BayesFlow Testing Patterns

This skill provides specialized guidance for testing BayesFlow-based simulation-based inference workflows. It enforces critical environment configurations, such as the mandatory Keras backend setup in conftest.py, and provides standardized patterns for shape testing, configuration serialization, and mock-based network testing. By implementing advanced patterns like invariance and equivariance testing, it ensures that neural networks within the BayesFlow ecosystem maintain their theoretical symmetry properties while keeping CI runs fast and deterministic.

Key Features

Use Cases