This repository contains a tutorial notebook demonstrating the implementation of Neural Posterior Estimation (NPE), a powerful method in simulation-based inference.
NPE can be employed in two primary configurations:
-
Sequential Neural Posterior Estimation (SNPE)
-
Amortized Neural Posterior Estimation (ANPE)
Explore the world of simulation-based inference, also known as likelihood-free inference (LFI) or SBI, through this Jupyter notebook. Gain insights into the Neural Posterior Estimation method and its applications.
To use this notebook, please ensure that you have the following libraries installed:
- PyTorch
- NumPy
- SBI
- kilonovanet [Simulator]
To install the SBI package, visit https://www.mackelab.org/sbi/
Refer to the last cell of the notebook for essential functions and libraries.
This project was developed by Phelipe Darc and contributors from CBPF (2024).
- E-mail: phelipedarc@gmail.com
-
Article: Simulation Based Inference of BNS Kilonova Properties: A Case Study with AT2017gfo
-
Webpage: SBI tutorial by Cranmer
Feel free to explore and contribute to advance simulation-based inference methods! 🚀