# Sum of weights

Dear experts,

I would like to ask you a more general question. I am wondering how the sum of weights is computed. I use as an example this sample: mc16_13TeV.364739.MGPy8EG_NNPDF30NLO_A14NNPDF23LO_lvlljjEW6_OFMinus.recon.AOD.e7421_s3126_r9364

If I assume that the sum of weights represents the number of event, how can a sum of weight=0.92, which is the sum of weights of the above sample, represent a number of events?

So, is it another "meaning" of the sum of weights in Madgrpah?

Eirini Kasimi

## Question information

Language:
English Edit question
Status:
For:
Assignee:
No assignee Edit question
Last query:
 Revision history for this message Olivier Mattelaer (olivier-mattelaer) said on 2023-01-26: #1

Hi,

Sorry but I do not know what mc16_13TeV.364739.MGPy8EG_NNPDF30NLO_A14NNPDF23LO_lvlljjEW6_OFMinus.recon.AOD.e7421_s3126_r9364
means, is this an experimental identifier from an experimental collaboration?
Independently, do you have a public link to such sample for me to take a look at them?

If your output is not in a .lhe format. then I might not be able to answer some of your question.

> I am wondering how the sum of weights is computed.

Here also I'm not sure about what information (i.e. line in the lhe file) you are talking about.

But let me go to the basics.
The weight of an event does not have a unique meaning and in madgraph we allow the user to generate events with three choice for the weight associated to an UN-WEIGTHED event sample:
1. all weight set to 1 (up to the sign if needed)
2. all weight set to the cross-section (up to the sign if needed) (so the average of the weight will give the cross-section)
3. all weight set to the cross-section divide by the number of event (up to the sign if needed) (so the sum of the weight will give the cross-section)

In general, normalization 2 or 3 are used, but one was requested by some MC generator group.

So I can not answer your answer but it certainly not an issue if the sum of the weight is not an integer.