Give a Welcome to the COMET

💫 COMET - Cosmological Observables Modelled by Emulated perturbation Theory.

COMET is a Python package that provides emulated predictions of large-scale structure observables from models that are based on perturbation theory. COMET substantially speeds up these analytic computations without any relevant sacrifice in accuracy, enabling an extremely efficient exploration of large-scale structure likelihoods.

At its core, COMET exploits the evolution mapping approach of Sanchez 2020 and Sanchez et al. 2021, which gives it a high degree of flexibility and allows it to cover a wide cosmology parameter space at continuous redshifts up to $z \sim 3$. Specifically, the current release of COMET supports the following parameters (for more details, see here):

Currently, COMET can be used to obtain the following quantities (the perturbation theory models are described here):

• the real-space galaxy power spectrum at one-loop order
• multipoles (monopole, quadrupole, hexadecapole) of the redshift-space power spectrum at one-loop order
• the linear matter power spectrum (with and without infrared resummation)
• Gaussian covariance matrices for the real-space power spectrum and redshift-space multipoles
• $\chi^2$'s for arbitrary combinations of multipoles

COMET provides an easy-to-use interface for all of these computations, and we give quick-start as well as more in-depth examples on our tutorial pages.

Our package is made publicly available under the MIT licence; please cite the papers listed above if you are making use of COMET in your own work.

Getting started

Install the code is as easy as

pip install comet-emu

Then you can follow the Jupyter Notebook for a small example on how to make predictions, compare with data and estimate the $\chi^2$ of your model.

Developer version

If you want to modify the code and play around with it, we provide a developer version so that you can make it and test it. Also, could be possible that you have your own theoretical predictions and you wish to train the emulator with your own computations. You can install the developer version as follow.

git clone git@gitlab.com:aegge/comet-emu.git
cd comet-emu
pip install -e .

Then you can follow the Jupyter Notebook to learn how to train the COMET and make predictions.

License

MIT License

Project status

.. note:: The COMET emulator is under constant development and new versions of the emulator become available as we improve them. Follow our public repository <https://gitlab.com/aegge/comet-emu>_ to make sure you are always up to date with our latest release.