ALICE Offline Week – 22 Oct Visualization of embedding Matevz Tadel, CERN Adam Kisiel, Ohio State University
ALICE Offline Week – 22 Oct Outline Embedding process description Requirements Proposition of features Plans for implementation Presentation of implemented features
ALICE Offline Week – 22 Oct Embedding goals The final goal is to study and quantify the effects of resolution and efficiency by performing embedding in the real data We need to make sure that the software tools to do that are available and ready. We must test whether our software tools can do that without compromising quality and introducing artificial distortions in the studied quantities Visualization is an important step to fulfill the goals
ALICE Offline Week – 22 Oct Embedding procedure Raw data background Monte-Carlo simulationRaw data signal “Fake” SDigits Simulated SDigits AliESD Summed SDigits Efficiency correction Kinematics information ConvertRaw2SDigitsConvertRaw2SDigitsSimulation Merge Merge OR Merge Reconstruction Analysis results Effiency&CorrectionFramework High-level analysis code
ALICE Offline Week – 22 Oct Embedding status Embedding has been tried and confirmed to work last year. Since then it has been a continuous effort to ensure it is kept in that state. Most recent tests: last week, AliRoot v4-15-Rev-06 for central barrel detectors (ITS TPC TRD TOF). Tested “real-life” scenario of embedding Pythia pp events into heavy-ion event background (Therminator PbPb midcentral events) as well as merging two pp events
ALICE Offline Week – 22 Oct Embedding visualization For embedding visualization is an important tool; it helps in understanding the effects of merging two signals and should be helpful in debugging Requirements for the visualization software are different from normal ones – the objective is to analyze the behavior of a rather small “signal” against an overall background It is important to compare the “bare” and “embedded” signal reconstruction
ALICE Offline Week – 22 Oct Software requirements The following “use cases” were considered when identifying tasks for embedding visualization Looking for a low multiplicity signal in a high multiplicity background – visual highlighting should reflect the origin of the signal Studying detector response – need to visually distinguish hits/digits/clusters from various sources as well as “merged” ones Comparing “bare” and “embedded” signals requires simultaneous inspection of two or more events
ALICE Offline Week – 22 Oct Software tasks The following software tasks for alieve were identified: Possibility to load more than one event into alieve in such a way that each event is a separate object Provide a way to show two events simultaneously “side-by-side” in a synchronized way Visually differentiate hits/digits/clusters/tracks based on the event they come from Visually differentiate hits/digits/clusters/tracks within a single event, based on the stream they come from Highlight “merged” hits/digits/clusters/tracks
ALICE Offline Week – 22 Oct Two-event display two events as separate objects requires patching ROOT and alieve
ALICE Offline Week – 22 Oct Ideas for two-event display Visually differentiate clusters/tracks from events Test the display for large-multiplicity Pb-Pb event, to make sure that the “signal” is still visible – reduce the displayed information for the background only to the necessary Provide a way to cut on the displayed data, to further reduce the amount of information: physics cuts and geometrical (acceptance) cuts Show Kinematics/full information only for signal
ALICE Offline Week – 22 Oct Finding a needle in a haystack embedded signal full PbPb event only clusters small marker
ALICE Offline Week – 22 Oct “Side-by-side” view mock-up
ALICE Offline Week – 22 Oct Ideas for the side-by-side view Synchronize data highlighting and selection on both views Lock the camera (zoom, rotation, shift) for both views so that actions in one view are automatically mirrored in the second Second view may show either detector response from a specific stream or the input kinematic information for the “real” case of embedding model signal in real data event
ALICE Offline Week – 22 Oct Summary Embedding visualization is under way Main software requirements identified Specific software tasks for alieve defined Implementation in alieve started Example embedding events provided both for simple testing (pp) and full-scale testing (PbPb)
ALICE Offline Week – 22 Oct The End