CWTHF: Identifying Dark Matter Halos with Continuous Wavelet Transform

Abstract

Cosmological simulations are an important method for investigating the evolution of the Universe. In order to gain further insight into the processes of structure formation, it is necessary to identify isolated bound objects within the simulations, namely, the dark matter halos. The continuous wavelet transform (CWT) is an effective tool used as a halo finder due to its ability to extract clustering information from the input data. In this study, we introduce CWTHF (Continuous Wavelet Transform Halo Finder), the first wavelet-based, MPI-parallelized halo finder, marking a novel approach in the field of cosmology. We calculate the CWT from the cloud-in-cell (CIC) grid and segment the grid based on the local CWT maxima. We then investigate the effects of the parameters that influence our program and identify the default settings. A comparison with the conventional friends-of-friends (FOF) method demonstrates the viability of CWT for halo finding. Although the actual performance is not faster than FOF, the linear time complexity of $\mathcal{O}(N)$ of our identification scheme indicates its significant potential for future optimization and application.