torfjelde

julia> using Bijectors

julia> using ForwardDiff

julia> using ComponentArrays, UnPack

julia> using Bijectors: LeakyReLU, Shift, Scale

julia> # Affine transformation is simply a `Scale` composed with `Shift`
       using Bijectors: Shift, Scale

torfjelde

torfjelde

julia> using Distributions, ForwardDiff

julia> import Distributions: StatsFuns # StatsFuns.jl is used within Distributions.jl, so we just grab from there

julia> ForwardDiff.derivative(λ -> StatsFuns.poiscdf(λ, 1), 1.0)
ERROR: MethodError: no method matching poiscdf(::ForwardDiff.Dual{ForwardDiff.Tag{var"#1#2", Float64}, Float64, 1}, ::Int64)
Closest candidates are:
  poiscdf(::Union{Float64, Int64}, ::Union{Float64, Int64}) at /home/tor/.julia/packages/StatsFuns/zJ1EI/src/rmath.jl:77
Stacktrace:
 [1] (::var"#1#2")(λ::ForwardDiff.Dual{ForwardDiff.Tag{var"#1#2", Float64}, Float64, 1})

torfjelde

julia> using DynamicPPL, Distributions, Bijectors

julia> using Symbolics

julia> using Symbolics: SymbolicUtils

julia> import StatsFuns

julia> # Make stuff from Distributions.jl primitives so that we don't recurse into them.
       @register Normal()

torfjelde

julia> using DynamicPPL

julia> @model function demo1(x)
           x ~ Normal()
       end
demo1 (generic function with 1 method)

julia> @model function demo2(x)
           x ~ Normal()
       end

torfjelde

julia> include("nested_samplers.jl")

julia> @model function demo()
           m ~ Normal()
           x ~ Normal(m, 1)
           y ~ Normal(x, 1)
           return (; m, x, y)
       end
demo (generic function with 2 methods)

torfjelde

using DynamicPPL

varnames_to_ranges(model::DynamicPPL.Model) = varnames_to_ranges(DynamicPPL.VarInfo(model))
varnames_to_ranges(varinfo::DynamicPPL.UntypedVarInfo) = varnames_to_ranges(varinfo.metadata)
function varnames_to_ranges(varinfo::DynamicPPL.TypedVarInfo)
    offset = 0
    dicts = map(varinfo.metadata) do md
        vns2ranges = varnames_to_ranges(md)
        vals = collect(values(vns2ranges))
        vals_offset = map(r -> offset .+ r, vals)

torfjelde

#!/usr/bin/env julia
using ArgParse

# Scenarios:
# 1. Project is specified, but no files => watch project + run `runtests.jl`.
# 2. Project is specified, and files => watch project + run files.

# Command line arguments.
s = ArgParseSettings()
@add_arg_table s begin

torfjelde

using DynamicPPL: OrderedDict, SamplingContext, AbstractContext, IsParent, VarName, Distribution, evaluate!!, VarInfo
import DynamicPPL: tilde_assume, dot_tilde_assume, childcontext, setchildcontext, NodeTrait

Base.@kwdef struct PriorExtractorContext{D,Ctx} <: AbstractContext
    priors::D=OrderedDict{VarName,Any}()
    context::Ctx=SamplingContext()
end

NodeTrait(::PriorExtractorContext) = IsParent()
childcontext(context::PriorExtractorContext) = context.context

torfjelde

julia> using PosteriorDB, StanDistributions, Turing, BridgeStan, LinearAlgebra

julia> # Necessary overloads to make it work with Turing.
       function DynamicPPL.init(rng, dist::StanDistribution, ::DynamicPPL.SampleFromPrior)
           # `init` uses `rand` by default but this is not supported for `StanDistribution`.
           return BridgeStan.param_constrain(dist.model, randn(rng, length(dist)))
       end

julia> function DynamicPPL.with_logabsdet_jacobian_and_reconstruct(f, dist::StanDistribution, x)
           # HACK: This is cheating.