/dsphs_gnn

Sampling the dark matter density profiles posterior with graph neural network and normalizing flows

Primary LanguagePythonMIT LicenseMIT

Neural Simulation-based Inference to uncover the dark matter density profiles of dSph

Package has migrated to https://github.com/trivnguyen/JeansGNN

Example code:

    python train_gnn.py --input data/training.h5 data/validation.h5 --output outdir --in-channels 2 --out-channels 3