skim.cxx File Reference

#include <stdlib.h>
#include <iostream>
#include <vector>
#include <set>
#include <string>
#include <regex>
#include <algorithm>
#include <cstdio>
#include "simpletimer.h"
#include "TChain.h"
#include "TFile.h"
#include "TTree.h"
#include "TH1D.h"
#include "TrigInDetAnalysis/TIDAEvent.h"
#include "utils.h"

Go to the source code of this file.

Functions
template<class T >
void	remove_duplicates (std::vector< T > &vec)
std::string	time_str ()
int	usage (int e=0)
template<class T >
std::ostream &	operator<< (std::ostream &os, const std::set< T > &s)
template<class T >
std::ostream &	operator<< (std::ostream &s, const std::vector< T > &v)
void	copyReleaseInfo (TFile *finput, TFile *foutdir)
	copy the TTree of release info from one directory to another More...
int	main (int argc, char **argv)

Detailed Description

Author

mark sutton

Date

Mon 30 Jan 2012 18:43:21 CET

Copyright (C) 2002-2023 CERN for the benefit of the ATLAS collaboration

Definition in file skim.cxx.

Function Documentation

copyReleaseInfo()

void copyReleaseInfo	(	TFile *	finput,
		TFile *	foutdir
	)

copy the TTree of release info from one directory to another

copy the TTree of release info from one directory to another

Definition at line 82 of file skim.cxx.

                                                      { 
 
  if ( finput && foutdir ) { 
 
    TTree* tree  = (TTree*)finput->Get("dataTree");    
    TTree* clone = tree->CloneTree();
 
    foutdir->cd();
    clone->Write("", TObject::kOverwrite);
 
    delete clone;
 
  }
  
}

main()

int main	(	int	argc,
		char **	argv
	)

need to properly fix the logic for deleting

open output file

create the main event TTree ...

event counters

open inputfile

now the main event TTree

this bit of code will drop any event where the HLT hasn't passed

now onto the business of deleting the chains ...

delete the listed chains

get the offline chain

Definition at line 101 of file skim.cxx.

                                { 
  
  if ( argc<1 ) return usage(-1);
  
  std::set<std::string>   require_chains;
  std::vector<std::string>   rchains;
 
  bool require  = false;
  bool deleting = false;
 
  std::string outfile = "tree.root";
 
  std::string infile="";
 
  bool force = false;
 
  double ptmin = 0;
 
  bool roi_filter = false;
 
  bool verbose = false;
 
  for ( int i=1 ; i<argc ; i++ ) { 
 
    std::string arg(argv[i]);
 
    
    if ( arg.find('-')!=0 ) { 
      if ( !deleting && infile!="" ) { 
    require = true;
    require_chains.insert(argv[i]); 
    rchains.push_back(argv[i]); 
    continue;
      }
      else { 
    if ( infile=="" ) infile = arg;
    else { 
      std::cerr << "more than one file specified: " << arg << std::endl;
      return usage(-2);
      infile=arg;
    }
      }
    }
    else { 
      if      ( arg=="-h" || arg=="--help" )   return usage(0);
      else if ( arg=="-o" || arg=="--output" ) { 
    if ( (i+1)<argc )  outfile = argv[++i];
    else               return usage(-1);
      }
      else if ( arg=="-r" || arg=="--require" ) { 
    if ( deleting ) { 
      std::cerr << "cannot require and delete chains" << std::endl;
      return usage(-4);
    }
    require = true;
      }
      else if ( arg=="-d" || arg=="--delete" ) {
    if ( require ) { 
      std::cerr << "cannot require and delete chains" << std::endl;
      return usage(-3);
    }
    deleting = true;
      }
      else if ( arg=="--pt" ) { 
    if ( (i+1)<argc )  ptmin = std::atof(argv[++i])*1000;
    else               return usage(-1);
    force = true;
      }
      else if ( arg=="-v" || arg=="--verbose" ) verbose = true;
      else if ( arg=="-f" || arg=="--force" )   force = true;
      else if (              arg=="--roi" )     { force=true; roi_filter = true; }
      else if ( infile=="" ) infile = arg;
      else { 
    std::cerr << "more than one file specified: " << arg << std::endl;
    return usage(-2);
      }
    }
  }
 
  std::cout << "required chains " << require_chains << std::endl;
 
  if ( require_chains.size()>0 ) require = true;
 
  if ( !force && require_chains.size()==0 ) { 
    std::cout << "no chains requested - not doing anything" << std::endl;
    return 0;
  }
 
  std::cout << "chains: " << rchains << std::endl;
 
 
  std::cout << "skim::start       " << time_str() << std::endl; 
 
 
  if ( require  ) std::cout << "require chains " << require_chains << std::endl;
  if ( deleting ) std::cout << "delete  chains " << require_chains << std::endl;
  
 
  if ( infile=="" ) { 
    std::cerr << "no files specified" << std::endl;
    return usage(-1);
  }
 
  std::cout << "reading from file: " << infile << std::endl;
  std::cout << "writing to file:   " << outfile << std::endl;
 
