fastjet 2.4.5
Public Types | Public Member Functions | Private Attributes
fastjet::CDFMidPointPlugin Class Reference

CDFMidPointPlugin is a plugin for fastjet (v2.1 upwards) that provides an interface to the CDF version of Run-II iterative cone algorithm with midpoint seeds (also known as the Iterative Legacy Cone Algorithm, ILCA). More...

#include <CDFMidPointPlugin.hh>

Inheritance diagram for fastjet::CDFMidPointPlugin:
Inheritance graph
[legend]
Collaboration diagram for fastjet::CDFMidPointPlugin:
Collaboration graph
[legend]

List of all members.

Public Types

enum  SplitMergeScale { SM_pt, SM_Et, SM_mt, SM_pttilde }
 the choice of scale to be used in the split-merge step More...

Public Member Functions

 CDFMidPointPlugin (double seed_threshold, double cone_radius, double cone_area_fraction, int max_pair_size, int max_iterations, double overlap_threshold, SplitMergeScale sm_scale=SM_pt)
 A CDFMidPointPlugin constructor that looks like the one provided by CDF.
 CDFMidPointPlugin (double cone_radius, double overlap_threshold, double seed_threshold=1.0, double cone_area_fraction=1.0)
 a compact constructor
double seed_threshold () const
double cone_radius () const
double cone_area_fraction () const
int max_pair_size () const
int max_iterations () const
double overlap_threshold () const
virtual std::string description () const
 return a textual description of the jet-definition implemented in this plugin
virtual void run_clustering (ClusterSequence &) const
 given a ClusterSequence that has been filled up with initial particles, the following function should fill up the rest of the ClusterSequence, using the following member functions of ClusterSequence:
virtual double R () const
 the plugin mechanism's standard way of accessing the jet radius

Private Attributes

double _seed_threshold
double _cone_radius
double _cone_area_fraction
int _max_pair_size
int _max_iterations
double _overlap_threshold
SplitMergeScale _sm_scale

Detailed Description

CDFMidPointPlugin is a plugin for fastjet (v2.1 upwards) that provides an interface to the CDF version of Run-II iterative cone algorithm with midpoint seeds (also known as the Iterative Legacy Cone Algorithm, ILCA).

The CDF code has been taken from Joey Huston's webpage http://www.pa.msu.edu/~huston/Les_Houches_2005/Les_Houches_SM.html

Note that the CDF midpoint code contains options that go beyond those described in the Tevatron run-II document (hep-ex/0005012), notably search-cones, as described in hep-ph/0111434, and midpoints bewteen multiplets of stable cones.

Additionally, the version of the CDF midpoint code distributed here has been modified by the FastJet authors, so as to allow one to choose the scale used in the split-merge step.

Definition at line 60 of file CDFMidPointPlugin.hh.


Member Enumeration Documentation

the choice of scale to be used in the split-merge step

Enumerator:
SM_pt 
SM_Et 
SM_mt 
SM_pttilde 

Definition at line 64 of file CDFMidPointPlugin.hh.


Constructor & Destructor Documentation

fastjet::CDFMidPointPlugin::CDFMidPointPlugin ( double  seed_threshold,
double  cone_radius,
double  cone_area_fraction,
int  max_pair_size,
int  max_iterations,
double  overlap_threshold,
SplitMergeScale  sm_scale = SM_pt 
) [inline]

A CDFMidPointPlugin constructor that looks like the one provided by CDF.

Its arguments should have the following meaning:

  • seed_threshold: minimum pt for a particle to be considered a seed of the iteration.
  • cone_radius: standard meaning
  • cone_area_fraction: stable-cones are searched for with a radius Rsearch = R * sqrt(cone_area_fraction), and then expanded to size R afterwards; note (hep-ph/0610012) that this introduces IR unsafety at NLO for X+2-jet observables (where X any hard object).
  • max_pair_size: "midpoints" can be added between pairs of stable cones, triplets of stable cones, etc.; max_pair_size indicates the maximum number of stable cones that are assembled when adding midpoints.
  • max_iterations: the maximum number of iterations to carry out when looking for a stable cone.
  • overlap_threshold: if (overlapping_Et)/(Et_of_softer_protojet) < overlap_threshold, overlapping jets are split, otherwise they are merged.
  • sm_scale: a choice for the scale to be used in the split-merge step (both for ordering the momenta and quantifying the overlap); the three options are

