ECoPANN

ECoPANN (Estimating Cosmological Parameters with Artificial Neural Network) is a new method to estimate cosmological parameters accurately based on Artificial Neural Network (ANN) proposed by Guo-Jian Wang, Si-Yao Li, Jun-Qing Xia (2020).

It is an alternative to the traditional Markov chain Monte Carlo (MCMC) method, and can obtain almost the same results as MCMC.

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

How to use ECoPANN

First, you are probably going to needs to see the Introduction guide to learn the basic principles of ECoPANN. After that, you may need to install ecopann 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:

• Introduction
  • A brief introduction to ANN
  • Network architecture
  • Training process
  • Parameter inference
• Installation
  • Dependencies
  • Package managers
  • Test the installation
• Quick Start
  • Using one dataset
  • Using multiple datasets
• Package Reference
  • ANN models
  • Training sets
  • Parameter estimation
  • Utils
• Release History
  • 1.0.0 - December 9, 2020

Attribution

If you use this code in your research, please cite our paper (ApJS, arXiv, ADS, BibTex).

License

Copyright 2020-2020 Guojian Wang

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