Release 6

.github/workflows/Changelog.yml

@@ -0,0 +1,14 @@
+name: Check CHANGELOG.md
+on:
+  pull_request:
+    types: [assigned, opened, synchronize, reopened, labeled, unlabeled]
+    branches:
+      - main
+jobs:
+  check-changelog:
+    name: Check Changelog Action
+    runs-on: ubuntu-latest
+    steps:
+      - uses: tarides/changelog-check-action@v2
+        with:
+          changelog: CHANGELOG.md

CHANGELOG.md

@@ -0,0 +1,228 @@
+# Changelog
+
+All notable changes to ReactiveMP.jl will be documented in this file.
+
+The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
+and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+
+## [Unreleased]
+
+## [6.0.0]
+
+### Added
+- `AbstractStreamPostprocessor` abstraction unifying the old pipeline stages and the per-node `scheduler` argument under a single concept that postprocesses outbound message streams, marginal streams, and score streams uniformly
+- `postprocess_stream_of_outbound_messages`, `postprocess_stream_of_marginals`, `postprocess_stream_of_scores` entry points with `::Nothing` pass-through fallbacks
+- `CompositeStreamPostprocessor` for chaining multiple postprocessors
+- `ScheduleOnStreamPostprocessor` — direct successor of `ScheduleOnPipelineStage` plus the per-node scheduler, applies a Rocket.jl scheduler to all three stream kinds
+- Marginal streams and score streams now go through stream postprocessors (previously only outbound message streams did)
+- Documentation page for stream postprocessors
+- Callback/event system for hooking into message passing steps (rule calls, message products, form constraints, marginal computation)
+- `MessageProductContext` struct to bundle product computation settings and callbacks
+- Labels for variables (`RandomVariable`, `ConstVariable`, `DataVariable`)
+- Docstrings for variable types, form constraints, and related functions
+- Documentation page for callbacks
+- `MethodError` hint for mismatched `handle_event` signatures
+- New annotations system: `AnnotationDict`, `AbstractAnnotations`, `LogScaleAnnotations`, `InputArgumentsAnnotations`
+- `post_rule_annotations!` and `post_product_annotations!` callbacks for annotation processors
+- `@logscale value` macro for setting log-scale annotations inside `@rule` bodies
+- `getannotations` function for `Message` and `Marginal`
+- Migration guide for v5 to v6
+- `skip_initial()`, `skip_clamped()`, `skip_clamped_and_initial()` filter operators replacing the `MarginalSkipStrategy` type hierarchy
+- `new_observation!(datavar, value)` for pushing observed values into a `DataVariable`
+- `get_stream_of_inbound_messages`, `get_stream_of_outbound_messages` accessors on `NodeInterface` and `IndexedNodeInterface`
+- `get_stream_of_marginals`, `set_stream_of_marginals!` accessors on variables
+- `get_stream_of_predictions`, `set_stream_of_predictions!` accessors on variables
+- `set_initial_marginal!`, `set_initial_message!` for seeding variables before inference
+- `create_new_stream_of_inbound_messages!` for allocating per-connection message streams
+- Docstrings for `MessageObservable`, `MarginalObservable`, `FunctionalDependencies`, `collect_functional_dependencies`, `RandomVariableActivationOptions`, `DataVariableActivationOptions`, `FactorNodeActivationOptions`, and `activate!` methods
+- Expanded documentation for variables (stream creation lifecycle per variable type), nodes (interfaces, activation), messages, and marginals
+
+### Changed
+- `FactorNodeActivationOptions` lost its `pipeline` and `scheduler` positional fields and gained a single `postprocessor` field
+- `RandomVariableActivationOptions` renamed its `scheduler` field to `stream_postprocessor`; the default is now `nothing` (no-op) instead of `AsapScheduler()`
