glasz

GitHub Actions Workflow Status GitHub Actions Workflow Status Codecov

PyPI - Version PyPI - Python Version

glasz is a package designed to aid in joint analyses of kSZ+GGL. For more information, please read Sunseri et al. 2025 (in prep). This code is built on pyccl and was derived from the work done in Amodeo et al. 2021.

Authors: James Sunseri

diagram of kSZ+GGL

Installation

This package is installable via pip. All one has to do is run

$ pip install glasz

to install the package into your current environment of choice.

Development Version

This package uses pixi as the default task runner and environment manager. It is significantly faster than any other competitor on the market (conda, miniconda, mambda, etc…). To install pixi, click on the link and install on your machine. It is very light weight and shouldn’t take long. From there, in this repository run the pixi install command and you will gain a pixi environment with this package installed including all of its dependencies. If you are not familiar with pixi, you can use the precomputed pixi.lock file to resolve all dependencies in the creation or modification of a conda environment. You can do this on your machine by running the following command

$ pixi project export conda-explicit-spec conda_env_files --ignore-pypi-errors

which will create a directory called conda_env_files loaded with 3 conda_spec.txt files (similar to a .yaml but these have all the dependencies precomputed and locked). Having the dependency conflicts precomputed is important as pyccl is a monstrous library which takes a very long time to install via conda. Take a look at the files, each one corresponds to a different operating system: linux, mac-OS intel, mac-OS arm (M-chips). Choose the file which resembles your machine and run the command

$ conda create --name ENV_NAME --file conda_env_files/default_{YOUR_SYSTEM}_conda_spec.txt

to create a new conda environment titled ENV_NAME. This environment is a standard conda environment with a development version of glasz installed alongside the dependencies. We provide a conda_env_files directory with precomputed conda_spec.txt files from the current pixi.lock file corresponding to the version on the main branch.