I’ve generate QCD events with MadGraph5_aMC@NLO&Pythia8.2 and reconstructed jets with the anti-kt algorithm; the (leading) jet as well as its constituent particles are stored as TLorentzVector object. Now I’d like to perform a Lorentz boost such that the pseudorapidity eta and the azimuthal angle phi of the reconstructed jet is at the position (0,0) in the eta-phi grid. However, I also want to boost the constituent particles of the jet with the same boost while keeping the angular separation between the constituents conserved. I’ve tried to use TLorentzVector::Boost(Double_t bx, Double_t by, Double_t bz) to solve this problem; however, the reconstructed mass of the jet (by summing the constituent’s TLorentzVectors and call TLorentzVector::M()) gives different results before and after the transformation is applied, while it should be invariant. I’d be very grateful to get a minimal working example. Many thanks in advance!