+- `getpipeline(options)` and `getscheduler(options)` replaced by `getpostprocessor(options)`
+- `EqualityChain` renamed its `pipeline` field to `postprocessor`
+- Switched from `ReTestItems` to `TestItemRunner` for tests ([#584](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/584))
+- Made formatting checks stricter
+- Removed `variables/generic.jl`; generic variable interface moved into `variable.jl`
+- Replaced hardcoded `DefaultMessageProdFn`/`DefaultMarginalProdFn` with `MessageProductContext`
+- `Message{D, A}` → `Message{D}` (type parameter `A` removed)
+- `Marginal{D, A}` → `Marginal{D}` (type parameter `A` removed)
+- `Message` and `Marginal` now carry an `AnnotationDict` instead of a typed addons tuple
+- Rules no longer return `(result, addons)` tuples — just the result
+- `@call_rule` no longer supports `return_addons` option; use `annotations` keyword with `AnnotationDict`
+- `MessageMapping.addons` field → `MessageMapping.annotations`
+- `MessageProductContext` gained `annotations` field for product-time annotation processors
+- `messagein(interface)` → `get_stream_of_inbound_messages(interface)`
+- `messageout(interface)` → `get_stream_of_outbound_messages(interface)`
+- `getmarginal(variable)` / `getmarginals` → `get_stream_of_marginals(variable)`
+- `getprediction(variable)` / `getpredictions` → `get_stream_of_predictions(variable)`
+- `setmarginal!(variable, value)` → `set_initial_marginal!(variable, value)`
+- `setmessage!(variable, value)` → `set_initial_message!(variable, value)`
+- `update!(datavar, value)` → `new_observation!(datavar, value)`
+
+### Removed
+- `AbstractPipelineStage`, `EmptyPipelineStage`, `CompositePipelineStage`, `ScheduleOnPipelineStage`, `apply_pipeline_stage`, `collect_pipeline`, `+` composition — replaced by the `AbstractStreamPostprocessor` abstraction (see migration guide)
+- `LoggerPipelineStage` — equivalent behaviour can be implemented via callbacks
+- `AsyncPipelineStage` — use `ScheduleOnStreamPostprocessor(AsyncScheduler())` instead
+- `DiscontinuePipelineStage` — was unused; implement a custom `AbstractStreamPostprocessor` if needed
+- `schedule_updates(vars; pipeline_stage = ...)` — construct a `ScheduleOnStreamPostprocessor` and pass it through the activation options instead
+- `getaddons` — use `getannotations` instead
+- `getlogscale(::Message)`, `getlogscale(::Marginal)` — use `getlogscale(getannotations(...))` instead
+- `getmemory`, `getmemoryaddon` — use `get_rule_input_arguments(getannotations(...))` instead
+- `AddonLogScale` — replaced by `LogScaleAnnotations` (calling `AddonLogScale()` throws a descriptive error)
+- `AddonMemory` — replaced by `InputArgumentsAnnotations` (calling `AddonMemory()` throws a descriptive error)
+- `AddonDebug` — use callbacks instead
+- `AbstractAddon`, `multiply_addons`, `@invokeaddon`
+- `message_mapping_addons`, `message_mapping_addon` helper functions
+- `MarginalSkipStrategy` abstract type and `SkipClamped`, `SkipInitial`, `SkipClampedAndInitial`, `IncludeAll` subtypes — use `skip_clamped()`, `skip_initial()`, `skip_clamped_and_initial()` filter operators instead
+- `apply_skip_filter`, `as_marginal_observable` — no longer part of the public API
+- `messagein`, `messageout` — use `get_stream_of_inbound_messages`, `get_stream_of_outbound_messages`
+- `getmarginal`, `getmarginals`, `getprediction`, `getpredictions` — use `get_stream_of_marginals`, `get_stream_of_predictions`
+- `setmarginal!`, `setmarginals!`, `setmessage!`, `setmessages!` — use `set_initial_marginal!`, `set_initial_message!`
+- `update!` — use `new_observation!`
