allow redefinition of inputs in logprob (#192)

Co-authored-by: David Widmann <[email protected]>

Author: mohamed82008

Project.toml

@@ -1,6 +1,6 @@
 name = "DynamicPPL"
 uuid = "366bfd00-2699-11ea-058f-f148b4cae6d8"
-version = "0.9.7"
+version = "0.9.8"
 
 [deps]
 AbstractMCMC = "80f14c24-f653-4e6a-9b94-39d6b0f70001"

src/context_implementations.jl

@@ -30,7 +30,7 @@ function tilde(rng, ctx::PriorContext, sampler, right, vn::VarName, inds, vi)
     return _tilde(rng, sampler, right, vn, vi)
 end
 function tilde(rng, ctx::LikelihoodContext, sampler, right, vn::VarName, inds, vi)
-    if ctx.vars !== nothing
+    if ctx.vars isa NamedTuple && haskey(ctx.vars, getsym(vn))
         vi[vn] = vectorize(right, _getindex(getfield(ctx.vars, getsym(vn)), inds))
         settrans!(vi, false, vn)
     end
@@ -169,7 +169,7 @@ function dot_tilde(
     inds,
     vi,
 )
-    if ctx.vars !== nothing
+    if ctx.vars isa NamedTuple && haskey(ctx.vars, getsym(vn))
         var = _getindex(getfield(ctx.vars, getsym(vn)), inds)
         vns, dist = get_vns_and_dist(right, var, vn)
         set_val!(vi, vns, dist, var)

src/prob_macro.jl

@@ -190,7 +190,7 @@ function Distributions.loglikelihood(
     if isdefined(right, :chain)
         # Element-wise likelihood for each value in chain
         chain = right.chain
-        ctx = LikelihoodContext()
+        ctx = LikelihoodContext(right)
         iters = Iterators.product(1:size(chain, 1), 1:size(chain, 3))
         logps = map(iters) do (sample_idx, chain_idx)
             setval!(vi, chain, sample_idx, chain_idx)

test/prob_macro.jl

@@ -128,4 +128,23 @@ Random.seed!(129)
         chain2 = sample(model2(y, group, n_groups), NUTS(0.65), 2_000; save_state=true)
         logprob"y = y[[1]] | group = group[[1]], n_groups = n_groups, chain = chain2"
     end
+
+    @testset "issue190" begin
+        @model function gdemo(x, y)
+            s ~ InverseGamma(2, 3)
+            m ~ Normal(0, sqrt(s))
+            x ~ filldist(Normal(m, sqrt(s)), length(y))
+            for i in 1:length(y)
+                y[i] ~ Normal(x[i], sqrt(s))
+            end
+        end
+        c = Chains(rand(10, 2), [:m, :s])
+        model_gdemo = gdemo([1.0, 0.0], [1.5, 0.0])
+        r1 = prob"y = [1.5] | chain=c, model = model_gdemo, x = [1.0]"
+        r2 = map(c[:s]) do s
+            # exp(logpdf(..)) not pdf because this is exactly what the prob"" macro does, so we test r1 == r2
+            exp(logpdf(Normal(1, sqrt(s)), 1.5))
+        end
+        @test r1 == r2
+    end
 end