Fermilab


MINOS Offline Documentation
[home] [installing MINOS software] [supported platforms] [package links] [mailing list] [HyperNews] [FAQ] [computing help] [MINOS glossary] [archives]

Package Glossary




Table of Context

Introduction
Terminology

index:    A   B   C   D   E   F   G   H   I   J   K   L   M 
    N   O   P   Q   R   S   T   U   V   W   X   Y   Z 

Discussion




Introduction

Object Oriented Programming Analysis and Design books generally emphasize the need for a problem domain specific glossary. The purpose of such a glossary is to make fuzzy concepts more concrete and ensure that all participants are speaking the same language when they use the same words. Specificity can help to eliminate the confusion that arises when two people use the same term with subtly different meanings or nuances.

To that end, this glossary is expected to be an evolving document with additions as new terms come into use and individual terms have their meanings more narrowly focussed. The discussion section at the end is intended as a place where justifications for particulars can be made without cluttering the glossary itself.




Terminology

A


active detector
A component of the overall detector that gives measurable signals.

B


C


collection
A generic term for a group of objects. Typically these are further classified as: sets (no duplicates), bags (duplicates), and sequences (duplicates, ordered). Any collection object should at minimum support the following operations:
size
return the number of elements in the collection; for bag or sequence duplicates are counted as separate items.
iteration
loop over the items in the collection visiting each item once; for a sequence the ordering is significant.


coordinate system
proposal: we continue to support both a global (detector-level) coordinate system as well as a local coordinate system for individual detector elements (e.g. strips (logical vs. physical?)).
question: do we continue to use cm as our basic unit of length?
question: do we continue to use the tpos (transverse position) variable. Is this useful? Is it well defined. A related question is the definition of the rotations X,Y,U,V. Perhaps, U and V should be redefined to straddle the X rather than the Y orientation.

D


digitization or digit
An unit of data corresponding to one electronic channel's response. For the MINOS detector this is expected to consists of two parts: an ADC value representing the photoelectron count, and a timestamp. As the channels are expected to be self triggering, there needn't be a digit for every channel. The timestamp may consist of more than one part, e.g., the output of a fast clock may be combined with a coarse bucket number to produce a unique time. (cf. hit and discussion on the structure of digits below).

E


event
This is a term fraught with ambiguity; see discussion below.

F


flat file


flux file
An source of particles, generated by an external program, that serves as input to the simulation. These simulate a variety of physics sources.
  • neutrino flux
    • NuMI beam simulation
    • atmospheric
  • muon flux
    • comic rays
    • rock interactions (CC neutrino events)

question: what about correlated muons from atmospheric showers ... are these fluxes or inputted events?

G


H


hit
A hit, following GEANT parlance, is a collection of unknowable truth information recorded while tracking particles in the simulation. The hit information relates to a single particle in a single sensitive (or active) volume. Multiple hits in the same volume will be collected together and used to compute the digitization after the tracking is complete. (cf. digit)

I


J


K


L


M


module
An unfortunately overused term; overloading of this term has lead to ambiguity.
  • A short sequence of passive and active detector planes that repeats itself an integral number of times to form a supermodule. This usage drove the origin of the term supermodule, which has achieved common usage even if module in this context has not.
    [used by the simulation code gminos]
  • A unit steel plane and active detector plane.
    [used by the TDR, page 3-15]
  • A light-tight package of 20 or 28 scintillator strips that make up a active detector plane. These come in a variety of shapes (boxes and trapazoids) and sizes.
    [used by the scintillator group, cf. TDR page 5-1]
  • A logical module consists of those strips that map, due to clear fiber routing in the MUX boxes, to different pixels. This term is nonsensical for some optical summing schemes.
    [used by the scintillator group, cf. TDR page 5-5]


N


O


overlay event
Overlay or pileup events are the results of multiple interactions (beam neutrinos, through-going muons associated with the beam, cosmic ray muons and neutrino interactions) occurring in the same TimeSlice.

P


passive plane
A component of the overall detector that serves solely as target material and absorber.

pixel
The input to a uniquely identifiable electronics channel.
question: is this to be exactly a channel, or should cross-talk impose a distinction between the two?

Q


R


S


sensitive volume
The sensitive volume of an active detector is that region that can return a measurable signal based on the energy deposition of a particle traversing the volume.

SR
Standard Reconstruction

strip
A length of scintillating plastic.
question: what about "logical" vs. "physical" strips, i.e., does the term refer to aligned strips that are in two parts due to the break imposed by the coil, or is each strip a single unbroken piece of plastic.

supermodule
A sequence of passive and active detector planes that constitute a logical unit.
proposal: The baseline MINOS detectors each have two (see discussion below).

