Examples

RxGP ships with Jupyter notebook examples in the examples/ directory of the repository. These demonstrate end-to-end workflows for common GP modelling scenarios.

GP Regression

Notebook: examples/GPRegression.ipynb

Standard sparse GP regression with scalar observations. Demonstrates:

• Setting up inducing points and a squared-exponential kernel
• Defining the model with UniSGP nodes
• Running variational inference with RxInfer
• Posterior prediction with predict_GP

GP State-Space Model

Notebook: examples/GPSSM.ipynb

A Gaussian process state-space model where the GP governs the transition dynamics. Demonstrates:

• Composing GP factor nodes with standard state-space model components
• Streaming (online) inference over time steps
• Uncertain inputs flowing through the GP node

GP Regression with Gradient Observations (AMF)

Notebook: examples/GP_Regression_AMF_Grads.ipynb

GP regression that incorporates derivative observations alongside function-value observations. Demonstrates:

• Using the UniSGP_dID node with :grad and :joint_fn_grad operators
• Automatic Mean Field (AMF) variational inference
• Hyperparameter optimisation via gradient-based methods

using Pkg
Pkg.activate("examples")
Pkg.develop(PackageSpec(path="."))   # link local RxGP
Pkg.instantiate()