|
|
KtJet User Guide v1-05
More information can be found in the
KtJet Paper and examples of using KtJet
are given in Examples.
Introduction
Performing KT-clustering using the KtJet package is extremely
simple, it basically involves instantiating a KtEvent object with a std::vector of
input "particles" and a number of flags to set the KtEvent parameters. The input
"particles" can be in the form of either HepLorentzVectors or KtLorentzVectors. The KtEvent objects can then be
interrogated about the event using its simple public interface.
KtEvent
KtEvent can be instantiated in inclusive,
exclusive or sub-jet mode. The inclusive mode has the following constructors:
KtEvent(const std::vector<KtLorentzVector> &,
int type, int angle, int recom, KtFloat rparameter);KtEvent(const std::vector<HepLorentzVector> &,
int type, int angle, int recom, KtFloat rparameter);KtEvent(const std::vector<KtLorentzVector> &,
int type, KtDistance *, KtRecom *, KtFloat rparameter);KtEvent(const std::vector<KtLorentzVector> &,
int type, KtDistance *, int recom, KtFloat rparameter);KtEvent(const std::vector<KtLorentzVector> &,
int type, int angle, KtRecom *, KtFloat rparameter);
The first two of the constructors are the simplest to understand. The
user passes KtEvent a std::vector
of input LorentzVectors, three integer flags and the rparameter. The
rparameter is of type KtFloat which
is simply a typedef which defaults to float (but can be changed to
double as a compiler option).
The three integer flags are: type,
which sets the collision type; angle, which sets the distance scheme;
recom, which sets the recombination scheme. A brief overview of the
available options for each of these is given in table 1.0. A full description is
given in the KtJet Paper.
| type |
angle |
recom |
|---|
| 1 | ee | 1 | angular | 1 | E |
|---|
| 2 | eP | 2 | delta R | 2 | Pt |
|---|
| 3 | Pe | 3 | QCD | 3 | Pt2 |
|---|
| 4 | PP | | | 4 | Et |
|---|
| | | | 5 | Et2 |
|---|
Table 1.0: KtJet Constructor flags
The other three constructors are much the same as the first two
except they allow the passing of pointers to user defined
recombination and distance schemes rather than integer flags. Defining
your own schemes is discussed in further detail here.
Although KtEvent can take either
HepLorentzVectors or KtLorentzVectors in its
constructor, it is recommended that KtLorentzVectors are used as
this gives greater functionality when interrogating the instantiated
KtEvent object. Further information on using KtLorentzVectors can be found
here.
Once that the KtEvent has been
constructed in inclusive mode the final state jets are fully
defined. They may be recovered from the KtEvent using the following
methods:
std::vector<KtLorentzVector> getJets(); |
Return final state jets without sorting |
std::vector<KtLorentzVector> getJetsE(); |
Return jets in order of decreasing E |
std::vector<KtLorentzVector> getJetsEt(); |
Return final state jets in order of decreasing Et |
std::vector<KtLorentzVector> getJetsPt(); |
Return final state jets in order of decreasing Pt |
std::vector<KtLorentzVector> getJetsRapidity(); |
Return final state jets in order of decreasing rapidity |
std::vector<KtLorentzVector> getJetsEta(); |
Return final state jets in order of decreasing pseudorapidity |
If KtLorentzVectors were used as input into the KtEvent constructor
there is a further method which will return the final state jet which
contains the input KtLorentzVector.
KtLorentzVector getJet(const KtLorentzVector &) const;
In addition to these Jet returning methods, KtEvent has the following
informational methods that are useful with the inclusive mode:
int getNJets() const; |
Return the number of final state jets |
int getNConstituents() const; |
Return the number of objects input to KtEvent |
std::vector<const KtLorentzVector *> getConstituents() const; |
Return pointers to the input particles |
std::vector<KtLorentzVector> copyConstituents() const; |
Return copies of the input particles |
KtFloat getETot() const; |
Return total energy in the event |
int getType() const; |
Return collision type |
int getAngle() const; |
Return distance ("angle") scheme |
int getRecom() const; |
Return recombination scheme |
bool isInclusive() const; |
Return inclusive flag: true if inclusive method constructor was used |
The exclusive mode has the following constructors:
KtEvent(const std::vector<KtLorentzVector> &,
int type, int angle, int recom);KtEvent(const std::vector<HepLorentzVector> &,
int type, int angle, int recom);KtEvent(const std::vector<KtLorentzVector> &,
int type, KtDistance *, KtRecom *);KtEvent(const std::vector<KtLorentzVector> &,
int type, KtDistance *, int recom);KtEvent(const std::vector<KtLorentzVector> &,
int type, int angle, KtRecom *);
These are the same as the inclusive mode, except without the
rparameter. Once exclusive KtEvent object has been instantiated the
final state jets can be defined by:
- setting either the stopping parameter d_cut
void findJetsD(KtFloat dCut);
- or by forcing the final state to decompose into N jets
void findJetsN(int nJets);
In the case of an e+e- analysis, the variable y_cut may be used.
