Flattenicity as "centrality" estimator in p-Pb collisions simulated with PYTHIA 8.312 Angantyr

In this paper, a "centrality" estimator based on flattenicity ($\rho$) is studied in proton-led (p-Pb) collisions at $\sqrt{s_{\rm NN}}=5.02$ TeV using PYTHIA 8 Angantyr. Although Angantyr is still under development, the existing implementation is enough to study the particle production in systems where medium effects are absent. Firstly, ALICE data on pseudorapidity distributions as a function of the forward multiplicity (V0M), as well as transverse momentum distributions of identified particles in non-single diffractive p-Pb collisions, are compared with Angantyr. Secondly, the average number of binary nucleon-nucleon ($N_{\rm coll}$) collisions for different "centrality" estimators are compared. The studies include the following "centrality" estimators: V0M, $\rho$ and midrapidity multiplicity (CL1). On one hand, the "centrality" dependence of $\langle N_{\rm coll} \rangle$ for the $\rho$ selection shows the smallest deviations ($<8$ %) with respect to that obtained using impact parameter $b$; on the other hand, the V0M and CL1 yield huge deviations (up to a factor 2) with respect to the results using $b$. The particle ratios and nuclear modification factors ($Q_{\rm pPb}$) as a function of $p_{\rm T}$ are also studied. The proton-to-pion ratio exhibits a flow-like peak at intermediate $p_{\rm T}$ (2-8 GeV/$c$) with little or no "centrality" dependence for V0M, $\rho$ and $b$ selections. The kaon-to-pion ratio as a function of $p_{\rm T}$ is "centrality" independent for the same selections. On the contrary, for the CL1 class the ratios exhibit the typical behaviour associated with hard physics. Regarding $Q_{\rm pPb}$, a peak at intermediate $p_{\rm T}$ ($2-8$ GeV/$c$) for different particle species is observed when the "centrality" is obtained with $b$ or $\rho$. The observed features diminish for the selections based on V0M and CL1.