Redshift determination through weighted phase correlation: a linearithmic implementation

Abstract

We present a new algorithm having a time complexity of O(N log N) and designed to retrieve the phase at which an input signal and a set of not necessarily orthogonal templates match best in a weighted chi-squared sense. The proposed implementation is based on an orthogonalization algorithm and thus also benefits from high numerical stability. We apply this method successfully to the redshift determination of quasars from the twelfth Sloan Digital Sky Survey (SDSS) quasar catalogue and derive the proper spectral reduction and redshift selection methods. Derivations of the redshift uncertainty and the associated confidence are also provided. The results of this application are comparable to the performance of the SDSS pipeline, while not having a quadratic time dependence.