NOTE: The on-line Databases with reprocessed info will be converted to root objects on ,at most, weekly basis to this directories.

For the Quality histograms

The procedure to use root to view the histograms is very simple:

1. Log in one of the hb-af14...20 machines
2. Make sure that the environment paths HBBINTYPE and BINTYPE are set to Linux_intel_glibc
3. Run the script with the settings with:
   . /afs/desy.de/group/hera-b/DATABASE/hblipcs4.1/settings.desy
4. cd /hb/database/DQ_HISTOS (which contains all the histos)
   or instead cd /hb/database2/DQ_REPRO/2 (which contains all the reprocessed histos)
5. Run the browser with: root viewhistos.C

Note: Should root report a missing .so file (specifically, HbBrowser.so), try setting the environment variable LD_LIBRARY_PATH with
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/afs/desy.de/group/hera-b/DATABASE/hblipcs4.1/Linux_intel_glibc/lib
before invoking root, so that root knows where to look for the aditional file(s). This is more likely to be needed if you use directory /hb/database2/DQ_REPRO/2 in step 4. This is for the bash shell; if you're using the tcsh shell, use instead: setenv LD_LIBRARY_PATH $LD_LIBRARY_PATH:/afs/desy.de/group/hera-b/DATABASE/hblipcs4.1/Linux_intel_glibc/lib

Now suppose you would like to see the histograms for the run 15012. The procedure continues as follows:

6. Select in the left window the folder that terminates with DQ_HISTOS
7. Then go to the folder 15000
8. In the right window, double-click the object DQ_15012.root to select it (if you want you can select others, and they will all accumulate in the folder ROOT_files)
9. In the left window select the folder ROOT_files, which should now contain the selected objects
10. Double-click on the object DQ_15012.root on the right window. Double-click again on the folder that it contains and you will see the list of all available histograms for that particular run. To see them just double-click

Note: It might happen that your looking for some runs that are not in the database (in the range of 16*** to 17***, which corresponds to a time lenght of about 3 weeks). Those runs probably correspond to the period of data acquisition where the Data Quality system was off due to some problems in the storage size.

It might have some interest manipulate the information contained inside the histograms using the powererful scripting facilities that ROOT provides. That's why in this page we describe some examples of simple scripts. Those scripts can be copied and easilly edited (just take a glimpse to the new ROOT Manual) in order to personalise the solution. To run them you just need to invoke the interpreter with root myscript If it doesn't work claiming that it doesn't find some *.so library, set the environment variable export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/afs/desy.de/group/hera-b/group/DATABASE/hblipcs4.1/Linux_intel_glibc/lib

• Script 1:

  This script opens a .root file (e.g., DQ_17042), checks for all histogram subsets inside the Run for one particular histogram (H10x1820a003) and gets the value of the bin content 5, printing it to the standart output.

  {
    TFile f("/hb/database/DQ_HISTOS/17000/DQ_17042.root");
    // Looking inside the root file
    TIter next(f.GetListOfKeys());
    TKey *key;
    while ((key=(TKey*)next())) {
      printf("Key:%s Class:%s at %d\n",
         key->GetName(),key->GetClassName(),key->GetSeekKey());
      f->cd(key->GetName());
      cout << ((TH1F*) gDirectory->Get("H10x1820a003"))->GetBinContent(5) << endl;
    f->cd();
    }
  }

• Script 2:

  The next script gets the first 10 bins of the histogram "H10x1820a00" and dumps those values to an example file named example.txt.

  {
    ofstream out("example.txt");
    TFile f("/hb/database/DQ_HISTOS/17000/DQ_17042.root");
    // Looking inside the root file
    TIter next(f.GetListOfKeys());
    TKey *key;
    while ((key=(TKey*)next())) {
      printf("Key:%s Class:%s at %d\n",
         key->GetName(),key->GetClassName(),key->GetSeekKey());
      f->cd(key->GetName());
      TH1F *h=(TH1F*) gDirectory->Get("H10x1820a003");
      out << key << endl;
      for (int i=1; i<=10;i++)
         out << h->GetBinContent(i) << endl;
      f->cd();
    }
    out.close();
  }

• Script 3:

  The following script is an evolution of the first example, being a little bit more complex but very usefull at the same time, it looks at a given directory and for each run's root file it retrieves the data of a given histogram that it contains.

  {
  
      char* dir = gSystem->ExpandPathName("/hb/database/DQ_HISTOS/17000/");
      void* dirp = gSystem->OpenDirectory(dir);
  
      if( !dirp )
      {
         printf("Failed to open directory: %d\n",inDir);
         return;
      }
  
      const char* entry;
      const char* filename;
      TString *str;
      TKey *key;
      TIter *iter;
      TFile *f;
  
      while( entry=gSystem->GetDirEntry(dirp) )
      {
         str = new TString(entry);
  
         if( str->EndsWith(".root") )
         {
            filename = gSystem->ConcatFileName(dir,entry);
            printf("File %s\n",filename);
  
            f = new TFile(filename);
  
            iter = new TIter(f->GetListOfKeys());
  
            while ((key=(TKey*)iter->Next()))
            {
               printf("Key:%s Class:%s at %d\n",
                  key->GetName(), key->GetClassName(),
                  key->GetSeekKey());
               f->cd(key->GetName());
  
  
               //INSERT YOUR CODE HERE
               if ( ((TH1F*)(gDirectory->Get("H10x1820a003")))!=NULL )
               {
                  cout << ((TH1F*)gDirectory->Get("H10x1820a003"))->GetBinContent(5) << endl;
               }
               else
               {
                  cout << "INFO: this histogram is not here!" << endl;
               };
  
  
               f->cd();
  
            }
  
            delete filename;
            delete iter;
            delete f;
         }
         delete str;
      }
  }