v4.6.0

ReactiveMP v4.6.0

Diff since v4.5.0

Merged pull requests:

• Generic implementaton of Transition node (#439) (@wouterwln)

v4.5.0

ReactiveMP v4.5.0

Diff since v4.4.6

Merged pull requests:

• Add BinomialPolya node with PolyaGamma augmentation for Binomial regression (#438) (@ismailsenoz)

v4.4.6

ReactiveMP v4.4.6

Diff since v4.4.5

Merged pull requests:

• fix #434: determine threads/cores via extra package Hwloc, if CpuId fails (#435) (@abpolym)
• Enhance functional dependencies customization in ReactiveMP (#437) (@bvdmitri)

Closed issues:

• Parallelization in runtests.jl using ReTestItems and CpuId will not work for Apple M1 Pro CPU (#434)
• Allow meta object to change default functional dependencies (#436)

v4.4.5

ReactiveMP v4.4.5

Diff since v4.4.4

Merged pull requests:

• Ux rule suggestions (#423) (@wmkouw)
• CVIProjection with constraints on interfaces (#428) (@Nimrais)
• Add a hint for the undefined factor nodes (#431) (@bvdmitri)

Closed issues:

• Expand RuleMethodError with list of defined rules for given node (#397)

v4.4.4

ReactiveMP v4.4.4

Diff since v4.4.3

Merged pull requests:

• Fix tests for Julia v1.11 (#422) (@bvdmitri)
• Update ExponentialFamily.jl dependency (#426) (@bvdmitri)
• add a missing method for the divisionOf (#427) (@bvdmitri)

v4.4.3

ReactiveMP v4.4.3

Diff since v4.4.2

v4.4.2

ReactiveMP v4.4.2

Diff since v4.4.1

Merged pull requests:

• Relax rules for structured VMP in Transition node (#418) (@wouterwln)
• Fix typo in DeferredMessage (#419) (@wouterwln)

v4.4.1

ReactiveMP v4.4.1

Diff since v4.4.0

Merged pull requests:

• Fix for MultiAgentTrajectoryPlanning#4 (#416) (@bvdmitri)
• Cleanup Linearization code (#417) (@bvdmitri)

Closed issues:

• Misleading of code in src/approximations/linearization.jl (#415)

v4.4.0

ReactiveMP v4.4.0

Diff since v4.3.2

Merged pull requests:

• Move the check for ProductOf to RxInfer (#414) (@bvdmitri)

v4.3.2

ReactiveMP v4.3.2

Diff since v4.3.1

Merged pull requests:

• Add conditioners to ProjectedTo to allow for conditioned distributions in marginal computations (#411) (@ismailsenoz)
• Adapt out message rule for delta node to new syntax of ExponentialFamilyProjection (#413) (@ismailsenoz)