multiweights

Hi Jakub,

Multiweights cannot be used yet. This is an on-going development and the reference cards only include those as place holders. If you tell us more about your ETA, maybe we can help and provide beta versions to play with.

Cheers,

Benj (and Jack ;) )

> On 15 Apr 2021, at 16:25, Jakub Salko <email address hidden> wrote:
>
> New question #696588 on MadAnalysis 5:
> https://answers.launchpad.net/madanalysis5/+question/696588
>
> Dear MA team,
>
> I am using the Expert mode of MA5 v.1.9.19 to analyse a sample (stored in the root file produced by Delphes) containing multiple weights associated to each event (currently they are all weights coming form the systematics evaluation using LHAPDF6 but I will also eventually use a reweight module of madgraph). I would like to use the information from the additional weights at the MA level and according to arxiv:1808.00480, I should be able to obtain this information by using multiweights() method of the MCEventFormat class. However, calling event.mc()->multiweights().size() in the beginning of the Execute function returns zero for all the events in my sample.
> I was unable to find any examples where the multiweights() method would be used so maybe I am just not using it right in my code. Could you please help me understand why I am unable to read the weights from the root file?
>
> Thank you very much for your help.
> Best regards,
> Jakub Salko
>
> --
> You received this question notification because you are an answer
> contact for MadAnalysis 5.

Hi Benjamin,

thank you for your fast reply and sorry for the delay on my side.

I am trying to cast bounds on a large number of Wilson coefficients (corresponding to flavourful 4-quark operators in the SMEFT) based on some kinematic (angular) observables coming from events passing all the selection criteria of a recasted search. I compute these observables within MA5 and then output them into a custom file for further processing outside of MA5.
It is computationally impossible to get reasonably large datasets generated under an assumption of only one non-zero WC . My idea was to use the reweight module of madgraph instead. The additional weights calculated by the reweight module are necessary for constructing the kinematic observables. I do not need to work with the weights within MA5 besides reading them and outputting them in my custom output.
I would be happy to hear any ideas/suggestions on how or where to look at for the weight information from the root file within MA5, I would very appreciate them. Please let me know if you need any further details.

Best,
Jakub

Hi Jakub,

Most of the weight infrastructure is already there (readers, accessors, etc.). What is missing is their treatment at the cut and histogramming level, as well as the output methods. I have no idea how urgent this development is for you, but maybe we can keep in touch on skype, so that you could beta-test anything we add to the framework. Please let us know.

Cheers,

Benj (& Jack & Eric)

> On 20 Apr 2021, at 11:30, Jakub Salko <email address hidden> wrote:
>
> Question #696588 on MadAnalysis 5 changed:
> https://answers.launchpad.net/madanalysis5/+question/696588
>
> Jakub Salko posted a new comment:
> Hi Benjamin,
>
> thank you for your fast reply and sorry for the delay on my side.
>
> I am trying to cast bounds on a large number of Wilson coefficients (corresponding to flavourful 4-quark operators in the SMEFT) based on some kinematic (angular) observables coming from events passing all the selection criteria of a recasted search. I compute these observables within MA5 and then output them into a custom file for further processing outside of MA5.
> It is computationally impossible to get reasonably large datasets generated under an assumption of only one non-zero WC . My idea was to use the reweight module of madgraph instead. The additional weights calculated by the reweight module are necessary for constructing the kinematic observables. I do not need to work with the weights within MA5 besides reading them and outputting them in my custom output.
> I would be happy to hear any ideas/suggestions on how or where to look at for the weight information from the root file within MA5, I would very appreciate them. Please let me know if you need any further details.
>
> Best,
> Jakub
>
> --
> You received this question notification because you are an answer
> contact for MadAnalysis 5.

Hi Benjamin,

thank you for your answer. It is not a burning issue right now as I am also struggling with the generation of event samples to analyse. On the other hand, for the purpose of my work, it might actually be sufficient to be able to read different weights at the MA level such that I can use the weight information by myself. I would certainly like to keep in touch about the issue.

Cheers,
Jakub