. SM_pt: pt (default -- source of small IR safety issue in purely hadronic events)

. SM_Et: Et (not boost invariant, reduces to mt at zero rapidity and to pt and infinite rapidity)

. SM_mt: transverse mass = sqrt(m^2+pt^2)

Definition at line 105 of file CDFMidPointPlugin.hh.

fastjet::CDFMidPointPlugin::CDFMidPointPlugin ( double  cone_radius,
double  overlap_threshold,
double  seed_threshold = 1.0,
double  cone_area_fraction = 1.0 
) [inline]

a compact constructor

NB: as of version 2.4, the default value for the overlap_threshold threshold has been removed, to avoid misleading people into using the value of 0.5 without thinking, which is known to have adverse effects in high-noise environments. A recommended value is 0.75.

Definition at line 128 of file CDFMidPointPlugin.hh.


Member Function Documentation

double fastjet::CDFMidPointPlugin::cone_area_fraction ( ) const [inline]

Definition at line 144 of file CDFMidPointPlugin.hh.

double fastjet::CDFMidPointPlugin::cone_radius ( ) const [inline]

Definition at line 143 of file CDFMidPointPlugin.hh.

{return _cone_radius        ;}
string fastjet::CDFMidPointPlugin::description ( ) const [virtual]

return a textual description of the jet-definition implemented in this plugin

Implements fastjet::JetDefinition::Plugin.

Definition at line 46 of file CDFMidPointPlugin.cc.

                                             {
  ostringstream desc;
  
  string sm_scale_string = "split-merge uses ";
  switch(_sm_scale) {
  case SM_pt:
    sm_scale_string += "pt";
    break;
  case SM_Et:
    sm_scale_string += "Et";
    break;
  case SM_mt:
    sm_scale_string += "mt";
    break;
  case SM_pttilde:
    sm_scale_string += "pttilde (scalar sum of pts)";
    break;
  default:
    ostringstream err;
    err << "Unrecognized split-merge scale choice = " << _sm_scale;
    throw Error(err.str());
  }

  
  if (cone_area_fraction() == 1) {
    desc << "CDF MidPoint jet algorithm, with " ;
  } else {
    desc << "CDF MidPoint+Searchcone jet algorithm, with ";
  }
  desc << "seed_threshold = "     << seed_threshold     () << ", "
       << "cone_radius = "        << cone_radius        () << ", "
       << "cone_area_fraction = " << cone_area_fraction () << ", " 
       << "max_pair_size = "      << max_pair_size      () << ", "
       << "max_iterations = "     << max_iterations     () << ", "
       << "overlap_threshold  = " << overlap_threshold  () << ", "
       << sm_scale_string ;

  return desc.str();
}
int fastjet::CDFMidPointPlugin::max_iterations ( ) const [inline]

Definition at line 146 of file CDFMidPointPlugin.hh.

{return _max_iterations     ;}
int fastjet::CDFMidPointPlugin::max_pair_size ( ) const [inline]

Definition at line 145 of file CDFMidPointPlugin.hh.

{return _max_pair_size      ;}
double fastjet::CDFMidPointPlugin::overlap_threshold ( ) const [inline]

Definition at line 147 of file CDFMidPointPlugin.hh.

{return _overlap_threshold  ;}
virtual double fastjet::CDFMidPointPlugin::R ( ) const [inline, virtual]

the plugin mechanism's standard way of accessing the jet radius

Implements fastjet::JetDefinition::Plugin.

Definition at line 154 of file CDFMidPointPlugin.hh.

