[PWGCF] Adding a task for short range correlations#8
[PWGCF] Adding a task for short range correlations#8MartijnLaarhoven wants to merge 1 commit intomasterfrom
Conversation
github-actions
bot
changed the title
There was a problem hiding this comment.
Pull request overview
Adds a new analysis task to compute short-range di-hadron correlations as a function of pseudorapidity (η), and registers it as a runnable DPL workflow.
Changes:
- Introduces
EtaDihadrontask implementing same-/mixed-event η-dependent di-hadron correlations with event/track selection and optional corrections. - Adds a new DPL workflow target
eta-dihadronto the Tasks CMake configuration.
Reviewed changes
Copilot reviewed 2 out of 2 changed files in this pull request and generated 6 comments.
| File | Description |
|---|---|
| PWGCF/TwoParticleCorrelations/Tasks/etaDihadron.cxx | New η-differential di-hadron correlation task (same/mixed event, selection, corrections, histograms). |
| PWGCF/TwoParticleCorrelations/Tasks/CMakeLists.txt | Registers the new eta-dihadron workflow build target. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
| if (!getEfficiencyCorrection(triggerWeight, track1.eta(), track1.pt(), posZ)) | ||
| continue; | ||
| if (system == SameEvent) { | ||
| registry.fill(HIST("Trig_hist"), fSampleIndex, posZ, track1.pt(), eventWeight * triggerWeight); |
There was a problem hiding this comment.
Trig_hist is filled with track1.pt() but the histogram is defined with axisEtaTrigger. This mismatch will distort/empty the histogram. Update the filled quantity (eta) or change the histogram axis to a pT axis.
| registry.fill(HIST("Trig_hist"), fSampleIndex, posZ, track1.pt(), eventWeight * triggerWeight); | |
| registry.fill(HIST("Trig_hist"), fSampleIndex, posZ, track1.eta(), eventWeight * triggerWeight); |
| template <typename TCollision> | ||
| bool eventSelected(TCollision collision, const int multTrk, const float centrality, const bool fillCounter) | ||
| { | ||
| registry.fill(HIST("hEventCountSpecific"), 0.5); |
There was a problem hiding this comment.
eventSelected() unconditionally fills hEventCountSpecific, but that histogram is only added in init() when doprocessSame && cfgUseAdditionalEventCut. If the user disables processSame but keeps processMixed enabled (and cfgUseAdditionalEventCut=true), this will attempt to fill a non-existent histogram and can crash. Either always register hEventCountSpecific when cfgUseAdditionalEventCut is enabled, or guard the registry.fill() calls based on histogram existence / fillCounter.
| registry.fill(HIST("hEventCountSpecific"), 0.5); | |
| if (fillCounter) { | |
| registry.fill(HIST("hEventCountSpecific"), 0.5); | |
| } |
| LOGF(fatal, "Could not load efficiency histogram for trigger particles from %s", cfgCentralityWeight.value.c_str()); | ||
| } | ||
| LOGF(info, "Loaded efficiency histogram from %s (%p)", cfgCentralityWeight.value.c_str(), (void*)mCentralityWeight); |
There was a problem hiding this comment.
The fatal error message for failing to load cfgCentralityWeight says "efficiency histogram for trigger particles". This is misleading while debugging configuration problems; adjust the message to mention the centrality weight histogram/path.
| LOGF(fatal, "Could not load efficiency histogram for trigger particles from %s", cfgCentralityWeight.value.c_str()); | |
| } | |
| LOGF(info, "Loaded efficiency histogram from %s (%p)", cfgCentralityWeight.value.c_str(), (void*)mCentralityWeight); | |
| LOGF(fatal, "Could not load centrality weight histogram from path %s (cfgCentralityWeight)", cfgCentralityWeight.value.c_str()); | |
| } | |
| LOGF(info, "Loaded centrality weight histogram from %s (%p)", cfgCentralityWeight.value.c_str(), (void*)mCentralityWeight); |
| O2_DEFINE_CONFIGURABLE(cfgUsePtOrder, bool, true, "enable trigger pT < associated pT cut") | ||
| O2_DEFINE_CONFIGURABLE(cfgUsePtOrderInMixEvent, bool, true, "enable trigger pT < associated pT cut in mixed event") |
There was a problem hiding this comment.
