ci: test turing models

.github/workflows/models.yml

@@ -0,0 +1,47 @@
+name: Test Models
+on:
+  push:
+    branches:
+      - master
+      - main
+  pull_request:
+    branches:
+      - master
+      - main
+
+jobs:
+  test-turing:
+    name: Test Turing Models
+    runs-on: ubuntu-latest
+    env:
+      JULIA_NUM_THREADS: "auto"
+    strategy:
+      matrix:
+        jl-file: [
+            "01-predictive_checks.jl",
+            "02-linear_regression-kidiq.jl",
+            "03-logistic_regression-wells.jl",
+            "04-ordinal_regression-esoph.jl",
+            "05-poisson_regression-roaches.jl",
+            "06-robust_linear_regression-duncan.jl",
+            "07-robust_beta_binomial_regression-wells.jl",
+            #"08-robust_robit_regression-wells.jl", # FIXME: broken
+            "09-robust_negative_binomial_regression-roaches.jl",
+            "10-robust_zero_inflated_regression-negative_binomial-roaches.jl",
+            "10-robust_zero_inflated_regression-poisson-roaches.jl",
+            "11-hierarchical_varying_intercept-cheese.jl",
+            "11-hierarchical_varying_intercept-non_centered-cheese.jl",
+            "12-hierarchical_varying_intercept_slope-cheese.jl",
+            "13-model_comparison-roaches.jl",
+          ]
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+      - uses: julia-actions/setup-julia@v1
+        with:
+          version: "1"
+      - uses: julia-actions/cache@v1
+      - name: Test ${{ matrix.jl-file }}
+        run: |
+          julia --project=turing -e 'using Pkg; Pkg.instantiate()'
+          julia --project=turing turing/${{ matrix.jl-file }}

flake.nix

@@ -39,7 +39,11 @@
             julia
           ];
 
-          inherit (self.checks.${system}.pre-commit-check) shellHook;
+          shellHook = ''
+            export JULIA_NUM_THREADS="auto"
+            export JULIA_PROJECT="turing"
+            ${self.checks.${system}.pre-commit-check.shellHook}
+          '';
         };
         packages.default = pkgs.stdenvNoCC.mkDerivation rec {
           name = "slides";

turing/04-ordinal_regression-esoph.jl

@@ -1,9 +1,10 @@
 using Turing
-using Bijectors
 using CSV
 using DataFrames
 using LinearAlgebra
 using CategoricalArrays
+using Bijectors: transformed, OrderedBijector
+using DataFrames: transform!
 
 # reproducibility
 using Random: seed!
@@ -36,7 +37,7 @@ y = esoph[:, :tobgp]
 # define the model
 @model function ordered_regression(X, y; predictors=size(X, 2), ncateg=maximum(y))
     # priors
-    cutpoints ~ Bijectors.ordered(filldist(TDist(3) * 5, ncateg - 1))
+    cutpoints ~ transformed(filldist(TDist(3) * 5, ncateg - 1), OrderedBijector())
     β ~ filldist(TDist(3) * 2.5, predictors)
 
     # likelihood

turing/07-robust_beta_binomial_regression-wells.jl

@@ -24,7 +24,14 @@ X = standardize(ZScoreTransform, X; dims=1)
 y = wells[:, :switch]
 
 # define alternate parameterizations
-BetaBinomial2(n, μ, ϕ) = BetaBinomial(n, μ * ϕ, (1 - μ) * ϕ)
+function BetaBinomial2(n, μ, ϕ)
+    α = μ * ϕ
+    β = (1 - μ) * ϕ
+    α = α > 0 ? α : 1e-4 # numerical stability
+    β = β > 0 ? β : 1e-4 # numerical stability
+
+    return BetaBinomial(n, α, β)
+end
 
 # define the model
 @model function beta_binomial_regression(X, y; predictors=size(X, 2))

turing/08-robust_robit_regression-wells.jl

@@ -22,7 +22,7 @@ y = wells[:, :switch]
 
 # define the model
 # setting ν to 4 since we have a lot of parameters
-# and Turing v0.21.9 samples slower than Stan
+# and Turing v0.29 samples slower than Stan
 @model function robit_regression(X, y; predictors=size(X, 2), N=size(X, 1), ν=4)
     # priors
     α ~ TDist(3) * 2.5

turing/13-model_comparison-roaches.jl

@@ -1,3 +1,10 @@
+# FIXME: ParetoSmooth is not compatible with Turing 0.29
+using Pkg
+Pkg.compat("Turing", "0.22")
+Pkg.compat("Bijectors", "0.10")
+Pkg.compat("StatsBase", "0.33")
+Pkg.add("ParetoSmooth")
+
 using Turing
 using CSV
 using DataFrames

turing/Project.toml

@@ -5,7 +5,6 @@ CategoricalArrays = "324d7699-5711-5eae-9e2f-1d82baa6b597"
 DataFrames = "a93c6f00-e57d-5684-b7b6-d8193f3e46c0"
 LazyArrays = "5078a376-72f3-5289-bfd5-ec5146d43c02"
 PDMats = "90014a1f-27ba-587c-ab20-58faa44d9150"
-ParetoSmooth = "a68b5a21-f429-434e-8bfa-46b447300aac"
 StatsBase = "2913bbd2-ae8a-5f71-8c99-4fb6c76f3a91"
 StatsFuns = "4c63d2b9-4356-54db-8cca-17b64c39e42c"
 TransformVariables = "84d833dd-6860-57f9-a1a7-6da5db126cff"
@@ -18,7 +17,6 @@ CategoricalArrays = "0.10"
 DataFrames = "1"
 LazyArrays = "1"
 PDMats = "0.11"
-ParetoSmooth = "0.7"
 StatsBase = "0.34"
 StatsFuns = "1"
 TransformVariables = "0.8"