Skip to content

Latest commit

 

History

History
41 lines (30 loc) · 4.31 KB

README.md

File metadata and controls

41 lines (30 loc) · 4.31 KB

cosmocnc

cosmocnc is a Python package for evaluating the number count likelihood of galaxy cluster catalogues in a fast, flexible and accurate way. It is based on the use of Fast Fourier Transform (FFT) convolutions in order to evaluate some of the likelihood integrals. The code was introduced in Zubeldia & Bolliet (2024), where the likelihood formalism and implementation are described in detail. If you use the code, please cite the paper.

If you have any questions or comments, please write Íñigo Zubeldia at inigo.zubeldia at ast cam ac uk.

Main features of the code

The main features of the code are the following:

  • It supports three types of likelihoods: an unbinned likelihood, a binned likelihood, and an extreme value likelihood.
  • It also supports the addition of stacked cluster data (e.g., stacked lensing profiles), which is modelled in a consistent way with the cluster catalogue.
  • It links the cluster mass observables (also known as mass proxies, e.g., tSZ signal-to-noise, richness, lensing mass estimate, or X-ray flux) to the cluster mass and redshift through a hierarchical model with an arbitrary number of layers, allowing for correlated scatter between the different mass observables. In each layer, the mass--observable scaling relations and the scatter covariance matrix can be easily defined in a custom way, and can depend on sky location and redshift.
  • It incorporates several widely-used halo mass functions.
  • The unbinned likelihood is the most general and flexible of the three. It supports an arbitrary number of cluster mass observables for each cluster in the sample and it allows for the set of mass observables to vary from cluster to cluster. It also allows for redshift measurement uncertainties. If some mass observables have completely uncorrelated scatter, cosmocnc takes advantage of this fact, boosting its computational performance significantly.
  • The binned and extreme value likelihoods, on the other hand, consider the cluster abundance only across one mass observable and/or redshift, and do not allow for redshift measurement uncertainties.
  • cosmocnc can produce mass estimates for each cluster in the sample, which are derived assuming the hierarchical model that is used to model the mass observables.
  • It also allows for the generation of synthetic cluster catalogues for a given observational set-up.
  • Several of cosmocnc's computations can be accelerated with Python's multiprocessing module.
  • The code is interfaced with the Markov chain Monte Carlo (MCMC) code Cobaya, allowing for easy-to-run MCMC parameter estimation.
  • The code is also interfaced with class_sz, allowing a wide range of cosmological models as well as enabling joint analyses with Cosmic Microwave Background (CMB) and Large Scale Structure (LSS) survey data.

Installation

Download the source code and do

$ pip install -e .

You'll then be able to import cosmocnc in Python with

import cosmocnc

Dependencies: astropy (optional), class_sz (optional), cosmopower (optional), hmf (optional), mcfit (optional).

Tutorials

The main computational capabilities of cosmocnc are illustrated in this Jupyter notebook. In this other one, the code is benchmarked against class_sz.

15/10/2024: Code upgrade

The code was significantly upgraded on 15/10/2024. In particular, the way the survey data (cluster catalogue and observable scaling relations) enter the code was modified, with the goal of making it easier and cleaner to add a new survey. Each survey now comes with two files, a catalogue file and a scaling relations file, where the survey details are specified and the data is formatted so that cosmocnc can process it. An example of these two files for the SO-like catalogues analysed in Zubeldia & Bolliet (2024) can be found in cosmocnc/surveys.