Analysis and Visualization Tools

Juraj Smieško (CERN)

Analysis tools and Productions

CERN

25 March 2026

FCC PED PS&C Groups

Group definitions during the Reference Design Phase of the FCC project

FCC PED PS&C groups

FCC PED Sub-structure

Organizational structure during the Reference Design Phase of the FCC project

FCC PED Sub-structure

Communication

Analysis tools

Analysis tools mandate

  • Lead the maintenance and development of the FCC analysis software ecosystem by exploring, validating, and integrating new technologies to enhance efficiency, functionality and user-friendliness
    • This includes, but is not limited to, utilities for machine learning, statistical combination, kinematic fitting, vertex fitting, ...
  • Develop mechanisms integrated in Key4hep to produce samples independent of the data model and easy to analyze
    • In close contact with other parties (Physics Performance, Detector Concept, MC Production ...), define the content of these samples for the various FCC applications
    • Support or perform the implementation of the producers and ensure they are included in the central workflows
  • Ensure that benchmark analyses (to be defined with the Physics work package) are centrally maintained and ready to run with various detector configurations
  • Provide user support regarding analysis and visualization tools
  • Write a comprehensive documentation for all processes, tasks, and systems under their responsibility

What is there

  • FCCAnalyses is a primary tool for the analysis at FCC
  • podio allows writing an analysis in C++ and Python
  • RDataFrame in C++ w/wo analyzers from FCCAnalyses
  • Coffea — awkward-arrays based columnar analysis framework
  • Julia: high-level, general-purpose dynamic programming language
    • Reading of the EDM4hep handled by the EDM4hep.jl package
  • FLARE — Workflow manager based on b2luigi
  • FCC Physics Events — dataset metadata catalogue
  • FCC{ee,hh}PhysicsPerformance — analysis catalogues
  • CMS Combine (in a container)

Future Plans / Challenges I.

  • Focus on the flagship analyses important for the benchmarking of the detector designs
    • Revive FCC{ee,hh}PhysicsPerformance repositories to also host benchmark analyses
      • FCCeePhysicsPerformance → FCCeePhysicsStudies
  • Expand distributed computing options
    • Allow running on GRID, Slurm or other distributed platforms
    • Either from within FCCAnalyses or from outside FCCAnalyses (FLARE, law, …)
  • Better ML support
  • Work on overcoming software ecosystem silos
    • Key4hep, PyHEP, DiracOS, Rucio, …
    • Streamline NTuple production mechanism
    • Improve the performance of the podio::DataSource
    • Invest time in tooling/approaches which can bridge the gap

Future Plans / Challenges II.

  • Make FCCAnalyses more robust and capable
    • Overhaul the analyzers to support EDM4Hep objects (normal and decayed), ROOT vectors, std::vectors
    • Merge staged and histmaker styles together
    • Bring Combine to the Key4hep stack
    • Revive the plotting capabilities
    • Support web based execution environments, e.g. SWAN
  • Keep up with the improvements in ROOT
    • Support processing of RNTuple files
    • ROOT's "R" classes
  • Utilize GPU resources through a specialized analyzers
  • Develop robust APIs for dataset metadata retrieval
  • Investigate and benchmark running of FCCAnalyses on non-CERN based computing resources
    • Datasets available locally or using Xrootd
  • Improve handling of metadata between analysis stages, support multiple event weights
  • Clean up the authorship and referencing

Visualization tools

Visualization tools mandate

  • Ensure that adequate visualization tools covering the needs of different purposes are available
    • The different purposes include: development and debugging of geometries and simulations, development and debugging of reconstruction algorithms, outreach
  • Provide user support regarding analysis and visualization tools
  • Write a comprehensive documentation for all processes, tasks, and systems under their responsibility

What is there

  • Phoenix@FCC
    • Developed under HSF, several detectors participating
    • Primarily targeted towards outreach, even has VR/AR capabilities
    • Web based, convoluted development / packaging
  • eedE
    • Event data explorer
    • Web based, Pixi.js as a rendering engine
  • Pandora Monitoring
    • Traditional ROOT based application
    • Heavily tied to Pandora
  • k4CEDViewer
    • Fresh port of CED Viewer form iLCSoft
    • Traditional application
  • JSROOT
    • Part of the ROOT project
    • Primarily visualizes histograms
    • Can also visualize detector geometries
    • Web based
  • Geant4 Qt visualization
    • Well established tool for Geant4
    • Traditional Qt based application
  • Other specialized tools …
    • calodisplay — ALLEGRO event display
    • FCCAnalyses — analysis graph generation

Plans / Challenges

  • No single tool for debugging and development of simulation, reconstruction (and analysis)
    • Support the usage and develop / contribute to development of existing tools
    • Avoid implementing specialized tools targeting specific sub-detector or use cases
  • Join the geometry converter efforts in Sim/Reco
    • ROOT to glTF conversion rudimentary
    • pyg4ometry seems promising
  • HSF Visualization not active anymore
  • Investigate where the interoperability of the OS graphics stacks can be improved
    • Wayland replacing Xorg
    • Xorg/OpenGL windows on MacOS