SAM File Metadata

SAM File Metadata

  • Table of contents
  • [SAM File Metadata]
    • [Running list of comments]
      • [29 January 2018]
    • [Useful links]

Running list of comments

29 January 2018

(From Wes)

  • SCD is setting up a SAM station for us and they need to know what file metadata values we may want to add. See here for a list of the default parameters
  • Ideally this would be the same as SBND, but SBND has no special parameters yet.
  • Talking to Marc Mengel (mengel@fnal.gov), we can add new parameters at any time, they just will be 'unset' in older files. Values can be set on older files one-at-a-time, or using helper scripts based on other queries of the files.
  • Tracy and I reviewed the MicroBooNE metadata parameters with Herb Greenlee. See here for MicroBooNE's custom parameters. Ones we would likely like to keep:
    • The 'name', 'stage', and 'version' on the ub_project parameters. This is used for specifying names of production jobs, stages in which they were run, and the version of software used to run them. Note: if/when we update to POMS, we will want to make sure this matches well with that production system.
    • The 'name' and 'version' for the fcl parameters. Currently only 'name' is really used in MicroBooNE, but there's an option for having a distinct fcl version if desired (which may be nice).
    • The 'start_time_usec' and 'end_time_usec' are useful for providing finer time precision than the defaults in SAM (which are one second?).
    • The 'blinding' parameters, which can be used to mark files for read protection as an option for implementing a blindness procedure. The 'blind' parameter is used to determine whether a file is or isn't "blind", and the 'processed' parameter is used as a flag if a file has been checked to see if it should be blind.
  • Herb also suggested to think about trigger and run configuration parameters, as this is something MicroBooNE does not have but would make use of if it had thought it through from the beginning. This is probably more an 'online' thing to apply to data, but should be noted. (Also to note: file_type, file_format, data_tier, data_stream, and run_type can be used for organizing files in such ways.)