Negative Weight Treatment (Madgraph)
Table of Contents
Negative weights?
How to handle it?
- Say if there are total of 100 events.. out of them 10 have -ve weights.
- so the 10 events that have -ve weights should be cancelled using 10 +ve weights so in general at the end of day we left we 80 events…
- but we don’t just use the number 80 as it is
- We apply the weight (+1 or -1) to each event as per the weight.
- and we so the scaling like:
(Cross-sectioon x Lumi) / (Nevents - 2*NNegevents)
References
- https://twiki.cern.ch/twiki/bin/view/CMSPublic/SWGuideMCatNLOInterface
- TMVA: Negative Weights in Reader; https://sourceforge.net/p/tmva/mailman/tmva-users/?viewmonth=200901
- Statistical power of negative weights: link
- https://cds.cern.ch/record/517223/files/cer-002273178.pdf
- Mulders, M; Direct measurement of the W boson mass in $e^{+}e_{-}$ collisions at LEP; Amsterdam U. Amsterdam 2001 http://documents.cern.ch/cgi-bin/setlink?base=generic&categ=public&id=cer-002273178; http://cds.cern.ch/record/517223/files/cer-002273178.pdf; http://cds.cern.ch/record/517223/files/cer-002273178.ps.gz
- Reason of negative weights; https://arxiv.org/pdf/1908.00167.pdf
- https://arxiv.org/pdf/hep-ph/0204244.pdf
- https://indico.cern.ch/event/459797/contributions/1961581/attachments/1181555/1800214/mcaod-Feb15-2016.pdf
Enjoy Reading This Article?
Here are some more articles you might like to read next: