Error in <TTree::Branch>: The pointer specified for event is not of a class known to ROOT

ROOT
5

Hi Andre,

since you are on 6.11, you can take advantage of TDataFrame. Basically, TDataFrame is a tool that allows you to formulate your analysis or MC generation in a declarative manner. TDataFrame is rather powerful: it aims to take away cumbersome tasks from users and get out of their way. It allows also to seamlessly write parallelised code. In the following example, I provide a recipe that will allow you:

  1. Download Pythia8
  2. Build a dictionary for the Event class
  3. Build a program to create a generated dataset in root format containing pythia events. Two parameters are available: number of cpus and total number of events to be generated.

The program tdfGeneration.cpp: As you can see the overhead of TDataFrame in terms of lines of code is minimal:

#include <ROOT/TDataFrame.hxx>
#include <Pythia8/Pythia.h>

int main(int argc, char** argv){

   int nworkers = 1;
   unsigned nevents = 1000;
   if (argc == 2) {
      nworkers = std::atoi(argv[1]);
   }
   if (argc == 3) {
      nworkers = std::atoi(argv[1]);
      nevents = std::atoi(argv[2]);
   }

   // Generate the Pythias. The random seed is different for each instance
   Pythia8::Pythia pythias[nworkers];
   auto seed = 1;
   for (auto&& pythia : pythias) {
      pythia.readString("Beams:idA = 2212");				// Specifies Proton Beam
      pythia.readString("Beams:idB = 2212");				// Specifies Proton Beam
      pythia.readString("HardQCD:all = on");
      pythia.readString("PhaseSpace:pTHatMin = 80.");
      pythia.readString("Beams:eCM = 7000.");
      pythia.readString("Random:setSeed = on");
      pythia.readString("Random:seed = " + std::to_string(seed++));
      pythia.init();
   }

   // go parallel if needed
   if (nworkers!=1) ROOT::EnableImplicitMT(nworkers);

   // The "generator function"
   auto genFunc = [&](unsigned int slot) { 
      while (!pythias[slot].next()) continue;
      return &pythias[slot].event;
   };

   // Use the TDF to write out
   ROOT::Experimental::TDataFrame tdf(nevents);
   tdf.DefineSlot("event", genFunc)
      .Snapshot<Pythia8::Event*>("tree", "hardQCD.root", {"event"});
}

Now, the instruction to get Pythia8:

wget http://home.thep.lu.se/~torbjorn/pythia8/pythia8226.tgz
tar -zxf pythia8226.tgz
mkdir pythia
cd pythia8226
./configure --with-root=$ROOTSYS --enable-shared --prefix=`pwd`/../pythia
make -j 6
make install
cd ..

And now, the instruction to setup the environment and build the dictionary:

export LD_LIBRARY_PATH=pythia/lib/:$LD_LIBRARY_PATH
export ROOT_INCLUDE_PATH=pythia/include
rootcling  -f pythiaDict.cc  -rml libpythiaDict.so -rmf libpythiaDict.rootmap pythia8226/examples/main92.h ./pythia8226/examples/main92LinkDef.h
g++ -fPIC -shared -o libpythiaDict.so pythiaDict.cc -I pythia/include/ -L pythia/lib/ -l pythia8 `root-config --cflags --libs`
g++ tdfGeneration.cpp -o tdfGeneration -I pythia/include/ -L pythia/lib/ -l pythia8 `root-config --cflags --libs` -lTreePlayer

Now, to generate 100 pp collisions resulting in hard QCD processes on 8 cpus in parallel, just type:

./tdfGeneration 8 100

This should allow you to profitably use all of the cores at your disposal.

I hope that helps.

Cheers,
D

1 Like