GitHub

Contributing

Contributions to RxGP.jl are welcome! This page covers how to get started.

Setting up a development environment

  1. Clone the repository:

    git clone https://github.com/ReactiveBayes/RxGP.jl.git
cd RxGP.jl

  2. Activate and instantiate the project:

    using Pkg
Pkg.activate(".")
Pkg.instantiate()

  3. Run the test suite:

    Pkg.test()

Project structure

DirectoryContents
src/Main source code
src/SparseGaussianProcessnode/Node definitions and average energy computations
src/rule/Message passing update rules, organised by node and edge
src/helper_functions/Kernels, caching, approximation, prediction, meta, utilities
test/Unit tests mirroring the src/ structure
examples/Jupyter notebook examples
docs/Documentation (Documenter.jl)

Adding a new message passing rule

  1. Create or edit the appropriate file in src/rule/<node_type>_rules/<edge>.jl.
  2. Use the @rule macro from ReactiveMP.jl to define the rule.
  3. Add corresponding tests in test/node_rule/.
  4. Make sure Pkg.test() passes.

Adding a new node

  1. Define the node struct and register it with @node in src/SparseGaussianProcessnode/.
  2. Define an @average_energy method.
  3. Implement rules for each edge.
  4. Create a corresponding meta type if needed.
  5. Export the node and meta from the main module.
  6. Add tests and update documentation.

Building the documentation locally

using Pkg
Pkg.activate("docs")
Pkg.instantiate()
include("docs/make.jl")

The built documentation will be in docs/build/.