Software Download

The current release is Version 2.7.1, released on 05 August 2015.

Version 2.7.1

Source tarball: VBFNLO-2.7.1.tgz August 2015)

For installation instructions see Documentation > Installation

For a list of known issues and fixes see Download > Known Issues

GitHub Repository

You can also get recent versions from our public repository at GitHub.

The tagged releases are identical to the tarballs. In the future, we will put small bug fixes to the code there also between releases, so you have faster access to those.

Previous Versions

Version 2.7.0

Source tarball: VBFNLO-2.7.0.tgz (April 2014)

Version 2.6.3

Source tarball: VBFNLO-2.6.3.tgz (June 2013)

Version 2.5.3

Source tarball: VBFNLO-2.5.3.tgz (September 2011)

Version 2.0.3

Source tarball: VBFNLO-2.0.3.tgz (November 2009)


All publicly available program versions since version 0.9 can be found here.

The Changelog from version 2.0.1 on can be found here.


VBFNLO is distributed under the GNU General Public License (GPL) version 2. This ensures that the source code will be available to users, grants them the freedom to use and modify the program and sets out the conditions under which it can be redistributed. However, it was developed as part of an academic research project and is the result of many years of work by the authors, which raises various issues that are not covered by the legal framework of the GPL. It is therefore distributed together with a set of guidelines, which were originally formulated and agreed on by the MCnet collaboration for event generator software.

Recent Changes

Changes in Version 3.0.0 beta 5 (2018-02-06)

  • Major Enhancements
    • Large speed improvement in real-emission calculation of Hjjj production (processes 110-117)
  • Bug fixes
    • Fix two bugs in NLO calculation of Hjjj production (processes 110-117)
    • Fix number of active flavours in BLHA interface

Changes in Version 3.0.0 beta 4 (2017-04-18)

  • Bug fixes
    • Fix bug in NLO calculation of $Z\gamma jj$ production (processes 290,291) when including anomalous couplings
    • Fix bug in event output for $Z\gamma jj$ production

Changes in Version 3.0.0 beta 3 (2017-01-18)

  • Major Enhancements
    • VBFNLO is now published also at
    • Processes formerly generated by running ggflo now use the binary vbfnlo as well
    • There is now a quick low-statistics version of the regress tests. They require the pytest package on python3. Run them using cd regress; py.test
  • Updates and new features
    • Add dimension-8 operator OS,2 and adjust K-matrix unitarization accordingly
    • Support for parallelization using MPI via –enable-MPI. See manual for details.
  • Other additions and bug fixes
    • Fix bug in BLHA interface for same-flavour boson decays
    • Fix PDF convolution in real-emission part of VBF-Zjj
    • Fix bug in event output for QCD-ZZjj and -ZAjj
    • Show timing information by default
    • Add git hash of source version to runtime version output
    • Fix compilation issue when flag –disable-nlo is used
    • Fix warnings for automake version >= 1.15
    • Tested with gfortran 5.4 and 6.2
    • Redirect LoopTools output for VBF-H EW corrections to LoopTools.out
    • Removed compilation flag –enable-spin2. The spin2 code is now always included at compile time and can be switched on/off at runtime as before
    • Reduce memory consumption of Histogram routine. Better scaling for many histograms.

Changes in Version 3.0.0 beta 2 (2015-10-30)

  • Add possibility for Breit-Wigner or Passarino CPS scheme distribution of Higgs boson (VBF-H only, process IDs 100-117,1010)
  • Fix bug in K-matrix unitarisation of $W^+W^-jj$

Changes in Version 3.0.0 beta 1 (2015-08-05)

  • Major enhancements (for details see manual):
    • Several new processes are available:
      WWj, ZZj production including the loop-induced gluon-fusion contribution
    • Interface according to the Binoth Les Houches Accord (BLHA) added for all VBF processes with fully leptonic decays
    • K-matrix unitarisation procedure implemented for the two dimension 8 operators O_{S,0} and O_{S,1}.

