We perform a Bayesian analysis to study possible features in the primordial inflationary power spectrum of scalar perturbations. In particular, we analyze the possibility of detecting the imprint of these primordial features in the anisotropy temperature power spectrum of the cosmic microwave background (CMB) and also in the matter power spectrum $P(k)$. We use the most recent CMB data provided by the Planck Collaboration and $P(k)$ measurements from the 11th data release of the Sloan Digital Sky Survey. We focus our analysis on a class of potentials whose features are localized at different intervals of angular scales, corresponding to multipoles in the ranges $10<\ell <60$ (Oscill-1) and $150<\ell <300$ (Oscill-2). Our results show that one of the step potentials (Oscill-1) provides a better fit to the CMB data than does the featureless $\mathrm{\Lambda }\mathrm{CDM}$ scenario, with moderate Bayesian evidence in favor of the former. Adding the $P(k)$ data to the analysis weakens the evidence of the Oscill-1 potential relative to the standard model and strengthens the evidence of this latter scenario with respect to the Oscill-2 model.

• Received 27 April 2016

DOI:https://doi.org/10.1103/PhysRevD.94.023526