Choose timezone
Your profile timezone:
Key4hep Discussion
→
Europe/Zurich
32/S-C22 (CERN)
Attendees: Katerina, Giulia, Juan, Swathi, Mateusz, Andre, Leonhard
Apologies:
Date: August 6, 2024
## News
## Round The Table
### Katerina
See slides
### Giulia
See slides
### Leonhard
* DD4hep lazy surfaces https://github.com/AIDASoft/DD4hep/pull/1304
* No improvement for CLD because all surfaces are then constructed by https://github.com/key4hep/k4geo/blob/main/plugins/LinearSortingPolicy.cpp
* DD4hep volume manager avoidance of duplicated entries
* https://github.com/AIDASoft/DD4hep/pull/1307
* Unfortunately also kills some wanted entries for some geometries
* Load speed increase for CLD from 56s to 49s in `Debug` build but disappears in`RelWithDebInfo`
* Rewrite parts of `TrackerBarrel_o1_v05`, `TrackerEndcap_o2_v06`
* Both have no volumes for part of the substructure
* Endcap is missing layers and rings
* Barrel is missing staves
* Not optimal for Geant4 navigation
* Also not optimal for DD4hep surface creation
* `MakePlacementPath` called for each module down the line
* Ensures that the placement hierarchy has no holes (would happen if e.g. layer is a placement but not a DD4hep `DetElement`)
* Does this by searching each part of the hierarchy by name -> search each module by name in all other modules of the layer
* Example CLD OuterTrackerBarrel outermost layer:
* without staves $n_\text{modules} = 37968$, lookup $\mathcal{O}(n_\text{modules}^2)$
* with staves $n_\text{mod-per-stave} = 84$, $n_\text{stave} = 452$, lookup $\mathcal{O}((n_\text{mod-per-stave}+n_\text{stave})\cdot n_\text{modules})$
### Andre
* DD4hep broken with root master https://github.com/AIDASoft/DD4hep/issues/1306
* https://github.com/root-project/root/issues/16167
* Investigated MCParticle <-> hit assignment
* https://github.com/AIDASoft/DD4hep/issues/1305
* Started looking into https://github.com/key4hep/CLDConfig/issues/43
* Debugging affected by broken associations for simhits--hits https://github.com/key4hep/CLDConfig/issues/48
### Mateusz
Back from vacation
* Josh gave his project presentation
* Attended Dagger meeting
* -> Use Julia release candidate version, some unstable package
### Juan
* New builds for nightlies, default builds with release mode, and there are also debug builds can be selected with `-d`
* Removed GaudiAlg, some algorithms from GaudiAlg gone (EventCounter)
### Swathi
* Investigating calibration constant for LAr, issue with LAr digitisation chain, not yet sorted out (Juraj, Brieuc)
* Segmentation changed from Phi to Theta
* Ported DDRec interface to version3 of InclinedEcalBarrel, digitiser no longer working for old geometry
* Writing proceedings for ICHEP
## AOB
* SFT BBQ, August 22, 2024
### ECFA H/E/T
## Summerstudents
* Dates:
* Katerina Kostova: 01/07 to 27/09
* Giulia Gianoli: 01/07 to 27/09
* Joshua Ott: 10/06 to 09/08
## Next Meeting
August 13, 2024
## TODO
There are minutes attached to this event.
Show them.