CoLFI¶

CoLFI (Cosmological Likelihood-Free Inference) is a framework to estimate cosmological parameters based on neural density estimators (ANN, MDN, and MNN) proposed by Guo-Jian Wang, Cheng Cheng, Yin-Zhe Ma, et al. (2023).

It is an alternative to the traditional Markov chain Monte Carlo (MCMC) method and has advantages over MCMC.

As a general method of parameter estimation, CoLFI can be used for research in many scientific fields. The code colfi is available for free from GitHub. It can be executed on GPUs or CPUs.

Attribution¶

If you use this code in your research, please cite Guo-Jian Wang, Cheng Cheng, Yin-Zhe Ma, et al., “CoLFI: Cosmological Likelihood-free Inference with Neural Density Estimators”, ApJS, 268, 7, (2023).

If you use the MDN method of this code, please also cite Guo-Jian Wang, Cheng Cheng, Yin-Zhe Ma, Jun-Qing Xia, “Likelihood-free Inference with the Mixture Density Network”, ApJS, 262, 24 (2022).

If you use the ANN method of this code, please also cite Guo-Jian Wang, Si-Yao Li, Jun-Qing Xia, “ECoPANN: A Framework for Estimating Cosmological Parameters Using Artificial Neural Networks”, ApJS, 249, 25 (2020).

How to use CoLFI¶

First, you are probably going to needs to see the Introduction guide to learn the basic principles of CoLFI. After that, you may need to install colfi on your computer according to the Installation guide, and then following the Quick Start guide to learn how to use it. If you need more detailed information about a specific function, the Package Reference below should have what you need.

Contents:¶

License¶

colfi is free software made available under the MIT License. For details see the LICENSE.