Merging

  1. Parameters taken from Pythia
  2. Additional Parameters

The merging procedure implemented in Vincia is (mostly) the same as the CKKW-L merging in Pythia. Therefore, most of the parameters are taken over from Pythia. A list of the used parameter is here. For how to use merging in Vincia have a look at the example program vincia26.cc.

Note that the merging has so far only been tested for e+ e- → Z → jj + jets and pp → Z + jets.

Parameters taken from Pythia

This is a list of the parameters taken over from the CKKW-L merging in Pythia. Note that we are currently only handeling doPTLundMerging. Please see the documentation in Pythia for more information about these parameters.

Merging:doPTLundMerging
Merging:Process
Merging:TMS
Merging:nJetMax
Merging:includeWeightInXsection
Merging:doXSectionEstimate
Merging:enforceCutOnLHE
Merging:applyVeto
Merging:mayRemoveDecayProducts
Merging:nRequested
Merging:muFac
Merging:muRen
Merging:muFacInME
Merging:muRenInME
Merging:unorderedScalePrescrip
Merging:unorderedASscalePrescrip
Merging:unorderedPDFscalePrescrip
Merging:incompleteScalePrescrip

Additional Parameters

In addition to the above parameters, we introduce a few new ones.

flag  Merging:useMECs   (default = off)
Apply matrix-element corrections to the splitting probabilities when a path is chosen during the merging.
Note 1: This option is only available for the processes and orders, for which Vincia has the respective matrix elements.
Note 2: For Merging:nonShowerPathsOnly = on this option should be switched off, as otherwise it will only slow down the code, but does not have any effect.

flag  Merging:showerPathsOnly   (default = off)
Only include events that can be reproduced by the shower with an ordered sequence of branchings scales, i.e. all events with only unordered and/or incomplete paths are assigned a weight of zero. Note that this includes the ordering with respect to the factorization scale (set dynamically) for all event samples.

flag  Merging:nonShowerPathsOnly   (default = off)
With this option, only those events are included that can not be reproduced by the shower in a sample with lower multiplicity. For the different event samples this means:
1) Merging:nRequested = 0: all events are accepted. Use Merging:enforceCutOnLHE = off and Merging:TMS = eCM when processing this sample in the event generation.
2) Merging:nRequested = 1: only events, where all scales are above the factorization scale are accepted. Use Merging:enforceCutOnLHE = off and Merging:TMS = eCM when processing this sample in the event generation.
3) Merging:nRequested = 2..n: only events, where all paths consist of unordered branchings scales are accepted. Note that in this sample no check with respect to the factorization scale is applied, as the shower can reproduce these events in the lower multiplicity samples. Depending on the choice of Merging:enforceCutOnLHE and Merging:TMS, the overlap is removed differently. For Merging:enforceCutOnLHE = on and Merging:TMS = TMS the overlap is removed by rejecting events with branchings larger than the merging scale in the samples with Merging:nRequested = 2..n-1. For Merging:enforceCutOnLHE = off and Merging:TMS = eCM the overlap is removed such that the lower multiplicity samples fill as much phase space as possible.
For including the samples without a merging cut, simply use Merging:TMS = 0 when running the cross section estimates.
Note 1: This option should not be used with Merging:TMS = eCM if incomplete histories are present, e.g. for W production.
Note 2: eCM acts as a placeholder for a larger number. Have a look at the example program vincia26.cc.

flag  Merging:checkShowerCutoff   (default = off)
Only used in combination with Merging:nonShowerPathsOnly = on. If this flag is switched on, any event that includes a branching scale below Merging:showerCutoff is assigned a weight of zero.

parm  Merging:showerCutoff   (default = 0.0; minimum = 0.0; maximum = 10.0)
Shower cutoff, used if Merging:checkShowerCutoff = on.

flag  Merging:useEffectiveScaleInUnorderedEvents   (default = on)
If switched on, and effective scale in unordered events is used and Merging:unorderedScalePrescrip is ignored. For more information on the effective scale see here todo.
Note 1: This option is only available for the processes and orders, for which Vincia has the respective matrix elements.
Note 2: This option has to be chosen if ordered matrix-element corrections are switched on and combined with merging.

parm  Merging:unorderedScaleVariation   (default = 1.0; minimum = 0.1; maximum = 10.0)
For Merging:useEffectiveScaleInUnorderedEvents = on this parameter is used to vary the effectice scale in unordered events. It represents the prefactor for the effective scale in GeV.