add regression test against master

.gitignore

@@ -1,4 +1,5 @@
 .vscode
 .history
+.DS_Store
 Manifest.toml
 test/Project.toml

Project.toml

@@ -38,6 +38,7 @@ Turing = "fce5fe82-541a-59a6-adf8-730c64b5f9a0"
 Plots = "91a5bcdd-55d7-5caf-9e0b-520d859cae80"
 UnicodePlots = "b8865327-cd53-5732-bb35-84acbb429228"
 Zygote = "e88e6eb3-aa80-5325-afca-941959d7151f"
+BSON = "fbb218c0-5317-5bc6-957e-2ee96dd4b1f0"
 
 [targets]
-test = ["Distributed", "Distributions", "ForwardDiff", "Plots", "MCMCDebugging", "Test", "Turing", "UnicodePlots", "Bijectors", "OrdinaryDiffEq", "Zygote"]
+test = ["Distributed", "Distributions", "ForwardDiff", "Plots", "MCMCDebugging", "BSON", "Test", "Turing", "UnicodePlots", "Bijectors", "OrdinaryDiffEq", "Zygote"]

test/regression.jl

@@ -0,0 +1,63 @@
+using Test, Random, BSON, AdvancedHMC, Distributions, ForwardDiff
+
+is_ef6de39 = isdefined(AdvancedHMC, :EndPointTS)
+
+function simulate_and_compare(hmc_variant)
+    filepath_ef6de39 = "$(@__DIR__)/regression/ef6de39/$hmc_variant.bson"
+
+    Random.seed!(1)
+
+    dim = 5; θ₀ = rand(dim)
+
+    ℓπ(θ) = logpdf(MvNormal(zeros(dim), ones(dim)), θ)
+
+    n_samples, n_adapts = 2_000, 1_000
+
+    metric = DiagEuclideanMetric(dim)
+    hamiltonian = Hamiltonian(metric, ℓπ, ForwardDiff)
+
+    ϵ₀ = find_good_stepsize(hamiltonian, θ₀)
+    integrator = Leapfrog(ϵ₀)
+
+    if is_ef6de39
+        κ = if hmc_variant == "hmc_mh"
+            StaticTrajectory{EndPointTS}(integrator, 10)
+        elseif hmc_variant == "hmc_multi"
+            StaticTrajectory{MultinomialTS}(integrator, 10)
+        elseif hmc_variant == "nuts_slice"
+            NUTS{SliceTS, GeneralisedNoUTurn}(integrator)
+        elseif hmc_variant == "nuts_multi"
+            NUTS{MultinomialTS, GeneralisedNoUTurn}(integrator)
+        end
+    else
+        κ = if hmc_variant == "hmc_mh"
+            HMCKernel(Trajectory(integrator, FixedNSteps(10)), MetropolisTS)
+        elseif hmc_variant == "hmc_multi"
+            HMCKernel(Trajectory(integrator, FixedNSteps(10)), MultinomialTS)
+        elseif hmc_variant == "nuts_slice"
+            HMCKernel(Trajectory(integrator, NoUTurn()), SliceTS)
+        elseif hmc_variant == "nuts_multi"
+            HMCKernel(Trajectory(integrator, NoUTurn()), MultinomialTS)
+        end
+    end
+
+    adaptor = StanHMCAdaptor(MassMatrixAdaptor(metric), StepSizeAdaptor(0.8, ϵ₀))
+
+    samples, stats = sample(
+        hamiltonian, κ, θ₀, n_samples, adaptor, n_adapts; verbose=false, progress=false
+    )
+
+    is_ef6de39 && return BSON.bson(filepath_ef6de39, samples=samples, stats=stats)
+
+    bson_ef6de39 = BSON.load(filepath_ef6de39)
+    @test samples == bson_ef6de39[:samples]
+    @test stats == bson_ef6de39[:stats]
+end
+
+if "REGRESSION_TEST" in keys(ENV) && ENV["REGRESSION_TEST"] == "1"
+    @testset "Regression" begin
+        for hmc_variant in ["hmc_mh", "hmc_multi", "nuts_slice", "nuts_multi"]
+            simulate_and_compare(hmc_variant)
+        end
+    end
+end

test/runtests.jl

@@ -10,6 +10,7 @@ println(ENV)
         "integrator",
         "trajectory",
         "adaptation",
+        "regression",
         "sampler",
         "sampler-vec",
         "demo",