fastjet 2.4.3
|
00001 00002 //STARTHEADER 00003 // $Id: ClusterSequenceAreaBase.cc 1601 2009-05-28 13:43:43Z soyez $ 00004 // 00005 // Copyright (c) 2005-2006, Matteo Cacciari and Gavin Salam 00006 // 00007 //---------------------------------------------------------------------- 00008 // This file is part of FastJet. 00009 // 00010 // FastJet is free software; you can redistribute it and/or modify 00011 // it under the terms of the GNU General Public License as published by 00012 // the Free Software Foundation; either version 2 of the License, or 00013 // (at your option) any later version. 00014 // 00015 // The algorithms that underlie FastJet have required considerable 00016 // development and are described in hep-ph/0512210. If you use 00017 // FastJet as part of work towards a scientific publication, please 00018 // include a citation to the FastJet paper. 00019 // 00020 // FastJet is distributed in the hope that it will be useful, 00021 // but WITHOUT ANY WARRANTY; without even the implied warranty of 00022 // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 00023 // GNU General Public License for more details. 00024 // 00025 // You should have received a copy of the GNU General Public License 00026 // along with FastJet; if not, write to the Free Software 00027 // Foundation, Inc.: 00028 // 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA 00029 //---------------------------------------------------------------------- 00030 //ENDHEADER 00031 00032 00033 00034 00035 #include "fastjet/ClusterSequenceAreaBase.hh" 00036 #include <algorithm> 00037 00038 FASTJET_BEGIN_NAMESPACE 00039 00040 using namespace std; 00041 00042 00044 LimitedWarning ClusterSequenceAreaBase::_warnings; 00045 LimitedWarning ClusterSequenceAreaBase::_warnings_zero_area; 00046 00047 //---------------------------------------------------------------------- 00055 double ClusterSequenceAreaBase::empty_area(const RangeDefinition & range) const { 00056 00057 if (has_explicit_ghosts()) {return 0.0;} 00058 else { return empty_area_from_jets(inclusive_jets(0.0), range);} 00059 00060 } 00061 00062 //---------------------------------------------------------------------- 00067 double ClusterSequenceAreaBase::empty_area_from_jets( 00068 const std::vector<PseudoJet> & all_jets, 00069 const RangeDefinition & range) const { 00070 00071 double empty = range.area(); 00072 for (unsigned i = 0; i < all_jets.size(); i++) { 00073 if (range.is_in_range(all_jets[i])) empty -= area(all_jets[i]); 00074 } 00075 return empty; 00076 } 00077 00078 double ClusterSequenceAreaBase::median_pt_per_unit_area(const RangeDefinition & range) const { 00079 return median_pt_per_unit_something(range,false); 00080 } 00081 00082 double ClusterSequenceAreaBase::median_pt_per_unit_area_4vector(const RangeDefinition & range) const { 00083 return median_pt_per_unit_something(range,true); 00084 } 00085 00086 00087 //---------------------------------------------------------------------- 00091 double ClusterSequenceAreaBase::median_pt_per_unit_something( 00092 const RangeDefinition & range, bool use_area_4vector) const { 00093 00094 double median, sigma, mean_area; 00095 get_median_rho_and_sigma(range, use_area_4vector, median, sigma, mean_area); 00096 return median; 00097 00098 } 00099 00100 00101 //---------------------------------------------------------------------- 00105 void ClusterSequenceAreaBase::parabolic_pt_per_unit_area( 00106 double & a, double & b, const RangeDefinition & range, 00107 double exclude_above, bool use_area_4vector) const { 00108 00109 int n=0; 00110 int n_excluded = 0; 00111 double mean_f=0, mean_x2=0, mean_x4=0, mean_fx2=0; 00112 00113 vector<PseudoJet> incl_jets = inclusive_jets(); 00114 00115 for (unsigned i = 0; i < incl_jets.size(); i++) { 00116 if (range.is_in_range(incl_jets[i])) { 00117 double this_area; 00118 if ( use_area_4vector ) { 00119 this_area = area_4vector(incl_jets[i]).perp(); 00120 } else { 00121 this_area = area(incl_jets[i]); 00122 } 00123 double f = incl_jets[i].perp()/this_area; 00124 if (exclude_above <= 0.0 || f < exclude_above) { 00125 double x = incl_jets[i].rap(); double x2 = x*x; 00126 mean_f += f; 00127 mean_x2 += x2; 00128 mean_x4 += x2*x2; 00129 mean_fx2 += f*x2; 00130 n++; 00131 } else { 00132 n_excluded++; 00133 } 00134 } 00135 } 00136 00137 if (n <= 1) { 00138 // meaningful results require at least two jets inside the 00139 // area -- mind you if there are empty jets we should be in 00140 // any case doing something special... 