desi-clustering

Collection of scripts to produce the DESI DR2 clustering measurements from the data / mocks catalogs to the parameter inferences.

📦 Installation

You can install the latest version directly from the GitHub repository:

pip install git+https://github.com/cosmodesi/desi-clustering.git

Alternatively, if you plan to contribute or modify the code, install in editable (development) mode:

git clone https://github.com/cosmodesi/desi-clustering.git
cd desi-clustering
pip install -e .

Overview

The package follows a simple structure:

• clustering_statistics: measurement of clustering statistics (power spectrum, correlation function, bispectrum, etc.), common to Full Shape, BAO and PNG Key Projects
• full_shape: full-shape fits
• bao: BAO fits
• local_png: local primordial non-Gaussianity

Environment

source /global/common/software/desi/users/adematti/cosmodesi_environment.sh main  # source the environment
# You may want to have it in the jupyter-kernel for plots
${COSMODESIMODULES}/install_jupyter_kernel.sh main  # this to be done once

You may already have the above kernel (corresponding to the standard GQC environment) installed. In this case, you can delete it:

rm -rf $HOME/.local/share/jupyter/kernels/cosmodesi-main

and rerun:

${COSMODESIMODULES}/install_jupyter_kernel.sh main

Note that you may need to restart (close and reopen) your notebooks for the changes to propagate.