FCC Software Meeting
Jana
Provided several tools to reconstruct calorimeter hits from geant deposits.
PR done, work ongoing.
- PrepareEmptyLarPhiEtaRCaloCellsTool: prepare the list of empty calo hits for the whole calorimeter.
- MergeHitsToCaloCellsTool: Merge CaloHits from Geant in a given cell into a single CaloHit
- CalibrateCaloCellsTool: Calibrate the merged CaloHits to the EM scale
- NoiseCaloCellsTool: Apply noise to the cells
- SaveLarPhiEtaRCaloCells: DD4HEP cell ID -> global position.
These tools are in fact very generic and could be used for all types of calorimeters.
Small reorganization discussed with Colin in this view. Still needs to be implemented before merging.
Joschka
Presented the new version of the event data model, in which most associations are replaced by relations.
Everybody agreed.
Finalizing the pull request to podio and fcc-edm today, and we merge tomorrow.
Then, Joschka will deal with the modifications in FCCSW while colin will take care of fcc-physics and heppy.
Goal: provide a new snapshot release after this is done, hopefully next week.
After the distribution of this new snapshot release, we will update the tutorials and announce.
For a full release, we will wait until we also have a new release of DD4hep (which he wants to request at DD4hep meeting, next week) because that'll allow us to use the event display.
Working on an issue reported by Colin, concerning memory-related crases when running papas (podio+fcc edm+simulation in heppy, using the python interface of podio) on the mac.
These random crashes occur every 1000 events on average.
Nicolò Foppiani and Lucas Torterotot reported similar issues on SLC6 machines (lxplus, lyon), albeit at a smaller rate.
Julia
ATS now has a web page.
Updated the DD4HEP plugin, and working on geometry building tools.
Attended a DD4HEP meeting. The DD4HEP team now uses github, and would like us to report issues with the github issue tracking system instead of Jira.
Ana
Looking at calorimeter data provided by Jana.
Valentin
Managed to install the FCCSW at DESY on SLC6 machines relatively easily.
There was only one problem: our scripts test that the user is on the CERN domain to decide to pick up the software through CVMFS.
Colin added this test to be able to install in Lyon before we start using CVMFS,
and also — maybe — to be able to install on the virtual machine.
Colin will implement a better test for the availability of CVMFS.
Zbynek & Valentin:
automatized tkLayout geometry output in XML format finished
-> work in progress on tracker geometry build based on the tkLayout XML description within the FCCSW framework
Alice
2 PRs going on,
- one in heppy : cleaning
- one in papas C++ : code review for the first version of this tool
In the process of synchronizing the python and C++ versions of papas, including the latest developments on the python side.
Clément
Provided a first version of the ttbar example on
https://twiki.cern.ch/twiki/bin/view/FCC/FccSoftwareGettingStarted
Need to discuss with Colin about the global integration of these instructions,
Noted that some people end up trying to install the FCCSW on their laptop just to generate pythia8 events,
while this could be done with fcc-physics.
Colin
Investigating the memory problem in papas / heppy.
Clean up in heppy: bug fixes, pylint, code refactoring, etc.
Helping Alice with the sync between python and C++ papas.
General plans
It would be good to organize a 1/2 day tutorial at CERN
Before that, we should organize a documentation sprint: all the interested people meet for up to a day,
and we just write the doc.
The conveners will set this up.