Changes in Version 2.7.1 (2015-08-05)

  • New processes:\\QCD-ZZjj, QCD-Zgammajj, VBF-Zgammajj, Z(→l+l-)Z(→nu nubar)gamma
  • Higgs decays for VBF-HHjj process
  • Enable linking with LHAPDF v6
  • Fix PDF convolution in QCD-(W-/W-Z/W-gamma)jj
  • Fixed missing entry of VBF-HHjj as VBF process in cut routine
  • Fixed missing axis labels on 2D histograms

Changes in Version 2.7.0 (2014-04-15)

  • Major enhancements (for details see manual):
    • Several new processes are available:
      Wgammajj in VBF, HHjj in VBF, W, Wj, WH, WHj, pp→Spin2jj in VBF (with Spin2→WW/ZZ→leptons), several QCD induced processes (WZjj, Wgammajj, same sign WWjj, Wjj).
      Anomalous gauge boson couplings are supported in VBF Wgammajj, WH, WHj.
    • Semileptonic decay modes are available in the following diboson, triboson and VBF diboson plus two jets processes: VV, VVV, VVgamma, VVjj with V=W,Z. Additionally, the Higgs production processes pp → H → WW/ZZ in VBF with semileptonic decays are available as well.
      The processes with semileptonic decays support anomalous gauge boson couplings.
    • The VBF VVjj processes can be run taking a second Higgs resonance into account (simplified 2HDM).
    • Anomalous gauge boson couplings are now available in all triboson (plus jet) processes and all VVjj production processes via VBF.
    • A larger set of operators leading to anomalous gauge boson couplings is included in VBFNLO.
    • Added ability to use anomalous HVV couplings in Higgs production via gluon fusion, where the Higgs decays into WW or ZZ (prodIDs 4105-7)
  • Further updates and new features:
    • It is now possible to specify a desired number of unweighted events for event output in vbfnlo.dat.
      In case VBFNLO is not able to achieve the desired number of unweighted events due to single events with very large weights it's also possible to opt for “partially unweighted events”, where most of the weights will have weight 1, and a few will have larger weights.
    • Added 'any leptons' switch (LEPTONS = 98/99), allowing for all possible lepton final state combinations of 2 or 3 generations. See manual for details.
    • m_ll cut can be switched between cutting only opposite-sign lepton combinations or all combinations
    • H→bb decay mode for Higgs boson production processes: jet cuts are applied to final state b-quarks
    • All default values of anomalous couplings are set to zero to avoid confusion
    • implemented staggered pT cuts for leptons and photons
  • Other additions and bug fixes:
    • Phase space improvements to increase unweighting efficiency for event output in several VBF, di- and triboson processes
    • Phase space optimizations for Z decays in VV and VVjj
    • Path to procinfo.dat is no longer hardcoded, now it's also searched in the input directory
    • Increased allowed file name length for slha file
    • Print out values of g, g' used in calculation
    • in raw histogram data output: include gnuplot command files
    • New command line option –version : provides version information of VBFNLO
    • add workaround for enormous memory consumption during compilation with gfortran 4.8
    • event output for W-W-jj in VBF: color information fixed
    • Fixed bug in WWZ virtuals implementation – new results are about a percent lower
    • Several dynamical renormalization and factorization scale choices corrected which gave wrong results for the real emission part of NLO calculations: min(pT_jet) (ID 2), min(ET_bosons) (ID 7)
    • fix particle-antiparticle-assignment in W+jj, Zjj
    • event output for W-W+W-: fixed momenta assignment
    • Fixed line length for ifort
    • default value for cut MDIJ_MIN is changed to 0 GeV
    • works with LHAPDF version 6
    • output of per-bin error in histograms now available in the ROOT output of histograms defined in histograms.F

Changes in Version 2.6.3 (2013-06-21)