{return cone_radius();}
void fastjet::CDFMidPointPlugin::run_clustering ( ClusterSequence ) const [virtual]

given a ClusterSequence that has been filled up with initial particles, the following function should fill up the rest of the ClusterSequence, using the following member functions of ClusterSequence:

  • plugin_do_ij_recombination(...)
  • plugin_do_iB_recombination(...)

Implements fastjet::JetDefinition::Plugin.

Definition at line 87 of file CDFMidPointPlugin.cc.

References fastjet::ClusterSequence::jets(), fastjet::ClusterSequence::plugin_record_iB_recombination(), and fastjet::ClusterSequence::plugin_record_ij_recombination().

                                                                        {
 
  // create the physics towers needed by the CDF code
  vector<PhysicsTower> towers;
  towers.reserve(clust_seq.jets().size());
  for (unsigned i = 0; i < clust_seq.jets().size(); i++) {
    LorentzVector fourvect(clust_seq.jets()[i].px(),
                           clust_seq.jets()[i].py(),
                           clust_seq.jets()[i].pz(),
                           clust_seq.jets()[i].E());
    PhysicsTower tower(fourvect);
    // misuse one of the indices for tracking, since the MidPoint
    // implementation doesn't seem to make use of these indices
    tower.calTower.iEta = i;
    towers.push_back(tower);
  }

  // prepare the CDF algorithm
  MidPointAlgorithm m(_seed_threshold,_cone_radius,_cone_area_fraction,
                      _max_pair_size,_max_iterations,_overlap_threshold,
                      MidPointAlgorithm::SplitMergeScale(_sm_scale));
    
  // run the CDF algorithm
  std::vector<Cluster> jets;
  m.run(towers,jets);


  // now transfer the jets back into our own structure -- we will
  // mimic the cone code with a sequential recombination sequence in
  // which the jets are built up by adding one particle at a time
  for(vector<Cluster>::const_iterator jetIter = jets.begin(); 
                                      jetIter != jets.end(); jetIter++) {
    const vector<PhysicsTower> & tower_list = jetIter->towerList;
    int jet_k = tower_list[0].calTower.iEta;
  
    int ntow = int(jetIter->towerList.size());
    for (int itow = 1; itow < ntow; itow++) {
      int jet_i = jet_k;
      // retrieve our misappropriated index for the jet
      int jet_j = tower_list[itow].calTower.iEta;
      // do a fake recombination step with dij=0
      double dij = 0.0;
      clust_seq.plugin_record_ij_recombination(jet_i, jet_j, dij, jet_k);
    }
  
    // NB: put a sensible looking d_iB just to be nice...
    double d_iB = clust_seq.jets()[jet_k].perp2();
    clust_seq.plugin_record_iB_recombination(jet_k, d_iB);
  }


  // following code is for testing only
  //cout << endl;
  //for(vector<Cluster>::const_iterator jetIter = jets.begin(); 
  //                                    jetIter != jets.end(); jetIter++) {
  //  cout << jetIter->fourVector.pt() << " " << jetIter->fourVector.y() << endl;
  //}
  //cout << "-----------------------------------------------------\n";
  //vector<PseudoJet> ourjets(clust_seq.inclusive_jets());
  //for (vector<PseudoJet>::const_reverse_iterator ourjet = ourjets.rbegin();
  //     ourjet != ourjets.rend(); ourjet++) {
  //  cout << ourjet->perp() << " " << ourjet->rap() << endl;
  //}
  //cout << endl;
}
double fastjet::CDFMidPointPlugin::seed_threshold ( ) const [inline]

Definition at line 142 of file CDFMidPointPlugin.hh.

{return _seed_threshold     ;}

Member Data Documentation

Definition at line 160 of file CDFMidPointPlugin.hh.

Definition at line 159 of file CDFMidPointPlugin.hh.

Definition at line 162 of file CDFMidPointPlugin.hh.

Definition at line 161 of file CDFMidPointPlugin.hh.

Definition at line 163 of file CDFMidPointPlugin.hh.

Definition at line 158 of file CDFMidPointPlugin.hh.

Definition at line 164 of file CDFMidPointPlugin.hh.


The documentation for this class was generated from the following files:
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Defines