Dear MG'ers, I'm trying to understand what is the best way to get the matrix element offline at any point in the phase-space, from a c++ code. The model I'm using is described here: https://cp3.irmp.ucl.ac.be/projects/madgraph/wiki/Models/ggHFullLoop#no1. I think that the "output standalone_cpp" functionality has problems when it comes to such an "external" model with custom definitions and functions but I'm not sure. So, my question divides into 2: (1) can MG give a c++ formatted matrix element (averaged and squared) for models which do not ship with MG by default (e.g. ggHFullLoop) ? (2) an alternative could be to link the objects coming from the generated fortran code, with a c++ code using the "extern" feature. The question is which function should be called and how to initialise the procedure ? For (2) I actually managed to do something like that after having the process generated (see code below). However, I'm not sure which of the fortran functions I should call from my c++ code. I have tried: SMATRIX1(...) --> runtime problems with initialising NTRY(2) DSIG1(...) --> same as for SMATRIX1 DSIG(...) --> executes well but seems to return sigma=0 always Any prompt help would be highly appreciated ! Cheers, Noam //////////////////////////////////////////////// //// C++ code to link with MG fortran //// //// Makefile is trivial ////////////////////// /////////////////////////////////////////////// #include using namespace std; extern "C" { void smatrix1_(double** pp, double* ans); } extern "C" { double dsig1_(double** pp, double* wgt, int* imode); } extern "C" { double dsig_(double** pp, double* wgt, int* imode); } int main() { double** pp = new double*[4]; for (int i=0; i<4; ++i) pp[i] = new double[5]; double* wgt = new double; *wgt = 1; int* imode = new int; *imode = 1; // example gg-->ttbar event, in the ggHFullLoop model with only the pseudo-scalar turned on, taken from the lhe file pp[0][0] = 0.00000000000E+00; pp[0][1] = 0.00000000000E+00; pp[0][2] = 0.40139316999E+03; pp[0][3] = 0.40139316999E+03; pp[1][0] = 0.00000000000E+00; pp[1][1] = 0.00000000000E+00; pp[1][2] = -0.25134778946E+03; pp[1][3] = 0.25134778946E+03; pp[2][0] = 0.50428638474E+02; pp[2][1] = 0.61397720859E+02; pp[2][2] = -0.18658101548E+03; pp[2][3] = 0.26623570956E+03; pp[3][0] = -0.50428638474E+02; pp[3][1] = -0.61397720859E+02; pp[3][2] = 0.33662639601E+03; pp[3][3] = 0.38650524989E+03; // DSIG(...) works but returns 0 *imode = 1; // initialize ? dsig_(pp,wgt,imode); *imode = 0; // run ? double xs = dsig_(pp,wgt,imode); cout << "xs=" << xs << endl; // DSIG1(...) issues with initialization // *imode = 5; // should give only the matrix element // double xs = dsig1_(pp,wgt,imode); // cout << "xs=" << xs << endl; // SMATRIX1(...) issues with runtime initialization // smatrix1_(pp,ans); // cout << "ans=" << ans << endl; return 0; }

When	Who	Action	New State
2016-02-15	Olivier Mattelaer	Answer	Answered
2016-02-15	Noam	Reopen	Open
2016-02-15	Olivier Mattelaer	Answer	Answered
2016-02-15	Noam	Confirm	Solved
2016-02-15	Noam	Confirm	Solved

MadGraph5_aMC@NLO

History of question #285523