Adding bayesian numerical tests

test/Project.toml

@@ -2,5 +2,6 @@
 Logging = "56ddb016-857b-54e1-b83d-db4d58db5568"
 RDatasets = "ce6b1742-4840-55fa-b093-852dadbb1d8b"
 StableRNGs = "860ef19b-820b-49d6-a774-d7a799459cd3"
+Statistics = "10745b16-79ce-11e8-11f9-7d13ad32a3b2"
 StatsModels = "3eaba693-59b7-5ba5-a881-562e759f1c8d"
 Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"

test/numerical/bayesian/LinearRegression.jl

@@ -0,0 +1,23 @@
+mtcars = dataset("datasets", "mtcars")
+
+tests = [
+    (Prior_Ridge(), 20.080877893580514),
+    (Prior_Laplace(), 20.070783434589128),
+    (Prior_Cauchy(), 20.019759144845644),
+    (Prior_TDist(), 20.042614331921428),
+    (Prior_HorseShoe(), 20.042984550677183),
+]
+
+for (prior, test_mean) in tests
+    CRRao.set_rng(StableRNG(123))
+    model = fit(@formula(MPG ~ HP + WT + Gear), mtcars, LinearRegression(), prior)
+
+    @test mean(predict(model, mtcars)) ≈ test_mean
+end
+
+gauss_test = 20.0796026428345
+
+CRRao.set_rng(StableRNG(123))
+model = fit(@formula(MPG ~ HP + WT + Gear), mtcars, LinearRegression(), Prior_Gauss(), 30.0, [0.0,-3.0,1.0], 1000)
+
+@test mean(predict(model, mtcars)) ≈ gauss_test

test/numerical/bayesian/LogisticRegression.jl

@@ -0,0 +1,58 @@
+turnout = dataset("Zelig", "turnout")[1:100,:] # Take a subset of rows to reduce input size
+
+tests = [
+    (
+        Prior_Ridge(),
+        (
+            (Logit(), 0.7690822208626806),
+            (Probit(), 0.7685999218881091),
+            (Cloglog(), 0.7751111243871245),
+            (Cauchit(), 0.7730511118602764)
+        )
+    ),
+    (
+        Prior_Laplace(),
+        (
+            (Logit(), 0.7718593681922629),
+            (Probit(), 0.7695587585010469),
+            (Cloglog(), 0.7714870967902365),
+            (Cauchit(), 0.7714839338283468)
+        )
+    ),
+    (
+        Prior_Cauchy(),
+        (
+            (Logit(), 0.7678814727043146),
+            (Probit(), 0.764699194194744),
+            (Cloglog(), 0.7642369367775604),
+            (Cauchit(), 0.7692152829967064)
+        )
+    ),
+    (
+        Prior_TDist(),
+        (
+            (Logit(), 0.588835403024102),
+            (Probit(), 0.7635642627091132),
+            (Cloglog(), 0.7609943137312546),
+            (Cauchit(), 0.772095066757767)
+        )
+    ),
+    (
+        Prior_HorseShoe(),
+        (
+            (Logit(), 0.38683395333332327),
+            (Probit(), 0.38253233489484173),
+            (Cloglog(), 0.7667553778881738),
+            (Cauchit(), 0.7706755564626601)
+        )
+    ),
+]
+
+for (prior, prior_testcases) in tests
+    for (link, test_mean) in prior_testcases
+        CRRao.set_rng(StableRNG(123))
+        model = fit(@formula(Vote ~ Age + Race + Income + Educate), turnout, LogisticRegression(), link, prior)
+
+        @test mean(predict(model, turnout)) ≈ test_mean
+    end
+end

test/numerical/bayesian/NegBinomialRegression.jl

@@ -0,0 +1,16 @@
+sanction = dataset("Zelig", "sanction")
+
+tests = [
+    (Prior_Ridge(), 6.8753100988051274),
+    (Prior_Laplace(), 6.908332048475347),
+    (Prior_Cauchy(), 6.9829255933233645),
+    (Prior_TDist(), 6.915515248823249),
+    (Prior_HorseShoe(), 6.703023191644206),
+]
+
+for (prior, test_mean) in tests
+    CRRao.set_rng(StableRNG(123))
+    model = fit(@formula(Num ~ Target + Coop + NCost), sanction, NegBinomRegression(), prior)
+
+    @test mean(predict(model, sanction)) ≈ test_mean
+end

test/numerical/bayesian/PoissonRegression.jl

@@ -0,0 +1,16 @@
+sanction = dataset("Zelig", "sanction")
+
+tests = [
+    (Prior_Ridge(), 7.177578002644547),
+    (Prior_Laplace(), 7.1454141602741785),
+    (Prior_Cauchy(), 7.148699646242317),
+    (Prior_TDist(), 7.165968828611132),
+    (Prior_HorseShoe(), 7.144190707091213),
+]
+
+for (prior, test_mean) in tests
+    CRRao.set_rng(StableRNG(123))
+    model = fit(@formula(Num ~ Target + Coop + NCost), sanction, PoissonRegression(), prior)
+
+    @test mean(predict(model, sanction)) ≈ test_mean
+end

test/runtests.jl

@@ -1,4 +1,4 @@
-using CRRao, Test, StableRNGs, Logging, RDatasets, StatsModels
+using CRRao, Test, StableRNGs, Logging, RDatasets, StatsModels, Statistics
 
 Logging.disable_logging(Logging.Warn)
 
@@ -23,4 +23,22 @@ CRRao.set_rng(StableRNG(123))
             include("basic/NegBinomialRegression.jl")
         end
     end
+
+    @testset "Numerical Tests" begin
+        @testset "Linear Regression" begin
+            include("numerical/bayesian/LinearRegression.jl")
+        end
+
+        @testset "Logistic Regression" begin
+            include("numerical/bayesian/LogisticRegression.jl")
+        end
+
+        @testset "Poisson Regression" begin
+            include("numerical/bayesian/PoissonRegression.jl")
+        end
+
+        @testset "Negative Binomial Regression" begin
+            include("numerical/bayesian/NegBinomialRegression.jl")
+        end
+    end
 end