v0.33.0

DynamicPPL v0.33.0

Diff since v0.32.2

Breaking changes

• values_as_in_model() now requires an extra boolean parameter, specifying whether variables on the lhs of := statements are to be included in the resulting OrderedDict of values. The type signature is now values_as_in_model([rng,] model, include_colon_eq::Bool [, varinfo, context])

Other changes

• Moved the implementation of predict from Turing.jl to DynamicPPL.jl; the user-facing behaviour is otherwise the same
• Improved error message when a user tries to initialise a model with parameters that don't correspond strictly to the underlying VarInfo used

Merged pull requests:

• Move predict from Turing (#716) (@sunxd3)
• Don't include lhs of := in results of predict() (#766) (@penelopeysm)
• Remove extra doctest filters (#769) (@penelopeysm)

Closed issues:

• Taking stochastic control flow a bit more seriously (#25)
• Adopt DensityInterface (#340)
• Remove NamedDist in favour of VarName interpolation (#400)
• Simplify assume/observe design (#402)
• Supporting mutating ADs in models that fill arrays of parameters (#412)
• Name clash caused by submodels is hard to debug (#427)
• Decouple from Distributions.jl (#523)
• Move TestUtils to an extension (#550)
• Issue with ReverseDiff and undef on Julia v1.7 (#612)
• Transfer Turing.Inference.predict to DynamicPPL. (#647)
• Distributions with latent variables (#689)
• Chain returned by predict will contain variables used in := statements (#765)
• Remove unneeded doctest filters and fix tests (#768)
• Do from_internal_transform() etc. require vi and vn as arguments? (#773)