ReactiveMP.jl: a Julia package for automatic Bayesian inference on a factor graph with reactive message passing.

ReactiveMP v3.9.3 Diff since v3.9.2 Closed issues: The documentation needs "how to add nodes and rules" (#296) @call_rule does not play nicely with addons (#332) CVI Support for the Categorical Distribution (#334) Rand for GammaShapeRate is broken. (#337) 'v_a_vT' is slow and named wrong (#343) Create sparse matrix structure for outer product of StandardBasisVector (#344) MethodError: *(::LinearAlgebra.Diagonal{Float64, Vector{Float64}}, ::StandardBasisVector{Float64}) is ambiguous. (#348) Merged pull requests: CompatHelper: bump compat for LazyArrays to 1, (keep existing compat) (#304) (@github-actions[bot]) CompatHelper: bump compat for StatsBase to 0.34, (keep existing compat) (#313) (@github-actions[bot]) Improving the prod function for Gaussian Distributions in CVI Approximation (#335) (@Nimrais) CVI Support for the Categorical Distribution (#336) (@Nimrais) Add Half-normal node (#338) (@albertpod) fix sampling from GammaShapeRate (#339) (@bvdmitri) Fix printing bug (#345) (@bartvanerp) Allow returning addons from the @call_rule (#346) (@bvdmitri) Fix v_a_vT and StandardBasisVector (#347) (@bvdmitri) Fix ambiguity error for * between Diagonal and StandardBasisVector (#349) (@bvdmitri)