+- `create_messagein!` — use `create_new_stream_of_inbound_messages!`
+
+## [5.6.6] - 2026-03-13
+
+### Fixed
+- Implemented effective rules with specialized dispatch for `MvNormalMeanScalePrecision` ([#579](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/579))
+
+### Tests
+- Added performance test for structured rule specialized for `MvNormalMeanScalePrecision`
+
+## [5.6.5] - 2026-02-02
+
+### Added
+- Implemented `MvNormalWishart` node and `out` rule ([#565](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/565))
+- Issue templates ([#558](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/558))
+- Rule interface name checking for `@rule`, `@marginalrule`, and `@average_energy` macros ([#545](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/545))
+
+### Changed
+- Removed `Requires` dependency (used for Julia <1.9, no longer supported) ([#564](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/564))
+- Removed vibe coded required fields from issue template ([#562](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/562))
+- Bumped compat for Optim to 2 ([#574](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/574))
+- Removed `Zygote` extension requirement for compatibility
+
+### Fixed
+- Fixed documentation build ([#567](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/567))
+
+### Performance
+- Pre-computed double loops in CT model ([#571](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/571))
+
+## [5.6.4] - 2025-11-18
+
+### Fixed
+- Fixed bug in average energy of `Uninformative` type ([#553](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/553))
+- Improved robustness of inverse precision matrix computation in `MvNormalMeanPrecision` rule ([#540](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/540))
+
+### Documentation
+- Added `MessageMapping` documentation ([#550](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/550))
+
+## [5.6.3] - 2025-11-04
+
+### Added
+- Added `IntegrationTest.yml` workflow ([#525](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/525))
+- Implemented missing marginal rule for multiplication node ([#531](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/531))
+
+### Changed
+- Changed dispatch to `AbstractVector` to allow other vector implementations ([#536](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/536))
+
+### Fixed
+- Renamed softdot marginal rules test file to include it in test runs ([#535](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/535))
+- Avoided `Vararg` deprecation warnings ([#537](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/537))
+- Defined `AverageEnergy` for `Mixture` node with warning ([#546](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/546))
+- Allowed different numeric types for `GammaShapeLikelihood` constructor ([#544](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/544))
+
+### Tests
+- Improved code coverage with tests for BIFM, Autoregressive, Mixture, GammaMixture, Wishart, InverseWishart, DotProduct, Multiplication, and Uniform nodes ([#539](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/539))
+
+## [5.6.2] - 2025-10-21
+
+### Fixed
+- Fixed `isonehot` to use approximate comparison for categorical rules ([#527](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/527))
+
+## [5.6.1] - 2025-10-21
+
+### Added
+- Categorical rules: check if probability vector of `q_out` is a one-hot encoded vector ([#510](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/510))
+- Support for non-linear node (univariate -> multivariate) with Unscented transform ([#508](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/508))
+
+### Changed