00141 a = 0.0; 00142 b = 0.0; 00143 } else { 00144 mean_f /= n; 00145 mean_x2 /= n; 00146 mean_x4 /= n; 00147 mean_fx2 /= n; 00148 00149 b = (mean_f*mean_x2 - mean_fx2)/(mean_x2*mean_x2 - mean_x4); 00150 a = mean_f - b*mean_x2; 00151 } 00152 //cerr << "n_excluded = "<< n_excluded << endl; 00153 } 00154 00155 00156 00157 void ClusterSequenceAreaBase::get_median_rho_and_sigma( 00158 const RangeDefinition & range, bool use_area_4vector, 00159 double & median, double & sigma, double & mean_area) const { 00160 00161 vector<PseudoJet> incl_jets = inclusive_jets(); 00162 get_median_rho_and_sigma(incl_jets, range, use_area_4vector, 00163 median, sigma, mean_area, true); 00164 } 00165 00166 00167 void ClusterSequenceAreaBase::get_median_rho_and_sigma( 00168 const vector<PseudoJet> & all_jets, 00169 const RangeDefinition & range, bool use_area_4vector, 00170 double & median, double & sigma, double & mean_area, 00171 bool all_are_incl) const { 00172 00173 _check_jet_alg_good_for_median(); 00174 00175 vector<double> pt_over_areas; 00176 double total_area = 0.0; 00177 double total_njets = 0; 00178 00179 for (unsigned i = 0; i < all_jets.size(); i++) { 00180 if (range.is_in_range(all_jets[i])) { 00181 double this_area; 00182 if (use_area_4vector) { 00183 this_area = area_4vector(all_jets[i]).perp(); 00184 } else { 00185 this_area = area(all_jets[i]); 00186 } 00187 00188 if (this_area>0) { 00189 pt_over_areas.push_back(all_jets[i].perp()/this_area); 00190 } else { 00191 _warnings_zero_area.warn("ClusterSequenceAreaBase::get_median_rho_and_sigma(...): discarded jet with zero area. Zero-area jets may be due to (i) too large a ghost area (ii) a jet being outside the ghost range (iii) the computation not being done using an appropriate algorithm (kt;C/A)."); 00192 } 00193 00194 total_area += this_area; 00195 total_njets += 1.0; 00196 } 00197 } 00198 00199 // there is nothing inside our region, so answer will always be zero 00200 if (pt_over_areas.size() == 0) { 00201 median = 0.0; 00202 sigma = 0.0; 00203 mean_area = 0.0; 00204 return; 00205 } 00206 00207 // get median (pt/area) [this is the "old" median definition. It considers 00208 // only the "real" jets in calculating the median, i.e. excluding the 00209 // only-ghost ones; it will be supplemented with more info below] 00210 sort(pt_over_areas.begin(), pt_over_areas.end()); 00211 00212 // now get the median & error, accounting for empty jets 00213 // define the fractions of distribution at median, median-1sigma 00214 double posn[2] = {0.5, (1.0-0.6827)/2.0}; 00215 double res[2]; 00216 00217 double n_empty, empty_a; 00218 if (has_explicit_ghosts()) { 00219 // NB: the following lines of code are potentially incorrect in cases 00220 // where there are unclustered particles (empty_area would do a better job, 00221 // at least for active areas). This is not an issue with kt or C/A, or other 00222 // algorithms that cluster all particles (and the median estimation should in 00223 // any case only be done with kt or C/A!) 00224 empty_a = 0.0; 00225 n_empty = 0; 00226 } else if (all_are_incl) { 00227 // the default case 00228 empty_a = empty_area(range); 00229 n_empty = n_empty_jets(range); 00230 } else { 00231 // this one is intended to be used when e.g. one runs C/A, then looks at its 00232 // exclusive jets in order to get an effective smaller R value, and passes those 00233 // to this routine. 