The final state jets are defined by the method
void findJetsY(KtFloat yCut);
The E_cut value can also be set (the default value is the total energy of the event):
void setECut(KtFloat eCut);
The exclusive KtEvent Object can also be interrogated about the d_cut and y_cut values when
n+1 jets merged into n jets:
KtFloat getDMerge(int nJets) const;
KtFloat getYMerge(int nJets) const;
In addition to these methods, any of those used in the inclusive mode may also be used.
KtJet may also be used to perform sub-jet analysis on any of the
output final state jets. This is a simple matter of constructing a
KtEvent object with one of the output final state jets. This can be
used in the same way as when performing exclusive jet finding in the
e+e- mode. KtEvent has the following constructors for sub-jet
analysis, they should by now have a familiar structure to the user,
except they take a single KtLorentzVector "Jet" instead of a
std::vector of "particles".
KtEvent(const KtLorentzVector jet, int angle, int recom);KtEvent(const KtLorentzVector & jet, KtDistance *, KtRecom *);KtEvent(const KtLorentzVector & jet, KtDistance *, int recom);KtEvent(const KtLorentzVector & jet, int angle, KtRecom *);
KtLorentzVector
All input "particle" and output "jet" objects in KtJet are KtLorentzVectors. KtEvents can
be instantiated with HepLorentzVectors, but the first thing KtEvent
does is construct KtLorentVectors from these. KtLorentzVector inherit
from CLHEPs HepLorentzVector and adds extra functionality so they can
be tracked easily inside KtEvent. It has the following
constructors:
KtLorentzVector();KtLorentzVector(const HepLorentzVector &);KtLorentzVector(KtFloat px, KtFloat py, KtFloat pz, KtFloat e);
When KtLorentzVectors
are added together the resulting KtLorentzVector contains a
std::vector of its constituents. Information about a KtLorentzVector's
constituents can be gained using the following methods.
const std::vector<const KtLorentzVector*> & getConstituents() const; |
return a reference to the vector of constituents |
std::vector<KtLorentzVector> copyConstituents() const; |
copy constituents |
inline int getNConstituents() const; |
returns the number of constituents KtLorentzVector is made up of |
bool contains(const KtLorentzVector &) const; |
check if a given KtLorentzVector is a constituent of this one |
A KtLorentzVector can also be queried as to whether
it is a Jet (has constituents) or not.
bool isJet() const;
KtLorentzVector has two extra methods for adding two
KtLorentzVectors together. One which does simple 4 vector
adding (E-scheme) and one which takes a recombination scheme as an argument and uses this to combine the
KtLorentzVectors.
void add(const KtLorentzVector &, KtRecom *recom);void add(const KtLorentzVector &);
There is also a operator+= which simply adds the the KtLorentzVectors using the E-scheme.
KtLorentzVector & operator+= (const KtLorentzVector &);
Finally there are four comparison operators.
bool operator== (const KtLorentzVector &) const;bool operator!= (const KtLorentzVector &) const;bool operator> (const KtLorentzVector &) const;bool operator< (const KtLorentzVector &) const;
It is important to note that these functions ONLY compare the
KtLorentzVectors IDs. So the opertator== function will return true for
two KtLorentzVectors with the same ID number even if their
four-momentums are different. The decision to do this was taken
because it improves the performance of KtJet and makes the use of STL
containers much simpler for the users. The HepLorentzVector
comparison, which just compares the four momentum, can still be used
on KtLorentzVectors, but involves much more convoluted calls. For
example, the four momentum of two KtLorentzVectors, vec0 and vec1, can
be compared as follows:
vec0.HepLorentzVector::operator==(vec1)
Example Code
For examples on how to run with KtJet see Examples
Defining distance and recombination schemes
For information on how to define your own distance and recombination schemes go
here
|