Power Spectrum, Bispectrum, 2- and 3-Point Correlation Function, and Beyond

TL;DR

This paper reviews the theoretical background, practical estimation methods, and recent applications of N-point correlation functions, including power spectrum and bispectrum, in cosmology to analyze large-scale structure clustering.

Contribution

It provides a comprehensive overview of the motivation, algorithms, and data applications of N-point correlation functions beyond the commonly used 2- and 3-point functions.

Findings

Discussion of how clustering emerges from these functions

Explanation of algorithms for estimating correlation functions

Linking to public code for practical application

Abstract

N-Point Correlation Functions, usually with N = 2, 3, and their Fourier-space analogs power spectrum and bispectrum, are major tools used in cosmology to capture the clustering of large-scale structure. We outline how the clustering these functions capture emerges, explain that inflation produces a 2PCF or power spectrum but that subsequent evolution eventually produces a 3PCF or bispectrum, and beyond (and that inflation may do so as well at some level). Furthermore, in principle the Universe also has a 4PCF or trispectrum, and even clustering beyond. For each of these tools, we discuss the motivation, the practical details of how they are estimated, the current algorithms used to compute them, the theory behind them, and recent applications to data. Throughout, we focus on positioning the reader to find and apply these algorithms with some understanding, linking to public code for each algorithm to the fullest extent possible.