- Introduction: What is Probabilistic Programming and why Pyro?
- A Simple but Complete Example
- Modelling in Pyro
- Minipyro (Model Sampling) and Event Handlers
- trace
- replay
- block
- condition
- do
- Tensor Shapes and Plate - Multi-Dimensional Inputs and Outputs
- Sampling and Parameterization
- The Funsor Backend
- Getting Probabilities From Your Model
- The Guide or Approximate Distribution
- Manual Guide
- Automatic Guide
- Minipyro (Inference)
- Posterior Inference
- Importance Sampling
- The ELBO
- SVI
- The Posterior Predictive Distribution
- Neural Networks in Pyro- PyroModule
- The Structural Cause Model and Writing a DAG
- Training the Structural Causal Model
- Intervention and The Interventional Distribution
- Calculating Causal Effects
- Counterfactual Inference
- Abduction
- Intervention
- Inference
- The Autoencoder
- The Variational Autoencoder
- The Causal Effect Variational Autoencoder