Soss test (update of #62)

test/Project.toml

@@ -7,6 +7,8 @@ FiniteDifferences = "26cc04aa-876d-5657-8c51-4c34ba976000"
 LinearAlgebra = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"
 Plots = "91a5bcdd-55d7-5caf-9e0b-520d859cae80"
 Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
+SampleChainsDynamicHMC = "6d9fd711-e8b2-4778-9c70-c1dfb499d4c4"
+Soss = "8ce77f84-9b61-11e8-39ff-d17a774bf41c"
 Statistics = "10745b16-79ce-11e8-11f9-7d13ad32a3b2"
 Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"
 Turing = "fce5fe82-541a-59a6-adf8-730c64b5f9a0"
@@ -19,6 +21,8 @@ Documenter = "0.24, 0.25, 0.26"
 FillArrays = "0.11"
 FiniteDifferences = "0.9.6, 0.10, 0.11, 0.12"
 Plots = "1"
+SampleChainsDynamicHMC = "0.1"
+Soss = "0.17"
 Turing = "0.14, 0.15"
 Zygote = "0.5, 0.6"
 julia = "1.3"

test/compat/soss.jl

@@ -0,0 +1,51 @@
+@testset "soss compat" begin
+    @testset "GP regression" begin
+        k = SqExponentialKernel()
+        y = randn(3)
+        X = randn(3, 1)
+        x = [rand(1) for _ in 1:3]
+
+        gp_regression = Soss.@model X begin
+            # Priors.
+            α ~ LogNormal(0.0, 0.1)
+            ρ ~ LogNormal(0.0, 1.0)
+            σ² ~ LogNormal(0.0, 1.0)
+
+            # Realized covariance function
+            kernel = α * transform(SqExponentialKernel(), 1 / ρ)
+            f = GP(kernel)
+
+            # Sampling Distribution.
+            y ~ f(X, σ²)
+        end
+
+        # Test for matrices
+        m = gp_regression(; X=RowVecs(X))
+        @test length(Soss.sample(DynamicHMCChain, (m | (y=y,)), 5)) == 5
+
+        # Test for vectors of vector
+        m = gp_regression(; X=x)
+        @test length(Soss.sample(DynamicHMCChain, (m | (y=y,)), 5)) == 5
+    end
+    @testset "latent GP regression" begin
+        X = randn(3, 1)
+        x = [rand(1) for _ in 1:3]
+        y = rand.(Poisson.(exp.(randn(3))))
+
+        latent_gp_regression = Soss.@model X begin
+            f = GP(Matern32Kernel())
+            u ~ f(X)
+            λ = exp.(u)
+            y ~ For(eachindex(λ)) do i
+                Poisson(λ[i])
+            end
+        end
+
+        m = latent_gp_regression(; X=RowVecs(X))
+        @test length(Soss.sample(DynamicHMCChain, (m | (y=y,)), 5)) == 5
+
+        # Test for vectors of vector
+        m = latent_gp_regression(; X=x)
+        @test length(Soss.sample(DynamicHMCChain, (m | (y=y,)), 5)) == 5
+    end
+end

test/turing.jl → test/compat/turing.jl

@@ -4,7 +4,8 @@
         y = randn(3)
         X = randn(3, 1)
         x = [rand(1) for _ in 1:3]
-        @model function GPRegression(y, X)
+
+        Turing.@model function GPRegression(y, X)
             # Priors.
             α ~ LogNormal(0.0, 0.1)
             ρ ~ LogNormal(0.0, 1.0)
@@ -19,26 +20,26 @@
         end
         # Test for matrices
         m = GPRegression(y, RowVecs(X))
-        @test length(sample(m, HMC(0.5, 1), 5)) == 5
+        @test length(sample(m, Turing.HMC(0.5, 1), 5)) == 5
         # Test for vectors of vector
         m = GPRegression(y, x)
-        @test length(sample(m, HMC(0.5, 1), 5)) == 5
+        @test length(sample(m, Turing.HMC(0.5, 1), 5)) == 5
     end
     @testset "latent GP regression" begin
         X = randn(3, 1)
         x = [rand(1) for _ in 1:3]
         y = rand.(Poisson.(exp.(randn(3))))
 
-        @model function latent_gp_regression(y, X)
+        Turing.@model function latent_gp_regression(y, X)
             f = GP(Matern32Kernel())
             u ~ f(X)
             λ = exp.(u)
             return y .~ Poisson.(λ)
         end
         m = latent_gp_regression(y, RowVecs(X))
-        @test length(sample(m, NUTS(), 5)) == 5
+        @test length(sample(m, Turing.NUTS(), 5)) == 5
         # Test for vectors of vector
         m = latent_gp_regression(y, x)
-        @test length(sample(m, NUTS(), 5)) == 5
+        @test length(sample(m, Turing.NUTS(), 5)) == 5
     end
 end

test/runtests.jl

@@ -23,17 +23,19 @@ using AbstractGPs:
 
 using Documenter
 using ChainRulesCore
-using Distributions: MvNormal, PDMat
+using Distributions: MvNormal, PDMat, Poisson, LogNormal
 using FillArrays
 using FiniteDifferences
 using FiniteDifferences: j′vp, to_vec
 using LinearAlgebra
 using LinearAlgebra: AbstractTriangular
 using Plots
 using Random
+using SampleChainsDynamicHMC
+using Soss: Soss
 using Statistics
 using Test
-using Turing
+using Turing: Turing
 using Zygote
 
 include("test_util.jl")
@@ -75,9 +77,15 @@ include("test_util.jl")
     println(" ")
     @info "Ran deprecation tests"
 
-    include("turing.jl")
-    println(" ")
-    @info "Ran Turing tests"
+    @testset "compat" begin
+        include(joinpath("compat", "turing.jl"))
+        println(" ")
+        @info "Ran Turing tests"
+
+        include(joinpath("compat", "soss.jl"))
+        println(" ")
+        @info "Ran Soss tests"
+    end
 
     @testset "doctests" begin
         DocMeta.setdocmeta!(