PyAutoFit - A Classy Probabilistic Programming Language For Data Science

Explore PyAutoFit, a powerful open-source probabilistic programming language for automated Bayesian inference, in this 29-minute EuroPython 2021 conference talk. Discover how PyAutoFit simplifies the composition and fitting of probabilistic models using various Bayesian inference libraries, handles model-fitting complexities, and excels in big-data analysis. Learn about its core features, advanced capabilities like multi-level models and automated pipelines, and its application in astronomical research for understanding dark matter. Gain insights into customizing models, analyses, and searches, as well as leveraging database functionalities for large-scale data processing. Suitable for those with basic object-oriented Python programming knowledge, this talk provides a comprehensive introduction to PyAutoFit's potential for diverse data science applications.

Intro
Overview
Model Fitting
What is Probabilistic Programming?
PyAutoFit: Classy Probabilistic Programming
Python Classes
Customizing the Model
Customizing the Analysis
Customizing the Search
Database
Advanced Modeling Tools
Strong Gravitational Lensing Machine Learning
Model Composition
Galaxy Class
Writing the Analysis
Summary