CovNet

This package contains the code required to design, train, and test neural networks to generate covariance matrices at different cosmologies.

This package is compatable with python 3.5 - 3.8 (other versions of python don't work with nbodykit, so use at your own risk!)

Installing the Code

This package should work for both Linux and MacOS. We recommend using anaconda for package management, but standalone pip should work as well.

1. Install CLASS-PT (https://github.com/Michalychforever/CLASS-PT) and follow the instillation instructions for the classy python extension there.
2. To enable GPU functionality for network training, make sure you have CUDA installed (or python 3.8+ if using apple arm64) and install the corresponding PyTorch version. If your machine doesn't have a GPU, you can skip this step.
3. Download this repository to your location of choice.
4. In the base directory, run python -m pip install ., which should install this repository as a package you can call anywhere on your machine.

NOTE: In order to calculate your own covariance window functions, you'll also need to install nbodykit. This package is not configured to run on the new apple processors, thus it isn't in the package dependency list. Some of the provided scripts also utilize mpi4py, so if you want to use those please make sure you have a working MPI implimentation (ex. openmpi) before installing that package.

Using the Code

A description of how to use CovNet is given here.

Citing the Code

Users of this code should cite the authors (http://arxiv.org/abs/2405.00125), CovaPT (https://arxiv.org/abs/1910.02914), and CLASS-PT (https://arxiv.org/abs/2004.10607)