/resources/gcd_validation/details/.

  • generate_gcd_snapshot.py - This pulls the data from the I3OmDb database.

  • correct_GCD.py

  • PMT Thresholds - Correct negative thresholds.

  • RDE (Relative DOM Efficiencies) - Set NaN RDEs to either 1.35(1.0) for High(Low) QE DOMs.

  • Low Noise DOMs - For DOMs with noise rates below 400Hz add 1kHz.

  • Clean out any triggers that aren’t InIce or IceTop or of the following types

  • SIMPLE_MULTIPLICITY

  • STRING

  • VOLUME

  • SLOW_PARTICLE

  • Remove any AMANDA OMs.

  • Inject SPE Templates.

  • Rescale noise parameters according the SPE charge distribution.

  • test_GCD.py - Makes a quick pass over the GCD file and makes some simple checks. It runs the following test (found in their respective projects).

  • vuvuzela_test

  • photon_propagators_test

  • pmt_response_sim_test

  • dom_launcher_test

  • topsimulator_test

  • trigger_sim_test

  • generate_IC86_stress_test_samples.py - Injects 20 PEs in each DOM spread uniformly over a 2 microsecond window.

  • validate_stress_test_samples.py - Checks the file that was generated by the previous script.

  • Verifies that the number of PEs is about what’s expected (~20). Currently the range is pretty large and only flags an error message if the charge is less than 20% of expected or greater than a factor of 2. We should be able to do better.

  • Checks that there are no LC bits set for SLC-only DOMs.

  • Checks that good DOMs do have hits, as expected.

Often many of the DOMs above that are flagged as “good” but produced no hits, really are bad DOMs and should be added to the bad DOM list. To be sure check out Wiki: Problem_DOMs and verify they are in fact bad. Please report this to the simulation group as well.