Major update: added the complete NNLO result for the reaction gg → tt ̄+ X from M. Czakon, P. Fiedler and A. Mitov arXiv:1303.6254. Now the program contain the full NNLO result for top pair production at hadron colliders. Additional functionality updates. Please see Appendix B in the program’s manual.

NOTE: Using Top++ with LHAPDF6. Top++ works (almost) out of the box with LHAPDF 6.0.5 onwards. The only change needed is to the Makefile: the path to the boost library needs to be given and linked. Here is a simple example of a modified Makefie which allows immediate use of Top++ with the new LHAPDF6 library:

CXX = g++

GSLDIR = $(shell gsl-config –prefix)

LHADIR = $(shell lhapdf-config –prefix)



CXXFLAGS = -O -I$(GSLDIR)/include -I$(LHADIR)/include -I$(BOOSTDIR)/include/boost-1_53

Please note that Top++ wouldn’t work with LHAPDF6 versions earlier than 6.0.5 due to backward incompatibility of the pdf initialisation in LHAPDF6 with respect to LHAPDF5 (it is restored starting from 6.0.5).


Previous versions:

  • Version 1.4 (Download it here)
    Functionality update: added the complete NNLO result for the reaction qg → tt ̄+ X from M. Czakon and A. Mitov arXiv:1210.6832. No other changes to the program with respect to version 1.3.1.
  • Version 1.3.1 (Download it here)
    Bug fix for ver.1.3 (in file top++.cpp): the program could call the wrong prescription for the calculation of pdf uncertainties with the NNPDF family of sets. The bug does not affect earlier versions of Top++ or any other functionality. In particular, all numbers in the manual are unaffected. We thank Sebastian Naumann-Emme for bringing it up to our attention.
  • Version 1.3 (please use ver.1.3.1 instead) Download it here.
    Added newly computed NNLO partonic reactions from M. Czakon and A. Mitov, arXiv:1207.0236 [hep-ph]: qqbar → tt + qqbar ; qq → tt + X ; qqbar’ → tt + X ; qq’ → tt + X. Adding the reaction qqbar → tt + qqbar (as defined in M. Czakon and A. Mitov, arXiv:1207.0236 [hep-ph]) completes the calculation of the reaction qqbar → tt + X at NNLO. Please note that, compared to version 1.2, version 1.3 adds few seconds to each evaluation due to the initialization of the newly added partonic fluxes qq,qq’ and qqbar’.
    Added an option that allows switching between “.LHgrid” and “.LHpdf” files in the LHAPDF interface.
    Added the symmetric prescription for computing PDF uncertainties. It is automatically invoked for the ABM11, ABKM09 and earlier Alekhin family of PDF sets.
  • Version 1.2 (Download it here)
    Major functionality update. Implemented is the exact NNLO result for the qq → tt + X reaction from the work of P. Bernreuther, M. Czakon and A. Mitov (2012).

This version allows full NNLO+NNLL phenomenology at the Tevatron and improved predictions for the LHC with respect to NLO+NNLL or approximate NNLO.

Please note: if only approximate NNLO calculation is desired (i.e. no exact NNLO), possibly including soft gluon resummation, then version 1.1 should be used.

Detailed description of the updates with respect to version 1.1 (see also the manual inside the program):

 – Remove the option Cbarqq which is now obsolete due to the inclusion of the exact NNLO qq → tt + X result.

 – The functionalities of the following options are modified with respect to ver.1.1 and lower: OrderFO, OrderRES, RESonORoff, H2qq, H2gg1, H2gg8, NNLO,NNLOonORoff, Cbargg, ECMLHC.

 – The description of the option MtopLimit is improved.

 – Included is a new set of examples, that reflects the new functionality.

 – Slight modifications to the Manual that reflect these changes.

 – Added functionality to the file top++.cpp that restricts the option MtopStep to be positive and prevents the program from possibly entering into an infinite loop.

  • Version 1.1 (Download it here)
    Minor bug fixes in the initial release ver.1.0. This is a fully functional version of the program. If only approximate NNLO calculations are needed, possibly including soft gluon resummation, then this version needs to be used.
    Detailed description of the bug fixes (see also the manual inside the program):
    Corrects the normalization of the hard matching constants H2qq, H2gg1, H2gg8 (the factor of 1/Pi^2 was missing in ver.1.0).
    Fixes a bug in the calculation of the matching of the all-order soft exponent. The bug was affecting the results in the unusual cases when lower logarithmic accuracy was requested than the fixed order one, i.e. the following three cases: NLO+LL, NNLO+LL and NNLO+NLL.
    These bug fixes do not affect any of the numbers published in this program’s manual or in the work of Cacciari et al 2011.