00234 empty_a = empty_area_from_jets(all_jets, range); 00235 mean_area = total_area / total_njets; // temporary value 00236 n_empty = empty_a / mean_area; 00237 } 00238 //cout << "*** tot_area = " << total_area << ", empty_a = " << empty_a << endl; 00239 //cout << "*** n_empty = " << n_empty << ", ntotal = " << total_njets << endl; 00240 total_njets += n_empty; 00241 total_area += empty_a; 00242 00243 for (int i = 0; i < 2; i++) { 00244 double nj_median_pos = 00245 (pt_over_areas.size()-1 + n_empty)*posn[i] - n_empty; 00246 double nj_median_ratio; 00247 if (nj_median_pos >= 0 && pt_over_areas.size() > 1) { 00248 int int_nj_median = int(nj_median_pos); 00249 nj_median_ratio = 00250 pt_over_areas[int_nj_median] * (int_nj_median+1-nj_median_pos) 00251 + pt_over_areas[int_nj_median+1] * (nj_median_pos - int_nj_median); 00252 } else { 00253 nj_median_ratio = 0.0; 00254 } 00255 res[i] = nj_median_ratio; 00256 } 00257 median = res[0]; 00258 double error = res[0] - res[1]; 00259 mean_area = total_area / total_njets; 00260 sigma = error * sqrt(mean_area); 00261 } 00262 00263 00268 vector<PseudoJet> ClusterSequenceAreaBase::subtracted_jets(const double rho, 00269 const double ptmin) 00270 const { 00271 vector<PseudoJet> sub_jets; 00272 vector<PseudoJet> jets = sorted_by_pt(inclusive_jets(ptmin)); 00273 for (unsigned i=0; i<jets.size(); i++) { 00274 PseudoJet sub_jet = subtracted_jet(jets[i],rho); 00275 sub_jets.push_back(sub_jet); 00276 } 00277 return sub_jets; 00278 } 00279 00284 vector<PseudoJet> ClusterSequenceAreaBase::subtracted_jets( 00285 const RangeDefinition & range, 00286 const double ptmin) 00287 const { 00288 double rho = median_pt_per_unit_area_4vector(range); 00289 return subtracted_jets(rho,ptmin); 00290 } 00291 00292 00294 PseudoJet ClusterSequenceAreaBase::subtracted_jet(const PseudoJet & jet, 00295 const double rho) const { 00296 PseudoJet area4vect = area_4vector(jet); 00297 PseudoJet sub_jet; 00298 // sanity check 00299 if (rho*area4vect.perp() < jet.perp() ) { 00300 sub_jet = jet - rho*area4vect; 00301 } else { sub_jet = PseudoJet(0.0,0.0,0.0,0.0); } 00302 00303 // make sure the subtracted jet has the same index (cluster and user) 00304 // (i.e. "looks like") the original jet 00305 sub_jet.set_cluster_hist_index(jet.cluster_hist_index()); 00306 sub_jet.set_user_index(jet.user_index()); 00307 00308 return sub_jet; 00309 } 00310 00311 00315 PseudoJet ClusterSequenceAreaBase::subtracted_jet(const PseudoJet & jet, 00316 const RangeDefinition & range) const { 00317 double rho = median_pt_per_unit_area_4vector(range); 00318 PseudoJet sub_jet = subtracted_jet(jet, rho); 00319 return sub_jet; 00320 } 00321 00322 00324 double ClusterSequenceAreaBase::subtracted_pt(const PseudoJet & jet, 00325 const double rho, 00326 bool use_area_4vector) const { 00327 if ( use_area_4vector ) { 00328 PseudoJet sub_jet = subtracted_jet(jet,rho); 00329 return sub_jet.perp(); 00330 } else { 00331 return jet.perp() - rho*area(jet); 00332 } 00333 } 00334 00335 00339 double ClusterSequenceAreaBase::subtracted_pt(const PseudoJet & jet, 00340 const RangeDefinition & range, 00341 bool use_area_4vector) const { 00342 if ( use_area_4vector ) { 00343 PseudoJet sub_jet = subtracted_jet(jet,range); 00344 return sub_jet.perp(); 00345 } else { 00346 double rho = median_pt_per_unit_area(range); 00347 return subtracted_pt(jet,rho,false); 00348 } 00349 } 00350 00351 00353 void ClusterSequenceAreaBase::_check_jet_alg_good_for_median() const { 00354 if (jet_def().jet_algorithm() != kt_algorithm 00355 && jet_def().jet_algorithm() != cambridge_algorithm 00356 && jet_def().jet_algorithm() != cambridge_for_passive_algorithm) { 00357 _warnings.warn("ClusterSequenceAreaBase: jet_def being used may not be suitable for estimating diffuse backgrounds (good options are kt, cam)"); 00358 } 00359 } 00360 00361 00362 00363 FASTJET_END_NAMESPACE