All events have weight = +/- 1 at NLO

Hi,

We perform an sample unweighting such that all event have the same weight (up to the sign for NLO sample).
So by design it is normal that all event have the same weights.

Now this is suspicious that you have exactly one but the normalization of those weight is subject to multiple convention and is not important as long as
the header of the file specify that those are unweighted events (up to the sign).

I’m suspecting that you look at the lhe file coming out of ma5 which has changed the normalization of the weight to one which is different of the default normalisation
that we use for the weights. You can check in the lhe produced by MG5 that indeed the weight is different of one (but all the same up to the sign)

Cheers,

Olivier

> On Dec 16, 2015, at 02:57, William Beggar <email address hidden> wrote:
>
> New question #279253 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/279253
>
> Hi,
>
> I'm generating g g > h > l+ l- l+ l- /a [QCD] using a slightly modified HEFT model and showering the result with Herwig++. I set the number of events to 1000, but other than that, everything is set to default. Then I load up MadAnalysis5 in reconstructed mode. I use the following commands:
>
> set main.fastsim.package = fastjet
> set main.fastsim.algorithm = antikt
> set main.outputfile = "output_from_ma5.lhe"
>
> Here is an event from the file:
>
> <event>
> 81 0 1.0000000E+00 1.2501400E+02 7.7909622E-03 1.1065897E-01
> 25 -1 0 0 0 0 0.00000000000E+00 0.00000000000E+00 -0.21063873291E+03 0.24494320679E+03 0.12501399994E+03 0.00000 0.00000
> -13 1 0 0 0 0 0.20571804047E+01 -0.13052531242E+02 -0.66948027611E+01 0.14813241005E+02 0.10565836728E+00 0.00000 0.00000
> 13 1 0 0 0 0 -0.60158805847E+01 0.19369634628E+02 -0.27178026199E+02 0.33912090302E+02 0.10565836728E+00 0.00000 0.00000
> -11 1 0 0 0 0 0.44253630638E+01 0.40972160339E+02 -0.68840034485E+02 0.80232490540E+02 0.00051100663E+00 0.00000 0.00000
> 11 1 0 0 0 0 -0.11812665939E+02 -0.41523220062E+02 -0.10901461792E+03 0.11725144958E+03 0.00051100395E+00 0.00000 0.00000
> 21 1 0 0 0 0 0.63648171425E+01 -0.10335512161E+01 -0.22234838486E+02 0.23399585724E+02 0.34019761086E+01 0.00000 0.00000
> 12 1 0 0 0 0 -0.00000014003E+00 0.00000107010E+00 0.00000000000E+00 0.00000107922E+00 0.00000000000E+00 0.00000 0.00000
> </event>
>
> Now, this event seems to be saying that we have an initial state Higgs which goes to 2 electrons, 2 muons and a gluon. So far so good. I understand that the third number on the first line of the event is the event weight, so this event has a weight of exactly 1. Here's another event:
>
> <event>
> 92 0 -1.0000000E+00 1.2500243E+02 7.7909541E-03 1.1066041E-01
> 25 -1 0 0 0 0 0.00000000000E+00 -0.00000000000E+00 -0.73672546387E+02 0.14509738159E+03 0.12500243378E+03 0.00000 0.00000
> -11 1 0 0 0 0 0.82857465744E+01 -0.12878288269E+02 0.55285077095E+01 0.16280918121E+02 0.00051099923E+00 0.00000 0.00000
> 11 1 0 0 0 0 -0.25284779072E+01 0.18114912033E+02 -0.19962181091E+02 0.27074562073E+02 0.00051099836E+00 0.00000 0.00000
> -11 1 0 0 0 0 0.32546413422E+02 0.23086227417E+02 -0.27244186401E+01 0.39995815277E+02 0.00051099481E+00 0.00000 0.00000
> 11 1 0 0 0 0 -0.93834238052E+01 -0.28194009781E+02 -0.58460498810E+02 0.65578811646E+02 0.00051098969E+00 0.00000 0.00000
> 21 1 0 0 0 0 -0.41323218346E+01 -0.15364439964E+02 -0.84173450470E+01 0.19220804214E+02 0.67413287163E+01 0.00000 0.00000
> 21 1 0 0 0 0 -0.54606485367E+01 0.20580186844E+02 0.52975814819E+02 0.57375621796E+02 0.56711773872E+01 0.00000 0.00000
> 21 1 0 0 0 0 -0.20772878647E+02 -0.10131204605E+02 -0.15698509216E+03 0.15886683655E+03 0.77589421272E+01 0.00000 0.00000
> 12 1 0 0 0 0 0.00000008571E+00 -0.00000010833E+00 0.00000000000E+00 0.00000013814E+00 0.00000000000E+00 0.00000 0.00000
> </event>
>
>
> Now, this event seems to be saying that we have an initial state Higgs which goes to 4 electrons and 3 gluons with a weight of exactly -1.
>
> Every other event has a weight of 1 or -1. I understand why negative events are necessary, but I'm just wondering whether or not it makes sense that EVERY event is +/- 1.
>
> Thank you for any advice.
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Great, thank you for your response. So, just to clarify, if I make something like a pz-histogram of the final state leptons out of this .lhe file from ma5 (taking the weights into account) then that should be comparable to the .lhe file from Pythia at LO? In other words, that would be a valid way of comparing NLO and LO?

Thanks Olivier Mattelaer, that solved my question.