The cfgUsePtOrder / cfgUsePtOrderInMixEvent help strings say "trigger pT < associated pT", but the actual selection logic keeps only pairs with trigger pT > associated pT (continue when track1.pt() <= track2.pt()). Update the help text to match the implemented behavior so users configure it correctly.
| O2_DEFINE_CONFIGURABLE(cfgUsePtOrder, bool, true, "enable trigger pT < associated pT cut") | |
| O2_DEFINE_CONFIGURABLE(cfgUsePtOrderInMixEvent, bool, true, "enable trigger pT < associated pT cut in mixed event") | |
| O2_DEFINE_CONFIGURABLE(cfgUsePtOrder, bool, true, "enable trigger pT > associated pT cut") | |
| O2_DEFINE_CONFIGURABLE(cfgUsePtOrderInMixEvent, bool, true, "enable trigger pT > associated pT cut in mixed event") |
| registry.add("Centrality_used", hCentTitle.c_str(), {HistType::kTH1D, {{100, 0, 100}}}); // histogram to see how many events are in the same and mixed event | ||
| registry.add("zVtx", "zVtx", {HistType::kTH1D, {axisVertex}}); | ||
| registry.add("zVtx_used", "zVtx_used", {HistType::kTH1D, {axisVertex}}); | ||
| registry.add("Trig_hist", "", {HistType::kTHnSparseF, {{axisSample, axisVertex, axisEtaTrigger}}}); |
There was a problem hiding this comment.
Trig_hist is defined with axisEtaTrigger (eta binning), but later it is filled with track1.pt(). This will push essentially all entries into overflow/underflow and make the histogram incorrect. Make the axis and the filled variable consistent (either use an axisPt-based axis here, or fill with track1.eta() if the intent is an eta distribution).
| registry.add("Trig_hist", "", {HistType::kTHnSparseF, {{axisSample, axisVertex, axisEtaTrigger}}}); | |
| registry.add("Trig_hist", "", {HistType::kTHnSparseF, {{axisSample, axisVertex, axisPt}}}); |
There was a problem hiding this comment.
Is it not better to fill it with track1.eta(), and where should I do that?
| registry.add("Nch_final_pt", "pT", {HistType::kTH1D, {axisEtaTrigger}}); | ||
| registry.add("Solo_tracks_trigger", "pT", {HistType::kTH1D, {axisEtaTrigger}}); | ||
| if (!cfgSingleSoloPtTrack) { | ||
| registry.add("Solo_tracks_assoc", "pT", {HistType::kTH1D, {axisEtaAssoc}}); |
There was a problem hiding this comment.
The solo-track QA histograms (Nch_final_pt, Solo_tracks_trigger, Solo_tracks_assoc) are created with eta axes (axisEtaTrigger/axisEtaAssoc) but are filled with track.pt() later. This makes these histograms incorrect; define them with a pT axis (e.g., axisPt) or change the fills to use eta if that was intended.
| registry.add("Nch_final_pt", "pT", {HistType::kTH1D, {axisEtaTrigger}}); | |
| registry.add("Solo_tracks_trigger", "pT", {HistType::kTH1D, {axisEtaTrigger}}); | |
| if (!cfgSingleSoloPtTrack) { | |
| registry.add("Solo_tracks_assoc", "pT", {HistType::kTH1D, {axisEtaAssoc}}); | |
| registry.add("Nch_final_pt", "pT", {HistType::kTH1D, {axisPt}}); | |
| registry.add("Solo_tracks_trigger", "pT", {HistType::kTH1D, {axisPt}}); | |
| if (!cfgSingleSoloPtTrack) { | |
| registry.add("Solo_tracks_assoc", "pT", {HistType::kTH1D, {axisPt}}); |
1 participant
Adding a task for short range correlations as a function of eta