Alternative Clustering Methods#
Warning
This documentation is still under construction. Please check back later for more information.
What is the ACM Pipeline ?#
The Alternative Clustering Methods (ACM) pipeline is a Python package that provides a set of tools for beyond-2pt galaxy clustering statistics for DESI. It allows the training of several simulation-based theoretical models to emulate the galaxy clustering statistics (see Pipeline). It contains several clustering methods that can be used to estimate the galaxy clustering statistics, such as densitysplit, bispectrum, Minkowski functionals and more (see Statistics).
Why are we doing this ?#
The non-Gaussianity of the density field at small scales limits the amount of information we can access with the galaxy power spectrum. Alternative clustering statistics can access this information at a limited computational cost, but most of them lack analytic theory models.
How are we doing this ?#
We train neural networks on clustering statistics measured on thousands of Abacus-based mocks, learning about how the clustering responds to cosmology and the galaxy-halo connection.
Getting Started
The pipeline
Credits
Note
🔎 See an issue with the documentation? Feel free to open an issue or a pull request on the GitHub repository to help us improve the documentation !