Real number of events generated after a cut

Hi,

Before answering this question, it is important to know what you want to do with this number.
This number is actually not physical at all and should be use only to compare the efficiency of the event generation.

Cheers,

Olivier

On Apr 23, 2014, at 10:51 PM, alefisico <email address hidden> wrote:

> New question #247527 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/247527
>
> Hi all
>
> Sorry for the clarification question.
> I am generating 100k events for a certain process and I am applying a cut in the HT of 500 GeV. My jobs works fine and at the end I got 100k events. I think all my 100k will have a HT bigger than 500, right? If that is the case, how I can know the number of events initially generated in order to have 100k events with this cut?
>
> thanks again for your help!!
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

well, I generated first the same sample without any cut but for our analysis we got poor statistics. So I want to generate more events with a cut in the HT to increase our statistics (otherwise I should generated 10 times the number of events that I generated before). But now, of course in order to weight our events properly I need to know the "original" number of events.

I hope that this will help.

cheers,

Still not sure to understand what you want.

But the only information that is physical is the cross-section before and after cut.
So the number of events that you should generate without cut to have the same number of event at the end is given by
N_orig = xsec_orig / xsec_new *N_new

Cheers,

Olivier

On Apr 23, 2014, at 11:26 PM, alefisico <email address hidden> wrote:

> Question #247527 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/247527
>
> Status: Answered => Open
>
> alefisico is still having a problem:
> well, I generated first the same sample without any cut but for our
> analysis we got poor statistics. So I want to generate more events with
> a cut in the HT to increase our statistics (otherwise I should generated
> 10 times the number of events that I generated before). But now, of
> course in order to weight our events properly I need to know the
> "original" number of events.
>
> I hope that this will help.
>
> cheers,
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Thanks for your patience. I think this is the number that I need, but just to make sure I will try to rephrase my question.

Let's say that without any cut I generate 100k events. Applying some selection I ended up with only 200 events that corresponds to something like 0.2 pb-1. If that is the case, I should generate like a million events to have more events to look at. Then I think will be useful to generate more events but now with an initial cut, to make sure that at the end I have a reasonable number of events to study.
Then, if I generate 100k events with a cut in the HT to make sure that I will not end up studying 200 events, to compare with data at 20 pb-1 I need to know how many generated events I start with, and I don't think that I can use anymore the 100k events with cut in the HT.

hopefully now what I need makes more sense.

thanks again!

Thanks Olivier Mattelaer, that solved my question.