LtU-ILI: An All-in-One Framework for Implicit Inference in Astrophysics and Cosmology

Abstract

This paper presents the Learning the Universe Implicit Likelihood Inference (LtU-ILI) pipeline, a codebase for rapid, user-friendly, and cutting-edge machine learning (ML) inference in astrophysics and cosmology. The pipeline includes software for implementing various neural architectures, training schemata, priors, and density estimators in a manner easily adaptable to any research workflow. It includes comprehensive validation metrics to assess posterior estimate coverage, enhancing the reliability of inferred results. Additionally, the pipeline is easily parallelizable and is designed for efficient exploration of modeling hyperparameters. To demonstrate its capabilities, we present real applications across a range of astrophysics and cosmology problems, such as: estimating galaxy cluster masses from X-ray photometry; inferring cosmology from matter power spectra and halo point clouds; characterizing progenitors in gravitational wave signals; capturing physical dust parameters from galaxy colors and luminosities; and establishing properties of semi-analytic models of galaxy formation. We also include exhaustive benchmarking and comparisons of all implemented methods as well as discussions about the challenges and pitfalls of ML inference in astronomical sciences. All code and examples are made publicly available at https://github.com/maho3/ltu-ili.

22 pages, 10 figures, accepted in the Open Journal of Astrophysics. Code available at https://github.com/maho3/ltu-ili

References in corpus (55)

• Stellar population synthesis at the resolution of 2003
• MCMC using Hamiltonian dynamics
• Simulating Galaxy Formation with the IllustrisTNG Model
• The 3D power spectrum of galaxies from the SDSS
• Gravitational Waves from Mergin Compact Binaries: How Accurately Can One Extract the Binary's Parameters from the Inspiral Waveform?
• Score-Based Generative Modeling through Stochastic Differential Equations
• Simba: Cosmological Simulations with Black Hole Growth and Feedback
• Frequency-domain gravitational waves from non-precessing black-hole binaries. II. A phenomenological model for the advanced detector era
• Properties of the Binary Black Hole Merger GW150914
• Robust parameter estimation for compact binaries with ground-based gravitational-wave observations using the LALInference software library
• Bayesian inference for compact binary coalescences with BILBY: Validation and application to the first LIGO--Virgo gravitational-wave transient catalogue
• The Quijote simulations
• Analysis of spin precession in binary black hole systems including quadrupole-monopole interaction
• Real-time gravitational-wave science with neural posterior estimation
• Colours and luminosities of z=0.1 simulated galaxies in the EAGLE simulations
• Validating Bayesian Inference Algorithms with Simulation-Based Calibration
• Fast likelihood-free cosmology with neural density estimators and active learning
• Gravitational-wave parameter estimation with autoregressive neural network flows
• Massive optimal data compression and density estimation for scalable, likelihood-free inference in cosmology
• Approximating Likelihood Ratios with Calibrated Discriminative Classifiers
• Likelihood-free inference with neural compression of DES SV weak lensing map statistics
• Parameter estimation for heavy binary-black holes with networks of second-generation gravitational-wave detectors
• Bayesian optimisation for likelihood-free cosmological inference
• Fast and Credible Likelihood-Free Cosmology with Truncated Marginal Neural Ratio Estimation
• A Full $w$CDM Analysis of KiDS-1000 Weak Lensing Maps using Deep Learning
• Neural posterior estimation for exoplanetary atmospheric retrieval
• Neural Posterior Estimation with guaranteed exact coverage: the ringdown of GW150914
• Robust Field-level Likelihood-free Inference with Galaxies
• The Cosmic Graph: Optimal Information Extraction from Large-Scale Structure using Catalogues
• Regulation of Star Formation by a Hot Circumgalactic Medium
• SBI++: Flexible, Ultra-fast Likelihood-free Inference Customized for Astronomical Applications
• SICRET: Supernova Ia Cosmology with truncated marginal neural Ratio EsTimation
• Robust Simulation-Based Inference in Cosmology with Bayesian Neural Networks
• Lightning-Fast Gravitational Wave Parameter Inference through Neural Amortization
• GWSim: Python package for creating mock GW samples for different astrophysical populations and cosmological models of binary black holes
• Investigating the Impact of Model Misspecification in Neural Simulation-based Inference
• Solving high-dimensional parameter inference: marginal posterior densities & Moment Networks
• Flow Matching for Scalable Simulation-Based Inference
• Learning Robust Statistics for Simulation-based Inference under Model Misspecification
• Sequential Neural Score Estimation: Likelihood-Free Inference with Conditional Score Based Diffusion Models
• Practical Guidance for Bayesian Inference in Astronomy
• Truncated Marginal Neural Ratio Estimation
• $\mathrm{S}\mathrm{IM}\mathrm{BIG}$: The First Cosmological Constraints from Non-Gaussian and Non-Linear Galaxy Clustering
• GATSBI: Generative Adversarial Training for Simulation-Based Inference
• Fishnets: Information-Optimal, Scalable Aggregation for Sets and Graphs
• Higher-order statistics of the large-scale structure from photometric redshifts
• Sensitivity Analysis of Simulation-Based Inference for Galaxy Clustering
• Misspecification-robust Sequential Neural Likelihood for Simulation-based Inference
• Simulation-based Inference for Exoplanet Atmospheric Retrieval: Insights from winning the Ariel Data Challenge 2023 using Normalizing Flows
• Constructing Impactful Machine Learning Research for Astronomy: Best Practices for Researchers and Reviewers
• Simulator-Based Inference with Waldo: Confidence Regions by Leveraging Prediction Algorithms and Posterior Estimators for Inverse Problems
• Inferring the Ionizing Photon Contributions of High-Redshift Galaxies to Reionization with JWST NIRCam Photometry
• PQMass: Probabilistic Assessment of the Quality of Generative Models using Probability Mass Estimation
• Albatross: A scalable simulation-based inference pipeline for analysing stellar streams in the Milky Way
• A point cloud approach to generative modeling for galaxy surveys at the field level