+- Updated `ForwardDiff` to version 1 ([#521](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/521))
+- Skip Aqua.jl checks during selective test runs ([#523](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/523))
+- Makefile: fixed selective test runs with `test_args` argument ([#517](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/517))
+
+### Fixed
+- Updated documentation for `as_marginal` ([#516](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/516))
+- Updated test cases for `GammaShapeScale`
+
+## [5.6.0] - 2025-09-23
+
+### Changed
+- Use `MvNormalMeanScaleMatrixPrecision` from ExponentialFamily package ([#509](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/509))
+
+## [5.5.12] - 2025-09-11
+
+### Fixed
+- Fixed `q_t1` dimensionality bug in delta node rules ([#504](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/504))
+
+## [5.5.11] - 2025-09-10
+
+### Added
+- Implemented `MvNormalMeanScaleMatrixPrecision` rules ([#497](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/497))
+
+## [5.5.10] - 2025-09-09
+
+### Fixed
+- Added new linearization method ([#500](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/500))
+
+## [5.5.9] - 2025-08-14
+
+### Changed
+- Reverted "Don't check for proper in division of" ([#496](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/496))
+
+## [5.5.8] - 2025-08-14
+
+### Added
+- Show meta suggestions in rule error printing ([#495](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/495))
+
+### Changed
+- Tightened dispatch for summation and other optimized rules ([#492](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/492), [#493](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/493))
+
+### Fixed
+- Don't check for proper in division of to accommodate inference ([#486](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/486))
+
+## [5.5.7] - 2025-07-24
+
+### Fixed
+- Fixed infinite RxInfer documentation build by not using lazy string ([#490](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/490))
+
+## [5.5.6] - 2025-07-23
+
+### Fixed
+- Fixed invalidations: removed bad `eltype` methods, `convert`, and `println` method ([#489](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/489))
+
+## [5.5.5] - 2025-07-23
+
+### Fixed
+- Resolved Gaussian division with proper Multivariate vs Univariate handling ([#479](https://github.com/ReactiveBayes/ReactiveMP.jl/pull/479))
+
+---
+
+[Unreleased]: https://github.com/ReactiveBayes/ReactiveMP.jl/compare/v5.6.6...HEAD
+[5.6.6]: https://github.com/ReactiveBayes/ReactiveMP.jl/compare/v5.6.5...v5.6.6
+[5.6.5]: https://github.com/ReactiveBayes/ReactiveMP.jl/compare/v5.6.4...v5.6.5
+[5.6.4]: https://github.com/ReactiveBayes/ReactiveMP.jl/compare/v5.6.3...v5.6.4
+[5.6.3]: https://github.com/ReactiveBayes/ReactiveMP.jl/compare/v5.6.2...v5.6.3
+[5.6.2]: https://github.com/ReactiveBayes/ReactiveMP.jl/compare/v5.6.1...v5.6.2
+[5.6.1]: https://github.com/ReactiveBayes/ReactiveMP.jl/compare/v5.6.0...v5.6.1
+[5.6.0]: https://github.com/ReactiveBayes/ReactiveMP.jl/compare/v5.5.12...v5.6.0
+[5.5.12]: https://github.com/ReactiveBayes/ReactiveMP.jl/compare/v5.5.11...v5.5.12
+[5.5.11]: https://github.com/ReactiveBayes/ReactiveMP.jl/compare/v5.5.10...v5.5.11
+[5.5.10]: https://github.com/ReactiveBayes/ReactiveMP.jl/compare/v5.5.9...v5.5.10
+[5.5.9]: https://github.com/ReactiveBayes/ReactiveMP.jl/compare/v5.5.8...v5.5.9
+[5.5.8]: https://github.com/ReactiveBayes/ReactiveMP.jl/compare/v5.5.7...v5.5.8
+[5.5.7]: https://github.com/ReactiveBayes/ReactiveMP.jl/compare/v5.5.6...v5.5.7
+[5.5.6]: https://github.com/ReactiveBayes/ReactiveMP.jl/compare/v5.5.5...v5.5.6
+[5.5.5]: https://github.com/ReactiveBayes/ReactiveMP.jl/compare/v5.5.4...v5.5.5