 
  TIDA::Event* track_ev = new TIDA::Event();
  TIDA::Event* h = track_ev;
 
  std::cout << "opening outfile:   " << outfile << std::endl;
 
  TFile fout( outfile.c_str(), "recreate");
 
  fout.cd();
 
 
  TTree *tree = new TTree("tree","tree");
  
  tree->Branch("TIDA::Event", "TIDA::Event", &h, 6400, 1);
    
  h->clear();
 
 
  int ev_in  = 0;
  int ev_out = 0;
  
  {
    
    TFile finput( infile.c_str() );
    if (!finput.IsOpen()) {
      std::cerr << "Error: could not open output file" << std::endl;
      exit(-1);
    }
 
 
    copyReleaseInfo( &finput, &fout );
 
    finput.cd();
 
 
    //   TChain* data = new TChain("tree");
 
    TTree* data = (TTree*)finput.Get("tree");
 
    TIDA::Event* track_iev = new TIDA::Event();
    
    data->SetBranchAddress("TIDA::Event",&track_iev);
    //    data->AddFile( argv[i] );
    
 
    unsigned entries = data->GetEntries();
 
    std::cout << "input has " << entries << " events" << std::endl; 
 
    std::string cck = "|/-\\";
        
    int ii=0;
 
    struct timeval tm = simpletimer_start();
 
    std::cout << "processing ..." << std::endl;
 
    for (unsigned int i=0; i<data->GetEntries() ; i++ ) {
 
      if ( verbose ) std::cout << "event " << i;
 
      if ( i%100==0 ) {
 
    double frac = i*1.0/entries;
 
    double t = simpletimer_stop(tm);
    
    double est = 0;
 
    if ( frac > 0 ) est = t/frac - t;
    
    double eventsps = 1000*i/t;
 
    std::printf( "\r%c    %6.2lf %%     time: %6.2lf s    remaining %6.2lf s   (%u at %5.2lf ps) ", 
             cck[ii%4], ((1000*(i+1)/entries)*0.1), t*0.001, est*0.001, i, eventsps );   
 
    std::fflush(stdout);
 
    ii++;
      }
      
      
 
      track_iev->clear();
      track_ev->clear();
 
      data->GetEntry(i);      
 
      ev_in++;
 
      *track_ev = *track_iev;
 
      //      std::cout << *track_ev << std::endl;
 
      //      std::cout << "track_ev  names " << track_ev->chainnames()  << std::endl;
      //      std::cout << "track_iev names " << track_iev->chainnames() << std::endl;
 
      if ( verbose ) std::cout << "----------------------------------------\n\tN chains " << track_ev->size() << " -> ";
 
      std::vector<TIDA::Chain>& chains = track_ev->chains();
 
 
 
      bool skip = true;
 
      std::vector<TIDA::Chain>::iterator citr = chains.begin();
      for ( ; citr!=chains.end() ; ++citr ) {
 
    if ( citr->name().find("HLT")==std::string::npos ) continue;
 
    if ( require ) { 
      for ( unsigned j=rchains.size() ; j-- ; ) { 
        if ( rchains[j].find("HLT")==std::string::npos ) continue;
        if ( citr->name().find(rchains[j])!=std::string::npos ) { 
          skip = false;
          if ( verbose ) std::cout << "keepin' " << citr->name()  << " " << rchains[j] << std::endl;
        }
 
      }
    }
 
      }
 
      if ( skip ) continue;
 
 
      { 
 
 
    std::vector<std::string> chainnames = track_ev->chainnames();
 
    for ( size_t ic=0 ; ic<chainnames.size() ; ic++ ) {
      
      
      bool matched = false;
      for ( std::set<std::string>::iterator it=require_chains.begin() ; it!=require_chains.end() ; ++it ) { 
 
        matched |= std::regex_match( chainnames[ic], std::regex(*it+".*") );
 
        if ( verbose && matched ) std::cout << "chain: " << chainnames[ic] << "\t :: reg " << *it << "\tmatched: " << matched << std::endl;
        
      }
        
      if ( ( require && !matched ) ) track_ev->erase( chainnames[ic] );
    }
            
    if ( verbose ) std::cout << track_ev->size() << std::endl;
      
 
    if ( roi_filter || ptmin>0 ) { 
      
      TIDA::Chain* offline = 0;
 
      std::vector<std::string> chainnames = track_ev->chainnames();
 
 
      for ( size_t ic=chainnames.size() ; ic-- ; ) {
        if ( chainnames[ic] == "Offline" ) {
          offline = &(track_ev->chains()[ic]);
          break;
        }
      }
 
      if ( offline ) { 
        // track_ev->addChain( "Offline" ); 
        
        std::vector<TIDA::Chain>& chains = track_ev->chains();
        std::vector<TIDA::Chain>::iterator citr = chains.begin();
        
        std::vector<std::pair<double,double> > philims;
        
