νGAN: A Deep Learning Emulator for Cosmic Web Simulations with Massive Neutrinos

This repository contains data and pretrained model checkpoint used in the nuGAN (neutrino GAN) paper.

Data:
- scaled_density_contrast_maps.npy: 2D density maps of shape (15000,256,256). Maps correpsonding to neutrino masses 0.0, 0.1, 0.4, 0.8, and 12.2 can be chosen by indices [:3000], [3000:6000], [6000:9000], [9000,12000], and [12000:] respectively.

- neutrino_masses.npy: corresponding neutrino masses of shape (15000,). Same indexing for as above.

- test_*: Test data

Model:
- Trained nuGAN checkpoint state dict

These resources are provided without restrictions.