Present: Bruce Baller, Eileen Berman, Giuseppe Cerati, Vito Di Benedetto, Lynn Garren, Krzysztof Genser, Herbert Greenlee, Thomas Junk, Robert Kutschke, Gianluca Petrillo, Brian Rebel, Erica Snider, Hans Wenzel Remote: Katherine Lato, others
Release and project report [Erica Snider]
Conclusions
-
no requests were raised for this week release at this meeting (one request is pending from Robert Sulej)
Tracking [Gianluca Petrillo]
-
[Robert Kutschke] the use case of fitters discarding input hits should be considered
-
[Bruce Baller] while we associate each hit to a point, tracks are 3D objects
-
[Thomas Junk] track direction might need to be inverted, and the fitter may be ask to fit both directions
-
[Herbert Greenlee] is the requirement hit/point enforced? [A] it is prescribed by policy
-
[Herbert Greenlee]
PFParticle should also be considered in the scenario [A] while PFParticle objects will remain distinct from Track, they should be able to connect tracks; this hasn't been considered so far
-
[Thomas Junk] inverting the direction may get to different fit quality and residuals, that is a different track
-
[Herbert Greenlee] approves the staged aspect of the implementation plan and that phase 1 is minimally disruptive
-
[Herbert Greenlee] what is the definition of covariance matrix? [Giuseppe Cerati] we have a start and end matrix, both 5x5 in dimension with the convention of Kalman fitter
-
[Bruce Baller] give a document about conventions, requirements and relations among data products
-
[Robert Kutschke] the track has proposed is the minimal unit, and it can be augmented by associations
-
[Robert Kutschke] considered to associate the track trajectory to the track rather than contain it
-
[Robert Kutschke] suggest to drop the concept of momentum "of the track", as opposed to a momentum at a certain point [A] trajectory and track interface access momentum point by point
-
[Robert Kutschke] is there support for biased residuals? [Giuseppe Cerati] the residuals associated to the points will be able to compute the biased residual
Conclusions
-
there are no vocal objections to the plan
-
a document describing the relations between data products and their requiremnents is needed
In-time filtering update and ideas on mixing modules in simulation + LArG4 re-structuring [Wesley Ketchum]
-
changes in CORSIKA database files allow a great speed up in event generation
-
random sampling has been implemented to extract the number of showers
-
the approach to detector simulation and charge transportation should be reconsidered with more modularity
Conclusions
-
should randomness of number of showers be enabled by default? or mandatory?
-
the changes are ready to be merged
-
to take advantage of the speed up, the new CORSIKA files need to be updated
-
CORSIKA "database" in LArSoft area needs to be updated
-
the backtracking has to be checked and potentially reconsidered
-
does MicroBooNE have resources to offer to help Hans Wenzel with this?
-
the topic is agreed, but we need to decide how to move forward
There are minutes attached to this event.
Show them.