MCFM-8.3/CHANGELOG.md

Changelog

[8.3] - August 2018

Changed

• Using seed 0 still uses a random seed. Using any seed unequal to 0 will give reproducible results even with OpenMP without further modifications.
• Improve QCDLoop 2 interface performance with OpenMP. We no longer initialize the library with each call and keep a global cache per thread.

Added

• New processes 164,169 for off-shell SM and SMEFT single-top-quark and anti-quark production, respectively. With this a new input.DAT block 'Single top SMEFT' is introduced which sets the SMEFT parameters, see mcfm.pdf. See also ref. [5].
• New dynamic scale 'fixed', which keeps the settings for the fixed scales but runs all dynamicscale code. This is mostly useful for debugging.
• Toggle for rapidity/pseudorapidity in definition of R and jet cuts, in mdata.f
• 'writetxt' option in input.DAT, which writes each histogram as a plain text file with space-separated columns.
• Add COPYING.txt GPL license file.

Fixed

• Fixed jets_opphem, which likely had no effect, depending on compiler.
• Fixed scheme in gg_2gam_v.f.

[8.2] - February 2018

Added

• Added top quark mass effects at high pT for Higgs+jet at NLO (nproc=200). Choosing mtex=0 gives the rescaling approach while mtex=100 uses the high pT expansion (valid for pT >~ 500 GeV). The low energy expansion mtex=2 can be used for pT ~< 250 GeV. See ref. [4].

[8.1] - November 2017

Added

• Electroweak one-loop corrections for $Z$, $t\bar t$ and di-jet production [1].
• $Z\gamma$ process at NNLO including anomalous couplings [2].
• $Z\gamma$ decay for H+jet production [2].
• $H$+2jet process with a finite top-quark mass and $H$+jet with finite top-quark mass effects [1]. This adds a new parameter [mtex] to the the input file, which specifies to which order in 1/mt^k (k=0,2,4) the finite part of the virtual corrections are computed.
• Support for random seeds by setting the seed value (previously [ij]) to 0.
• Support for boosted (as opposed to hadronic) definition of jettiness, which is now also used as default.
• Support for $p_T$ and rapidity ranges for most cuts in the input file.
• Native implementation of two PDF sets containing photons: mrstqed and CT14qed.
• EXPERIMENTAL: Integration routine that adaptively selects the cross section contribution with the largest integration uncertainty. It also adapts integration calls in the warmup phase and then continuously increases them to not run into grid bias problems. This ingration can be enabled by setting newIntegration to .true. in Need/mcfmmain.f. After each iteration a snapshot is saved for resumption. When readin is set to .true. in the gridinfo_logic common block, the integration will be resumed from any stage. Currently this integration mode does not stop the integration, but the user can at any point abort manually when the desired precision has been reached. It has only been thoroughly tested with $Z\gamma$ and $H+$jet. Other processes might need tweaks or specific initializations.

Changed

• Fixed FROOT implementation.
• Upgraded QCDLoop to QCDLoop 2.
• Using new random number generator (Mersenne Twister) from libstdc++ (C++11).
• Fixed bug concerning integration uncertainty of histogram bins.

Removed

• Removed flags virtonly,realonly,vanillafiles in input.DAT, as well as broken parameters for resuming integrations.
• Removed support for PDFLIB.
• Removed noomp version. If just a single thread is wanted, please set OMP_NUM_THREADS to 1.

[1] Campbell, Wackeroth, Zhou https://arxiv.org/abs/1608.03356 [2] Neumann, Williams https://arxiv.org/abs/1609.00367 [3] Campbell, Neumann, Williams https://arxiv.org/abs/1708.02925 [4] Neumann https://arxiv.org/abs/1802.02981 [5] Neumann, Sullivan