  • Fixed particle ID in LHE output for process 191 (Spin2→gamgam in VBF)
  • Fixed bug in couplings for spin-2 VBF→ZZ→llnunu (process 211)
  • Fixed bug with pdf sum for LO+j VBF processes
  • Fixed several bugs in les houches output concerning tau mass output and helicity output
  • Added tau mass output for LHE to all processes
  • Fix linking HepMC on some newer distributions
  • In pp→Hjjj: fix PDF sum if nfl<4
  • In ggf Higgs processes: remove debug output when using event output
  • Process 191 is now treated as VBF process in JetVeto-cuts
  • LHE output for process W-W-jj: parton and beam particle IDs fixed
  • Add higher-order corrections to H→gg partial width
  • Fixed particle ID in LHE output for gg→VV processes
  • Fixed bugs in electroweak corrections to VBF-H
  • For event output, the default choice is now to generate unweighted events.
  • Consistent use of nflVBF for Hjjj processes

Changes in Version 2.6.2 (2012-12-07)

  • Fixed lepton assignment in H→(WW,ZZ)→4l decays.
  • Fixed event output for processes with more than one phase space.
  • For the Ajj production in VBF (procID 150) a different form factor is used with respect to the other processes.
  • Phase space optimizations in WWW, ZZW and ZZZ production
  • Fixed bug in dipole subtraction for ZZZ production
  • Phase space optimization for VBF→4l including spin-2 case
  • Improved output of event numbers written to event file
  • Faster code for process 191 (spin2 → gamma gamma)

Changes in Version 2.6.1 (2012-08-24)

  • Fixed problem with TREEFAC for anomalous Higgs couplings
  • Fixed problem with root-only histograms (only histograms in rootusershist.cpp were affected)
  • Fixed coefficient of FB_ODD in a3_HZZ coupling
  • Fixed symmetry factor in 2106 (pp → HAjj → ZZAjj → llllAjj) and 4106 (gluon fusion pp → Hjj → ZZjj → lllljj) for identical final leptons
  • Increased allowed path length for input/output files

Changes in Version 2.6.0 (2012-07-20)

  • New processes added (see manual for details):
    • W+W+jj and W-W-jj via VBF: process IDs 250-260
    • diboson processes WZ, ZZ, WA, ZA, AA: process IDs 310-370
    • triboson + jet processes WAAj: process IDs 800-810
    • gluon-induced diboson production WW, ZZ, ZA, AA: process IDs 4300-4370
  • Updates and new features:
    • Anomalous couplings included in vector boson production via VBF: process IDs 120-150
    • Spin-2 model included in diboson + 2 jet production via VBF: process IDs 200-240
    • The switch FERMIONLOOP controls the inclusion of the gluon-induced higher order processes to diboson production: process IDs 300-370, 4300-4370
    • Anomalous couplings (VVV and HVV) included in diboson production: process IDs 300-370, 4300-4370
  • Other additions and bug fixes:
    • Added input HWIDTH: the Higgs width can now be fixed by the user
    • When anomalous VVV couplings are used, the Higgs width is now calculated with the appropriate anomalous HVV couplings in relevant processes
    • Altered histograms to correct the errors and allow the user to control the smearing between bins
    • Altered formfactor momentum dependence for anomalous gauge boson couplings
    • Increased allowed width of virtuality of resonant bosons in phasespace generation affected processes: 106,117,116,117,2107,4107,211
    • Fixed compilation error to allow both HepMC and ROOT to be linked
    • Streamlined calls to LHAPDF (no change to results)
    • Fixed compilation error with LHAPDF for newer versions of gcc
    • Fixed bug in print out of alfa_s for scale choice 6 (no change to results)
    • Fixed problem in TREEFAC for anomalous HVV couplings, and separated into TREEFACZ and TREEFACW
    • Fixed problem in MASS_SCALE_KAPPA for anomalous VVV couplings - separated into MASS_SCALE_AKAPPA and MASS_SCALE_ZKAPPA
    • Fixed bug in HVV anomalous coupling formfactor F2
    • Input grid files are now taken from the –input=INPUT directory: output files still written to the working directory
    • Added ability to include electroweak loop squared contributions in SUSY (switch LOOPSQR_SWITCH)
    • Added ability to switch between tree-level and corrected Higgs masses in electroweak loop corrections (switch MH_LOOPS)
    • Bug fixed in virtual contributions to H+3j processes: process IDs 110-117
    • The main files, vbfnlo_main.F and ggflo_main.F, have been moved from src/ to lib/ – this is to enable compilation with newer versions of gfortran

