LAST MODIFIED: 3rd September 1999

Glossary of LHCb software terms

This glossary is not intended to be complete, it will evolve as needs arise

0. Related information

Glossary of terms, LCB Software Architecture Study Group

Glossary of OO Terminology, Advanced Object Group

1. Glossary of data types

MonteCarlo truth.

Consists of:

MC Kinematics

The output of the event generator, before tracking through the detector (e.g. KINE+VERT banks), plus secondary tracks created in the detector via material interactions or decay in flight

MC Hits

Simulated hits resulting from tracking the kinematics particles through the active regions of the detector. This is the input to the digitisation step of the simulation.

 

Raw data

Can be real or simulated:

Real raw data

is the data format at the output of the data acquisition system (e.g. ADC counts) and is the input to the higher level (software) triggers and the coordinate reconstruction.

Simulated raw data

has identical format to the real raw data, with the addition of links to the MonteCarlo truth. It is produced in the digitisation step of the simulation by applying a detector response function.

 

Reconstructed data

Consists of:

Reconstructed coordinates

These are derived from the raw data by applying an apparatus function (such as a drift time-distance relation), using appropriate calibration and/or alignment constants. They are the input to the pattern recognition. They may be produced, with different levels of refinement, by the detector front ends, the software triggers or the reconstruction program. They contain links to the raw data.

In the simulation, reconstructed coordinates be produced directly from the MonteCarlo truth. This can be the case in a fast MonteCarlo, where the digitisation step is skipped, but is forbidden in the complete detector simulation.

Reconstruction objects

This is the output of the reconstruction program and consists of physics quantities such as tracks and clusters, plus links to the coordinates.

 

Analysis objects

These are a different view of the reconstructed data, for use in the analysis. They may be a condensed form of the reconstruction objects (e.g. tracks defined only at the origin, with no reference to coordinates, but with some kind of association to the original MC Truth track) or a different abstraction (e.g. "tracks" could include also neutral calorimeter objects)

 

Event tag

A fixed length, contiguous, summary of the event, intended to optimise the preselection of events. It consists of variables that can be queried by an event selector. There may be many event tags per event (raw tag, reconstruction tag, private tag etc.).

The event tag is analogous to an Aleph event directory entry.

 

Event collection

A container of event tags. Our current understanding is that the event tags are physically contained in an event collection. Deletion of the event collection implies deletion of the contained tags. This also implies that there may be several copies of a given tag, if there are overlapping event collections. Examples of event collections are a "run" or a "physics group selection"

This type of event collection is analogous to an Aleph event directory file.

An alternative implementation would be that event collections contain references to event tags.

 

Event summary

Part of the event data model, containing often accessed information about the event, and intended to optimise data access within an event: most analyses should access only the event summary section of the majority of preselected events.

Analogous to an Aleph MINI-DST event

2. Glossary of GAUDI framework terms

Option

An option consists of a new set of values for a property of an algorithm

 

Property

A property is an attribute of an algorithm, visible from the outside world, whose values can be altered to modify the default behaviour of the algorithm

 

Event selector

A piece of code which, given an event collection and some selection criteria, returns references to events that satisfy the criteria.

 

Selection criteria

A set of requirements that an event tag must satisfy in order for an event to be selected. One possible implementation is an SQL query on the event tag variables.