[PWGCF] Fixes in femto framework (AliceO2Group#16097)

Author: ariedel-cern

PWGCF/Femto/Core/pairHistManager.h

@@ -32,6 +32,7 @@
 #include <Math/Vector4D.h> // IWYU pragma: keep (do not replace with Math/Vector4Dfwd.h)
 #include <Math/Vector4Dfwd.h>
 
+#include <algorithm>
 #include <array>
 #include <cmath>
 #include <cstdint>
@@ -903,15 +904,20 @@ class PairHistManager
 
   float getKstar(ROOT::Math::PtEtaPhiMVector const& part1, ROOT::Math::PtEtaPhiMVector const& part2)
   {
-    // compute pair momentum
-    auto sum = part1 + part2;
-    // Boost particle 1 to the pair rest frame (Prf) and calculate k* (would be equivalent using particle 2)
-    // make a copy of particle 1
-    auto particle1Prf = ROOT::Math::PtEtaPhiMVector(part1);
-    // get lorentz boost into pair rest frame
-    ROOT::Math::Boost boostPrf(sum.BoostToCM());
-    // boost particle 1 into pair rest frame and calculate its momentum, which has the same value as k*
-    return static_cast<float>(boostPrf(particle1Prf).P());
+    // Use Cartesian 4-vectors: addition/M2() become pure arithmetic
+    const ROOT::Math::PxPyPzEVector p1(part1);
+    const ROOT::Math::PxPyPzEVector p2(part2);
+
+    // Mandelstam s = (p1 + p2)^2
+    const double s = (p1 + p2).M2();
+    const double m1sq = p1.M2();
+    const double m2sq = p2.M2();
+
+    // Källen function λ(s, m1^2, m2^2) = (s - m1^2 - m2^2)² - 4*m1^2*m2^2
+    const double kallen = (s - m1sq - m2sq) * (s - m1sq - m2sq) - 4.0 * m1sq * m2sq;
+
+    // k* = 0.5 * sqrt(λ/s)
+    return static_cast<float>(0.5 * std::sqrt(std::max(0.0, kallen) / s));
   }
 
   o2::framework::HistogramRegistry* mHistogramRegistry = nullptr;

PWGCF/Femto/Core/tripletHistManager.h

@@ -383,23 +383,24 @@ class TripletHistManager
   float getQ3() const { return mQ3; }
 
  private:
-  ROOT::Math::PxPyPzEVector getqij(ROOT::Math::PtEtaPhiMVector const& pi, ROOT::Math::PtEtaPhiMVector const& pj)
+  ROOT::Math::PxPyPzEVector getqij(ROOT::Math::PxPyPzEVector const& vi, ROOT::Math::PxPyPzEVector const& vj)
   {
-    // Convert to PxPyPzEVector to get proper Lorentz dot product
-    ROOT::Math::PxPyPzEVector vi(pi);
-    ROOT::Math::PxPyPzEVector vj(pj);
-
     auto trackSum = vi + vj;
     auto trackDifference = vi - vj;
-    double scaling = trackDifference.Dot(trackSum) / trackSum.Dot(trackSum);
+    const double s = trackSum.M2();
+    const double scaling = (s != 0.0) ? (vi.M2() - vj.M2()) / s : 0.0;
     return trackDifference - scaling * trackSum;
   }
 
   float getQ3(ROOT::Math::PtEtaPhiMVector const& part1, ROOT::Math::PtEtaPhiMVector const& part2, ROOT::Math::PtEtaPhiMVector const& part3)
   {
-    auto q12 = getqij(part1, part2);
-    auto q23 = getqij(part2, part3);
-    auto q31 = getqij(part3, part1);
+    // upfront conversion to PxPyPzEVector
+    const ROOT::Math::PxPyPzEVector p1(part1);
+    const ROOT::Math::PxPyPzEVector p2(part2);
+    const ROOT::Math::PxPyPzEVector p3(part3);
+    auto q12 = getqij(p1, p2);
+    auto q23 = getqij(p2, p3);
+    auto q31 = getqij(p3, p1);
     double q = q12.M2() + q23.M2() + q31.M2();
     return static_cast<float>(std::sqrt(-q));
   }

PWGCF/Femto/Tasks/femtoPairEfficiency.cxx

@@ -431,7 +431,7 @@ struct FemtoPairEfficiency {
     ROOT::Math::PtEtaPhiMVector particle1;
     ROOT::Math::PtEtaPhiMVector particle2;
 
-    for (auto const& [p1, p2] : o2::soa::combinations(o2::soa::CombinationsUpperIndexPolicy(tracks, tracks))) {
+    for (auto const& [p1, p2] : o2::soa::combinations(o2::soa::CombinationsStrictlyUpperIndexPolicy(tracks, tracks))) {
 
       bool order1 = checkTrack(p1, TrackSel1) && checkTrackPid(p1, TrackSel1) &&
                     checkTrack(p2, TrackSel2) && checkTrackPid(p2, TrackSel2);