Getting Started with LHCb Open Data

In 2022, LHCb has released the first 200 terabytes of the data via CERN OpenData Portal, making it available to the public.

• The data comes in .DST and .MDST format, the same format is used by LHCb internally.
• Every data set released is narrated by an "Open Data Record" accessible through Open Data Portal.
• Open Data Records contain various bits of information about the selected data set (this is called metadata). An example of the types of metadata provided in the record is:
• Number of events in the dataset
• Number of files in the dataset
• Combined size in TB of the dataset
• Production ID
• Production Type
• Detector conditions (condb, dddb tags)
• List of Trigger Configuration Keys (TCKs)
• Scripts used for each production step
• List of Logical File Names (LFNs) on LHCb DIRAC.

The metadata provided should help the user to navigate, select and work with with LHCb Open Data.

Index of files is accessible both via a GUI or as a machine readable file.

• Some instructions on how to use open data are pointed out in the records themselves.
• As well as the data records, an extensive list of LHCb stripping lines and their descriptions is provided as well.
• After selecting the desired stream, a stripping line description can be followed to obtain a number of cuts/conditions which could be used to filter the data further.
• Data can be accessed directly (eg. using xrootd protocol) or downloaded locally.
• It is suggested to further filter and categorize the data by writing out smaller data files in .root format (called ntuples).
• This is done in LHCb with the help of software called DaVinci.
• 'DaVinci' and other LHCb Software is available through CVMFS.
• Some initial instructions on working with DaVinci are provided in LHCb Starterkit web page.