Changes in Version 2.5.3 (2011-09-26)

  • Fixed implementation of anomalous couplings in diboson plus jet processes (610 - 640)
  • Fixed problem in ROOT histogram for invisible particles (neutrinos)

Changes in Version 2.5.2 (2011-08-26)

  • Fixed write-out of $\alpha_{s}$
  • Fixed bug in HepMC output
  • Fixed bug in $H \rightarrow ZZ \rightarrow 4l$
  • Fixed bug in anomalous $WW\gamma$ coupling for $VVV$ production
  • Stopped calculation of diboson plus jet processes with anomalous couplings while we fix a bug in their implementation

Changes in Version 2.5.1 (2011-07-21)

Fixed bug in real emission part of VBF processes.

Changes in Version 2.5.0 (2011-07-20)

  • New processes added (see manual for details):
    • H + photon + jj: process IDs 2100-2107
    • photon + jj: process ID 150
    • W + photon + j, W + Z + j: process IDs 610-640
    • WW + photon, ZZ + photon: process IDs 460,470
    • WZ + photon: process IDs 480,490
    • W + 2 photons, Z + 2 photons: process IDs 500-521
    • 3 photons: process ID 530
  • Updates to existing processes:
    • Anomalous quartic gauge boson couplings for VVV production
    • Extension to the MSSM (via FeynHiggs, SLHA input, user input)
    • Electroweak corrections to VBF Higgs production in the SM and MSSM (requires link to LoopTools)
    • Triple vector production in Higgsless KK
  • Other additions:
    • Ability to vary histogram range via input histogram.dat file
    • More cut options implemented
    • New drawing option
  • Bug fixes:
    • Bug fixed in readinput routine
    • Correct evolution of bottom mass for gluon fusion
    • Allow low $p_{Tj}$ for Higgs production via VBF
    • Updates to LHAPDF interface
    • ROOT interface updated

Changes in Version 2.0.3 (2009-11-25)

Fixed bug and improved mapping in WWW phase space

Changes in Version 2.0.2 (2009-04-08)

Bug fixed in the real emission part of the WWjj process (VBF process)

Changes in Version 2.0.1 (2009-02-25)

Fixed –disable-NLO vs –disable-nlo inconsistency in the manual. Manual now consistent with configure using –disable-NLO

Release of Version 2.0.0 (2008-11-27)

Known Issues

Mac OS X:

Compilation of VBFNLO on MacOSX only works with static libraries. This can be requested with the configure-option --enable-shared=no. With this option the current version of Vbfnlo has been compiled successfully on different versions of MacOSX between 10.6.8 and 10.12.3 using GFortan 5 or 6 from Homebrew. However, linking with the library LoopTools, which is only needed for the calculation of the electroweak corrections in the Hjj process, is known to fail.

32-bit compilers:

If you want to compile VBFNLO with a 32-bit compiler (e.g. ifort) on a 64-bit machine, the following configure command should work:

FC=ifort CFLAGS=-m32 CXXFLAGS=-m32 LDFLAGS=-m32 ./configure

followed by your standard configure options (e.g. --prefix=)

VBFNLO on clusters:

When VBFNLO is linked to LHAPDF version 5 it can require more memory than some clusters allow (although VBFNLO version 2.7.0 needs significantly less memory than previous versions). The amount of memory required can be decreased (slightly) by only enabling certain processes at compilation – e.g.

./configure --enable-processes=vbf

LHAPDF 5 can also be reconfigured such that it only stores one PDF set per initialisation (in the LHAPDF code, set nmxset=1 in src/ or configure with the option -with-max-num-pdfsets=1, as described here).

Alternatively, the new LHAPDF version 6 can be used, which is rewritten in C++ with dynamical memory allocation. This new version needs much less memory than previous LHAPDF versions.

Fixed in version 2.7.1

The full list of bugfixes can be found in the ChangeLog.

