Opened 8 years ago
Closed 7 years ago
#366 closed defect (fixed)
Eta asymmetry in LocalKFactor with Comix RS and Amegic LO process
| Reported by: | Frank Siegert | Owned by: | Frank Siegert |
|---|---|---|---|
| Priority: | major | Milestone: | rel-2.3.0 |
| Component: | ME Generator | Version: | 2.2.1 |
| Keywords: | Cc: |
Description
(just for documentation and further reference, now that the problem is understood and fixed)
When using a setup with
- MEPS@NLO
- Comix ME generator for the subtracted real emission matrix elements (RS)
- Amegic ME generator for the additional higher-multiplicity LO matrix elements
an eta asymmetry of final state particles appears.
This has now been understood to stem from a bug in the LocalKFactor(*). The problem was that the Amegic LO process did not supply a point in colour space (because Amegic's calculations are colour-summed), but Comix needs a colour point for the calculation of the RS matrix element within the LocalKFactor.
This has now been fixed on trunk (r28915) and provided as a patch on the rel-2-2-1 SVN branch as well.
Note that the most commonly used MEPS@NLO setup (Comix RS and Comix LO matrix elements) is not affected.
(*) The LocalKFactor is a correction applied to the higher-multiplicity LO matrix elements and calculated using the next available NLO matrix element based on backward clustering. The purpose is to allow a smooth merging of these multiplicities despite their (different) accuracies.
This is fixed in Sherpa 2.2.2.