2nd Physics Lists and Validation Tools working group meeting
Monday, 19 May 2014

Participants:
Gunter Folger, Alberto Ribon, Witek Pokorski, Vladimir Ivantchenko, Andrea Dotti , Julia Yarba, Daniel Elvira, Krzysztof Genser, Robert Hatcher, Alex Howard, Kihyeon Cho 

Minutes/Notes:
(Q) = question
(A) = answer
ACTION = todo

===============
First presentation (Intensity Frontier Physics List):

NuBeam PL questions:
Andrea (Q): Is this NuBeam PL better w.r.t. FTFP_BERT independently of nu specific requirements? 
Julia (A): From these preliminary plots this tuning shows a better agreement w.r.t. FTFP_BERT to projectile / target combination relevant for nu experiments at FNAL

Gunter (Q): This looks like model level tests, not PL. Can you comment? The results here should be the same as “model level” testing.
Julia (A): For user point of view, they use PL, not models, however in a certain sense this is similar to what in G4 we call model level.

Alberto (Q): What is the “reference flux distribution”?
Robert (A): FLUKA. 
Alberto (Q): Thus FTFP_BERT is more in agreement with FLUKA.
Alberto: Conclusions, for thin target NuBeam seems better than FTFP_BERT, but considering full simulation the FTFP_BERT is probably closer to the reference

Comments:
Gunter: This discussion is not for this WG but it is more for HAD WG.
Vladimir: In G4, differently from generators, we do not have a limited set of parameters with a physical meaning that form a phase space that can be spanned. In HAD WG there is discussion on how extend parts of processes to allow some tuning (e.g. vary a XS), for models often is more missing parts of physics and not a tuning.
Alberto: LHC experiments do not plan to vary parameters in G4 as done for generators.
Daniel: Actually in CMS there were these kind of requests
Andrea: Proposal in the HAD WG make a list of what are the parameters that can be exposed and can be tuned. 

ACTION: Krzysztof will repeat the talk to HAD WG to discuss how to propose these interfaces because these are tuning that are needed for tuning

ACTION: The two new physics lists will be compared with Version 10.0 and then to be discussed for inclusion and try to have them in the 10.1.beta release.


==========================
Second presentation (Validation Repository)

Comments:
Witek: upload intermediate solution for CERN DB. In the long term should read CERN DB
Gunter (Q): Concern on how to integrate with CDAsh, can scale with nightly system? 
Andrea (A): Probably not needed. We could foresee a special cdash “run” for upload of DB once per reference tag.
Vladimir: Concern on how to search and focus on specific result that is interested by developer. The web page needs to facilitate life of developers and reduce time spent for validation. E.g. how to search a result for a given test without interacting with several buttons/menus?

ACTION: By the middle / end of summer (end of summer student work) have the production system in place. Advertise tool as completed at the Collaboration Meeting.


===========================
Third presentation (Physics List factory)

Comments:
From several pleple: Avoid duplication of code, improve what is existing now.
Witek: Recognize this as a good plan to integrate system with GenericPhysicsList.
Vladimir: Since 9.6 variants for EM physics are not existing anymore, this work should take this into account.
Andrea: There are several technical issues to be discussed (an example MT). Some more iterations are needed.

Due to lack of time the discussion on PL factory has to be followed up at the next meeting beginning of July.

ACTION: Gunter will poll for next meeting date

ACTION:  Via mail people interested (Andrea, Gunter, Robert, Vladimir, Witek) will follow-up on this issue of physics lists.

Topic for next meeting: What are the PLs used in examples?
ACTION : Andrea will make a survey