CONTRIBUTING.md

@@ -0,0 +1,38 @@
+# Contributing to ReactiveMP
+
+**ReactiveMP.jl** is a community-driven reactive message passing engine and we welcome contributions of all kinds!
+
+ReactiveMP is part of the [ReactiveBayes](https://github.com/ReactiveBayes) organization and the broader [RxInfer](https://github.com/ReactiveBayes/RxInfer.jl) ecosystem for reactive Bayesian inference in Julia.
+
+## Getting Started
+
+* Browse **beginner-friendly issues**: https://github.com/ReactiveBayes/ReactiveMP.jl/issues?q=is%3Aopen+is%3Aissue+label%3A%22good+first+issue%22
+* Check out the [RxInfer Contributing Guide](https://docs.rxinfer.com/stable/contributing/guide) for general conventions shared across the ecosystem
+
+We welcome contributions such as:
+
+* Bug reports and fixes
+* New message passing rules and nodes
+* Documentation improvements
+* Tests and performance improvements
+* Feature suggestions
+
+## Development Workflow
+
+1. Fork the repository and create a feature branch
+2. Make your changes and ensure tests pass locally with `make test`
+3. Run `make format` to ensure consistent code formatting
+4. Open a pull request against `main` — the CI will check tests, formatting, and that `CHANGELOG.md` has been updated
+
+## Contributing to the RxInfer Ecosystem
+
+ReactiveMP is one of several packages in the RxInfer ecosystem. Contributions to any of these projects are very welcome:
+
+- [RxInfer.jl](https://github.com/ReactiveBayes/RxInfer.jl) — the high-level inference package
+- [GraphPPL.jl](https://github.com/ReactiveBayes/GraphPPL.jl) — probabilistic model specification
+- [ExponentialFamily.jl](https://github.com/ReactiveBayes/ExponentialFamily.jl) — exponential family distributions
+- Browse other packages at https://github.com/ReactiveBayes
+
+If you're unsure where to start or where a contribution belongs, feel free to open an issue or start a discussion.
+
+Thank you for helping improve ReactiveMP!

Project.toml

@@ -1,7 +1,7 @@
 name = "ReactiveMP"
 uuid = "a194aa59-28ba-4574-a09c-4a745416d6e3"
-version = "5.6.6"
 authors = ["Dmitry Bagaev <d.v.bagaev@tue.nl>", "Albert Podusenko <a.podusenko@tue.nl>", "Bart van Erp <b.v.erp@tue.nl>", "Ismail Senoz <i.senoz@tue.nl>"]
+version = "6.0.0"
 
 [deps]
 BayesBase = "b4ee3484-f114-42fe-b91c-797d54a0c67e"
@@ -30,6 +30,7 @@ StatsFuns = "4c63d2b9-4356-54db-8cca-17b64c39e42c"
 TinyHugeNumbers = "783c9a47-75a3-44ac-a16b-f1ab7b3acf04"
 Tullio = "bc48ee85-29a4-5162-ae0b-a64e1601d4bc"
 TupleTools = "9d95972d-f1c8-5527-a6e0-b4b365fa01f6"
+UUIDs = "cf7118a7-6976-5b1a-9a39-7adc72f591a4"
 
 [weakdeps]
 ExponentialFamilyProjection = "17f509fa-9a96-44ba-99b2-1c5f01f0931b"
@@ -68,6 +69,7 @@ StatsFuns = "1.3.0"
 TinyHugeNumbers = "1.0.0"
 Tullio = "0.3"
 TupleTools = "1.2.0"
+UUIDs = "1"
 julia = "1.10"
 
 [extras]
@@ -87,9 +89,9 @@ Optimisers = "3bd65402-5787-11e9-1adc-39752487f4e2"
 Pkg = "44cfe95a-1eb2-52ea-b672-e2afdf69b78f"
 PkgBenchmark = "32113eaa-f34f-5b0d-bd6c-c81e245fc73d"
 REPL = "3fa0cd96-eef1-5676-8a61-b3b8758bbffb"
-TestItemRunner = "f8b46487-2199-4994-9208-9a1283c18c0a"
 StableRNGs = "860ef19b-820b-49d6-a774-d7a799459cd3"
 Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"
+TestItemRunner = "f8b46487-2199-4994-9208-9a1283c18c0a"
 
 [targets]
 test = ["Aqua", "TestItemRunner", "Test", "Pkg", "Logging", "InteractiveUtils", "Coverage", "Dates", "Distributed", "Documenter", "BenchmarkTools", "JET", "PkgBenchmark", "StableRNGs", "Optimisers", "DiffResults", "ExponentialFamilyProjection", "REPL", "Manopt"]