Cited by in corpus (20)

• A Parameter-Masked Mock Data Challenge for Beyond-Two-Point Galaxy Clustering Statistics
• Learning the Universe: Cosmological and Astrophysical Parameter Inference with Galaxy Luminosity Functions and Colours
• COmoving Computer Acceleration (COCA): $N$-body simulations in an emulated frame of reference
• Cosmology from Point Clouds with Dark Matter Halos from the Quijote Simulations
• Hybrid summary statistics: neural weak lensing inference beyond the power spectrum
• Informed total-error-minimizing priors: Interpretable cosmological parameter constraints despite complex nuisance effects
• Cosmological Constraints with Void Lensing I: the Simulation-Based Inference Framework
• Field-Level Comparison and Robustness Analysis of Cosmological N-body Simulations
• Dark from light (DfL): Inferring halo properties from luminous tracers with machine learning trained on cosmological simulations. I. Method, proof of concept & preliminary testing
• Cosmology with Persistent Homology: Parameter Inference via Machine Learning
• Cosmology with Topological Deep Learning
• Reconstructing galaxy star formation histories from COSMOS2020 photometry using simulation-based inference
• Quantifying the Milky Way, LMC and their interaction using all-sky kinematics of outer halo stars
• Exploring the BSM parameter space with Neural Network aided Simulation-Based Inference
• Towards an optimal extraction of cosmological parameters from galaxy cluster surveys using convolutional neural networks
• Statistical Patterns in the Equations of Physics and the Emergence of a Meta-Law of Nature
• Diagnosing Systematic Effects Using the Inferred Initial Power Spectrum
• Beyond single tracers: CNN-based inference of galaxy mass profiles from combined gas and stellar kinematics
• Testing subhalo abundance matching with galaxy kinematics
• Semiresolved Stellar Populations as Distance Indicators