Helmholtz-Gemeinschaft New Implementation Updates

mitglied der helmholtz gemeinschaft n.w
1 / 15
Embed
Share

Explore the latest updates and future implementations in the Helmholtz-Gemeinschaft, focusing on event-based and time-based data processing. Learn about current challenges and proposed solutions for optimizing data linkage and storage efficiency.

  • Helmholtz-Gemeinschaft
  • Data Implementation
  • Event-Based
  • Time-Based
  • Future Updates
rayaanacharya
devorysc Follow

Uploaded on | 0 Views

Download Presentation

Please find below an Image/Link to download the presentation.

The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author. If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.

You are allowed to download the files provided on this website for personal or commercial use, subject to the condition that they are used lawfully. All files are the property of their respective owners.

Download Presentation

The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author.

E N D

Presentation Transcript

  1. Mitglied der Helmholtz-Gemeinschaft New Implementation FairLinks and a Fast Event Builder | Tobias Stockmanns 1. Juli 2025

  2. Current Implementation Event Based EntryNr MC Track MVD Point Pixel Digi Pixel Cluster Pixel Hit 1 2 FairLink (FileID / EntryNr / BranchID / Index / Weight) DataObject (FairLink / Data) 1. Juli 2025 Folie 2 Tobias Stockmanns

  3. Current Implementation Each DataObject contains a set of links to the objects used to create the DataObject Set of links stored in each DataObject MC match between reconstructed hit and MC: follow the links (Hit->Cluster->Digi->Point) Works well for event based data For time based data it works but it is extremely slow! 1. Juli 2025 Folie 3 Tobias Stockmanns

  4. Current Implementation Time Based EntryNr MC Track MVD Point Pixel Digi Pixel Cluster Pixel Hit 1 20 32 1. Juli 2025 Folie 4 Tobias Stockmanns

  5. Future Implementation DataObjects not anymore in same entry Follow FairLinks needs permanent switch between entries very time consuming Solution: store in each DataObject all FairLinks to previous data (not only to stage directly before) Drawback: Memory and disk storage consumption rises 1. Juli 2025 Folie 5 Tobias Stockmanns

  6. Future Implementation Current Future Data Data FairLinks TRef FairLinks Manager + FairLinks stored in object Always on FairLinks Selector: Which links to store Which to ignore Separate Branch Can be disabled 1. Juli 2025 Folie 6 Tobias Stockmanns

  7. Results MCTrack Cluster MCPoint Digi 37 : MVDHitsPixel 0 : [(-1/0/1/4/2) (-1/0/36/0/1) (0/0/3/0/2) (0/20/25/0/1) (0/20/25/17/1) ] 1 : [(-1/0/1/3/4) (-1/0/36/1/1) (0/0/3/7/4) (0/20/25/1/1) (0/20/25/3/1) ] 2 : [(-1/0/1/0/1) (-1/0/36/6/1) (0/0/3/14/1) (0/20/25/12/1) ] 3 : [(-1/0/1/4/3) (-1/0/36/2/1) (0/0/3/2/3) (0/20/25/2/1) (0/20/25/19/1) (0/20/25/22/1) ] 4 : [(-1/0/1/3/2) (-1/0/36/4/1) (0/0/3/6/2) (0/20/25/9/1) (0/20/25/20/1) ] (-1/0/1/4/2) FileID EventID BranchID Index in TClonesArray 1. Juli 2025 Folie 7 Tobias Stockmanns

  8. Summary New concept for storage of FairLinks to better cope with TimeBased simulation All FairLinks directly available in each DataObject Storage of FairLinks separated from Data Connection via persistant TRef FairLinks can be disabled FairLinksManager to select what to store and what not Where to get it? Is part of FairRoot in the latest releases of the trunk in Git repository 1. Juli 2025 Folie 8 Tobias Stockmanns

  9. TimeGapEventBuilder Looks for time gaps in data stream of one sub detector Selects hits between gaps and groups them into pseudo event Takes time width of pseudo event to select data of other sub detectors depending on their time resolution to build complete event MVD STT Event 1 Event 2 Event 3 1. Juli 2025 Folie 9 Tobias Stockmanns

