EDM4hep Discussion

Tuesday 13 Dec 2022, 09:00 → 10:00 Europe/Zurich

Zoom

Live Notes

EDM4hep Live Notes
==================

Date: December 13
Indico: https://indico.cern.ch/event/1231126/

Connected: Tao, Wenxing, Andre, Erica, Brieuc, Juraj, Alvaro, Valentin, Frank, Sang Hyun, Juan

Apologies: Benedikt, Thomas, Gerri

## Introduction and General Points

### Upcoming workshops / conferences
https://github.com/orgs/key4hep/projects/4/views/1
   

## Progress and discussion

## CEPC Progress on extension for Drift Chamber/ClusterCounting
* See presentation
* https://github.com/key4hep/EDM4hep/pull/179/files 
* https://github.com/key4hep/EDM4hep/issues/124
* TPCHit too specific: https://github.com/key4hep/EDM4hep/issues/25
   * [ ] Rename TPCHit to TrackerRawData
   * [ ] https://github.com/key4hep/EDM4hep/pull/179
* RecDndx
   * Make components of VectorMembers (grouping them into appropriate ones)
   * Relation to Track; Does it make sense to have the relation to the track?
       * In reconstruction track comes first and in analysis the RecDndx information is used
       * "Logically" it looks the wrong way around, but in podio philosophy filling information "later" is not currently easily done.
* Do we have to revisit some of EDM4hep in light of the fact that changing quantities after the fact is not possible?
   * Could also run collections in parallel (potential issues with filtering)
   * Could use external relations/associations (gets cumbersome if there is more than two types involved) 
   * [ ] Check how this is done in (other) FWs (CMSSW, ATLAS, LHCb, Marlin)
   * Should dEdx information be removed from `Track`?

- more general discussion needed on how to deal w/ higher level auxiliary information that is added by a different algorithm
   - in EDM4hep modification of data objects not possible after they have been added to a frame
   - need to look at solutions by running experiments:
       - CMS, LHCb (Benedikt)
       - ATLAS (Graeme)
       - LCs/Marlin (Frank)

## Dual Readout with small data model extension

* sparse vector: 
   * https://github.com/HEP-FCC/dual-readout/blob/master/edm4dr.yaml
   * What is it used for?
       * Used during the digitisation, histogram containing amplitudes in timing window. Used in reconstruction
   * objectID could be replaced by relation?
       * Would need "untyped" relations, or HitWrapper
   * Specialisatoin of RawCaloHitType?
   * Discuss in one of the next meeting, or open PR for discussion

## Store yaml in output files
* https://github.com/key4hep/EDM4hep/pull/179
* Why convert to json?
* discuss next time or in the PR

## Podio
* https://github.com/AIDASoft/podio/pulls
* https://github.com/AIDASoft/podio/issues
* https://github.com/orgs/AIDASoft/projects/1/views/1

### Frame python bindings, etc.
* Merged three related PRs to make them available for the Terascale tutorial:
  * python bindings for frame (https://github.com/AIDASoft/podio/pull/343) 
  * "legacy" readers for reading EventStore based files into the "Frame world" (https://github.com/AIDASoft/podio/pull/345)
  * Make `podio-dump` work with Frames. Using the legacy readers, all files are now treated as Frame based here (https://github.com/AIDASoft/podio/pull/344)
* Plan to make a tag with these changes shortly

## EDM4hep
* https://github.com/key4hep/EDM4hep/pulls
* https://github.com/key4hep/EDM4hep/issues

## k4CLUE

* https://github.com/key4hep/k4Clue/issues/34
* Do we need an nvcc build? (Including I/O)

## LCIO tool to show missing collections in some events
* https://github.com/iLCSoft/LCIO/pull/158
* Allow converting LCIO files to EDM4hep.root where some collections are not present all the time

 

## Next meeting:
* January 17, 2023; 09:00 CET

09:00 → 09:05 Introduction

Speakers: Andre Sailer (CERN), Frank-Dieter Gaede (Deutsches Elektronen-Synchrotron (DE)), Thomas Madlener (Deutsches Elektronen-Synchrotron (DESY))

EDM4hep Github

Podio Github

09:55 → 09:56 Discussion

Speaker: Dr All