This repository includes the trained normalizing flow describing the galaxy star-forming sequence from Leja et al. 2021. The normalizing flow is trained using code from Green & Ting 2020.
The trained flow is stored in data. The code to interact with this flow is in sample_nf_probability_density.py, including stellar mass-complete limits.
You can interact with the trained data using the do_all function in this file. A second file, torch_light.py, contains all of the helper functions
needed to interpret the flow. Normal usage will not require interacting with this code.
The package dependencies include torch, astropy, scipy, numpy, and matplotlib.