        for ( ; citr!=chains.end() ; ++citr ) {
          if ( citr->name().find("HLT_")!=std::string::npos ) { 
        for ( size_t ir=0 ; ir<citr->size() ; ir++ ) {
          TIDARoiDescriptor& roi = citr->rois()[ir].roi();
          if ( roi.composite() ) { 
            for ( size_t isub=0 ; isub<roi.size() ; isub++ ) { 
              philims.push_back( std::pair<double,double>( roi[isub]->phiMinus(), roi[isub]->phiPlus() ) ); 
            }
          }
          else philims.push_back( std::pair<double,double>( roi.phiMinus(), roi.phiPlus() ) ); 
        }
          }
        }
 
        remove_duplicates( philims );
 
        for ( size_t iroi=0 ; iroi<offline->size() ; iroi++ ) {
 
          std::vector<TIDA::Track>& tracks = offline->rois()[iroi].tracks();
          
          for ( std::vector<TIDA::Track>::iterator it=tracks.begin() ; it<tracks.end() ; ) {
        bool inc = true;
        if ( ptmin>0 ) { 
          if ( std::fabs(it->pT())<ptmin ) { inc=false; tracks.erase( it ); }
        }
        if ( inc && roi_filter ) { 
          bool remove_track = true;
          for ( size_t isub=0 ; isub<philims.size() ; isub++ ) { 
 
            if ( philims[isub].first < philims[isub].second ) { 
              if ( it->phi()>=philims[isub].first && it->phi()<=philims[isub].second ) { 
            remove_track = false; 
            break;
              }
            }
            else  {
              if ( it->phi()>=philims[isub].first || it->phi()<=philims[isub].second ) { 
            remove_track = false; 
            break;
              }
            }
          }
          if ( remove_track ) { inc=false; tracks.erase( it ); }
        }
        if ( inc ) ++it;
          }
 
        }
 
      }
 
 
    }
    
 
    if ( verbose ) std::cout << "writing event:\n" << *track_ev << std::endl;
 
    tree->Fill();
    ev_out++;
    
    
#if 0
    for (unsigned int ic=0 ; ic<chains.size() ; ic++ ) {
      if ( chains[ic].name()=="Offline" ) {
        const std::vector<TIDA::Track>& tracks = chains[ic].rois()[0].tracks();
        
        for ( unsigned it=0 ; it<tracks.size() ; it++ ) {
          h->Fill( tracks[it].pT()*0.001 );
          itracks++;
          
        }
        break;
      }
    }
#endif     
    
      }
 
    }
 
 
    double t = simpletimer_stop(tm);
    
    std::printf( "\r%c    %6.2lf %%     time: %6.2lf s\n", 
         cck[ii%4], 100., t*0.001 );   
 
 
    finput.Close();
    
 }
 
  std::cout << "skim::done        " << time_str() << std::endl; 
 
  std::cout << "skim::events in:  " << ev_in  << std::endl;
  std::cout << "skim::events out: " << ev_out << std::endl;
  
  fout.Write();
  fout.Close();
  
  return 0;
}

operator<<() [1/2]

template<class T >

std::ostream& operator<<	(	std::ostream &	os,
		const std::set< T > &	s
	)

Definition at line 62 of file skim.cxx.

                                                             {
  typename std::set<T>::const_iterator sitr = s.begin();
  os << "[  ";
  while ( sitr!=s.end() ) os << (*sitr++) << "\t";
  os << " ]";
  return os;
}

operator<<() [2/2]

template<class T >

std::ostream& operator<<	(	std::ostream &	s,
		const std::vector< T > &	v
	)

Definition at line 72 of file skim.cxx.

                                                               { 
  typename std::vector<T>::const_iterator sitr = v.begin();
  s << "[  ";
  while ( sitr!=v.end() ) s << (*sitr++) << "\t"; 
  s << " ]";
  return s;
}

remove_duplicates()

template<class T >

void remove_duplicates

(

std::vector< T > &

vec

)

Definition at line 34 of file skim.cxx.

                                          {
  std::sort(vec.begin(), vec.end());
  vec.erase(std::unique(vec.begin(), vec.end()), vec.end());
}

time_str()

std::string time_str

(

)

Definition at line 39 of file skim.cxx.

                     { 
  time_t t;
  time(&t);
  std::string s(ctime(&t));
  return s.substr(0,s.find('\n'));
}

usage()

int usage

(

int

e = 0

)

Definition at line 46 of file skim.cxx.

                   { 
    std::cerr << "usage: skim <filename> [OPTIONS]" << std::endl;
    std::cerr << "\nremoves chains from the ntple\n";
    std::cerr << "\nOptions:\n";
    std::cerr << "\t-d | --delete\tremoves specified chains\n";
    std::cerr << "\t-o | --output\toutput filename (default tree.root)\n";
    std::cerr << "\t-r | --require\trequire that this chain has tracks\n";
    std::cerr << "\t-f | --force\tforce processing even if no chains specified\n";
    std::cerr << "\t     --pt value\tremove offline tracks pt < value\n";
    std::cerr << "\t     --roi  \tfilter offline tracks by roi phi\n";
    std::cerr << "\nIf option -d is not given, the specified chains are retained and all others are removed" << std::endl;
    std::cerr << "\nIf no chains are specifed simply all events with no L2 or EF chains are excluded" << std::endl;
    return e;
}