T


TimeSlice
Proposal: a term for a collection of digitizations within a time window (marked by a beginning and ending timestamp). In the simulation there should also be a mechanism for relating this window to the individual events that contributed to the list of digitizations, even those events that originated before the start of the TimeSlice but generated digits within the window due to propagation times.

timestamp
A datum associated with each digitization that uniquely identifies the time at which the information was recorded. The structure of this item must be fleshed out in conjunction with the DAQ software group, but it is expected to have two components: a finely segmented clock tick and a coarse time bucket.

track
This is a term, like event, fraught with ambiguity; see discussion below.

U


V


W


X


Y


Z





Discussion

What is an event?

This might seem to be a silly question, but actually it is quite subtle on several levels. Different definitions can come into play if one is talking about the simulation or the reconstruction levels.

The intuitive definition involves assigning to an event the properties of a single physics interaction. Furthermore one normally does not differentiate between the "hard" interaction (say the initial neutrino interaction) and the subsequent interactions of the daughters products (the shower development). Thus an "event" in such a usage would involve a particular neutrino flavor and 4-vector, the interaction (NC vs. CC), and resultant kinematics (outgoing lepton and fragmentation products).

While a hit can obviously be associated with a single unique event, the same is not true of a digit when one considers pileup or overlapping interactions. Similarly, are digits due to noise (photodetector, radioactivity, cosmic background) part of an event or not?


Proposal: adopt and use the term event to mean a physics interaction, and designate a separate distinct term for reference to a collection of digits within a time window. Such digits may or may not be logically related via a single event. During reconstruction one can then assign individual digits from a collection, say TimeSlice to an object, say RecoEvent. The digits in the data store are not owned by either TimeSlice nor RecoEvent which thus allows them to be used by multiple versions of each.

hits vs. digits

This bugaboo is a pet peeve of Robert's. When doing simulations one must clearly distinguish between the information recorded while tracking particles (hit) and the detector response (digit). The first contains information that is simply unknowable in the real world, while the second may be the result of multiple particles in the same sensitive volume. This choice of terminology follows that used by GEANT. Sloppiness in common usage (saying hits when one means digits) is usually decipherable, but will occasionally lead to temporary confusion. It would be better if all involved would attempt to stick to precise usage. Alternatively, if there is a strong upswelling for the use of hits as the term for data that is delivered out of the detector, then we must find an alternative term for the truth information.

What is an digit?

We need some discussion on what constitutes a digit. The GMINOS concept of a digit as realized by the FLSDigit structure. FLSDigit associates two ADC counts and two TDC values with each logical strip. This makes the assumption that the event is either demultiplexed or that ghosts digits occupy each of the 7+7 alternative strips (assuming non-zero values on each side).
Proposal: It is recommended that we define the base digit as the output of a single electronics channel, e.g. PixelDigit. These would be the data coming from the DAQ system. Tentative assignment to the physical strips via a demultiplexing algorithm can be collected into StripDigit objects, which would not have ghosts. In reference to the "What is an event?" discussion above, one model might have TimeSlice collections contain PixelDigits, while RecoEvent collections contain StripDigit objects.

What is a track?

Tracks, like the term event, has subtly different meaning depending on the context. The issue is further confused by the question of assigning digits (not hits) to tracks; like events a digit may truly belong to more than one track. This implies that any "track object" should not own the digits that make it up. ....needs more discussion

collections

As written, collections are a general term. When it is relevant one should use a more specific term (set, bag, sequence) for the particular collection type under discussion. It is expected that in the MINOS software most collections will be either sets or sequences. The choice has implications on what one can expect out of the iterator.

supermodule(s) in the near detector

The term supermodule has achieved a general usage and understanding in the context of the baseline MINOS far detector. In the near detector there remains an ambiguity. One can view the near detector in three different ways.
  1. A single unit. There is only one coil, thus only one supermodule.
  2. As four parts. This distinguishes between identically constructed regions based on their functional use.
    • veto region
    • target region
    • hadron shower region
    • muon spectrometer
  3. As two parts. This only distingishes between the forward region and the downstream spectrometer.

Proposal: It is recommended that we adopt the third usage, with the near detector having two supermodules. This allows flexibility in designating where the "target" region starts and ends, while retaining the distinction between the forward region and the spectrometer.


Last Modified: $Date: 2002/08/13 16:40:42 $
Contact: rhatcher@slac.stanford.edu
Page viewed from http://www-numi.fnal.gov/offline_software/srt_public_context/WebDocs/glossary.html
Fermilab
Security, Privacy, Legal Fermi National Accelerator Laboratory