NLO calculations of W-jj, W-Zjj, W-gamma jj production

A bug in the PDF convolution of the QCD-induced production processes $W^-jj$, $W^-Zjj$ and $W^-\gamma jj$ has been fixed. The size of the effect, which can be observed both in the real-emission part of the NLO calculation and the corresponding LO+jet processes, can reach several percent, depending on the parameter settings.

VBF cuts for HHjj

VBF cuts now also apply to VBF production of a Higgs pair.

Fixed in version 2.7.0

Reading in slha-file fails sometimes: FIXED IN VERSION 2.7.0

In version 2.6.3 VBFNLO crashes on some computers when reading in slha-files. This will be fixed in the next release. If you stumble on this problem, please contact us for a solution.

Color flow in LHE output for process W-W-jj in VBF: FIXED IN VERSION 2.7.0

In version 2.6.3 the color flow information in Les Houches event output is set up incorrectly. This issue will be fixed in the next release. If you need a patch beforehand, please contact us.

gfortran 4.8: FIXED IN VERSION 2.7.0

Unfortunately, the gfortran versions 4.8.x (which are included e.g. in openSUSE 13.1 and Ubuntu 13.10) seem to have a bug during optimization, resulting in a much longer compile time of the VBFNLO loop functions. Additionally, up to 3 GB of RAM are needed during compilation.

The next release will include a workaround for the memory requirements, and will compile significantly faster than previous releases (however still not as fast as with gfortran 4.7).

Fixed in version 2.6.3

Les Houches event output: FIXED IN VERSION 2.6.3

There are some small problems in the les Houches output of the current VBFNLO release:

  • The inclusion of the tau mass gives wrong particle momenta for the H→WW and H→ZZ processes.
  • The helicity information output is wrong for all Higgs decay processes and processes with a final state photon.

These issues will be fixed in the next release.

Fixed in version 2.6.2

ZZZ production (process 450): FIXED IN VERSION 2.6.2

The QCD real emission part of the NLO computation gives no reliable result due to a problem with the dipole subtraction. This issue will be fixed in the next release.

LesHouches event output for processes with more than one phase space: FIXED IN VERSION 2.6.2

At the moment the fraction of events coming from the different phase spaces is not sampled correctly (process list: WA, ZA, WA+j, ZA+j, ZZZ, WWA, WAA, WZA, ZAA, AAA, WAA+j). This will be fixed in the next release.

Distributions and cross section for H→WW in VBF and GF: FIXED IN VERSION 2.6.2

In the current release the distributions and cross section for Higgs decaying in W bosons in vector boson fusion and gluon fusion production processes (IDs 105, 115, 2105, 4105) are slightly off. This will be fixed in the next release.

Fixed in version 2.6.1

User generated ROOT histograms: FIXED IN VERSION 2.6.1

The NLO histograms for the root-only histograms do not include the virtual contributions. This will be fixed in the next release.

Symmetry factors for processes 4106 and 2106: FIXED IN VERSION 2.6.1

Processes 4106 ($pp \rightarrow Hjj \rightarrow ZZjj \rightarrow \ell_{1}^{+} \ell_{1}^{-} \ell_{1}^{+} \ell_{1}^{-} jj$ via gluon fusion) and 2106 ($pp \rightarrow H \gamma jj \rightarrow ZZ \gamma jj \rightarrow \ell_{1}^{+} \ell_{1}^{-} \ell_{1}^{+} \ell_{1}^{-} \gamma jj$ via vector boson fusion) are incorrect a factor of 2 if identical lepton pairs are chosen. (This is the default setting.) This will be fixed in the next release.

Anomalous HVV couplings: FIXED IN VERSION 2.6.1

When working with anomalous HVV couplings in VBF processes, the factors multiplying the SM HWW and HZZ couplings are not properly taken into account if the default settings (TREEFACW=TREEFACZ=1) are not used. This problem will be fixed in the next release. To fix this problem

  • Download and unpack the file qqhqq.tar.gz
  • Place the new file in the amplitudes/vvjj/ subdirectory
  • Do make all install (there is no need to make clean beforehand)
Last modified:: 2011/03/07 16:01