  10. Simulation Results 400 DPM events 50 ns average time between events complete time based simulation MVD Pixel as event builder, MVD Strip and STT as addition detectors 10 15 20 25 30 35 40 0 5 0 5 10 15 MVDHitsPixel_event_times 20 25 30 35 40 time width (ns) MVDHitsPixel_event_times MVDHitsPixel_event_times RMS RMS Mean Mean Entries Entries 45 7.992 7.992 10.37 10.37 292 292 1. Juli 2025 Folie 10 Tobias Stockmanns

  11. Time Width of Event Data 10 15 20 25 30 35 40 0 5 0 5 10 15 MVDHitsPixel_event_times 20 25 30 35 40 time width (ns) MVDHitsPixel_event_times MVDHitsPixel_event_times RMS RMS Mean Mean Entries Entries 45 7.992 7.992 10.37 10.37 292 292 10 15 20 25 30 35 40 45 10 20 30 40 50 60 0 5 0 0 0 50 10 100 STTSortedHits_event_times MVDHitsStrip_event_times 150 20 200 30 250 300 40 time width (ns) time width (ns) MVDHitsStrip_event_times MVDHitsStrip_event_times 350 STTSortedHits_event_times STTSortedHits_event_times RMS RMS Mean Mean Entries Entries RMS RMS Mean Mean Entries Entries 400 50 8.102 8.102 9.256 9.256 13.14 13.14 266.6 266.6 243 243 290 290 1. Juli 2025 Folie 11 Tobias Stockmanns

  12. Event Quality 100 120 140 160 180 200 220 20 40 60 80 0 0 1 2 MVDHitsPixel_event_EventQuality 0) Number of merged events 1) All hits in one pseudo event from one real event 2) More the one real event in pseudo event 5) Data of one event in more than one pseudo event 3 4 5 6 7 8 MVDHitsPixel_event_EventQuality MVDHitsPixel_event_EventQuality RMS RMS Mean Mean Entries Entries 9 10 1.131 1.131 1.125 1.125 424 424 100 120 140 160 180 100 150 200 250 300 20 40 60 80 50 0 0 0 0 1 1 2 2 STTSortedHits_event_EventQuality MVDHitsStrip_event_EventQuality 3 3 4 4 5 5 6 6 7 7 8 8 MVDHitsStrip_event_EventQuality MVDHitsStrip_event_EventQuality STTSortedHits_event_EventQuality STTSortedHits_event_EventQuality RMS RMS Mean Mean Entries Entries RMS RMS Mean Mean Entries Entries 9 9 10 10 1.132 1.132 0.962 0.962 1.881 1.881 2.932 2.932 421 421 689 689 1. Juli 2025 Folie 12 Tobias Stockmanns

  13. Number of Events in PseudoEvent 100 120 140 160 180 200 220 20 40 60 80 0 0 MVDHitsPixel_event_NEventsInEvent 2 4 6 8 MVDHitsPixel_event_NEventsInEvent MVDHitsPixel_event_NEventsInEvent RMS RMS Mean Mean Entries Entries 10 0.8073 0.8073 1.46 1.46 401 401 Events per PseudoEvent 100 120 140 160 180 100 20 40 60 80 20 40 60 80 0 0 0 0 1 2 MVDHitsStrip_event_NEventsInEvent STTSortedHits_event_NEventsInEvent 5 3 4 5 10 6 7 15 8 MVDHitsStrip_event_NEventsInEvent MVDHitsStrip_event_NEventsInEvent RMS RMS Mean Mean Entries Entries RMS RMS Mean Mean Entries Entries STTSortedHits_event_NEventsInEvent STTSortedHits_event_NEventsInEvent 9 10 20 0.8529 0.8529 1.318 1.318 4.502 4.502 6.784 6.784 401 401 401 401 Events per PseudoEvent Events per PseudoEvent 1. Juli 2025 Folie 13 Tobias Stockmanns

  14. Event Display 1. Juli 2025 Folie 14 Tobias Stockmanns

  15. Summary and Outlook TimeGapEventBuilder very simple event builder for first stage of processing Already quite good for MVD (> 50% clean events for MVD) Possible to use different sub-detectors as start event builders Possible to vary window size to improve purity or completeness of events Can be used for all type of detectors 1. Juli 2025 Folie 15 Tobias Stockmanns

Related


No related presentations.

More Related Content