Docs failing to build due to environment issues

.github/workflows/CI.yml

@@ -19,8 +19,8 @@ jobs:
     runs-on: ${{ matrix.os }}
     strategy:
       matrix:
-        julia-version: ['1.10', '1.11']
-        os: [ubuntu-latest, windows-latest, macOS-latest]
+        julia-version: ['1']
+        os: ['ubuntu-latest']
     steps:
       - uses: actions/checkout@v4
 
@@ -30,6 +30,22 @@ jobs:
 
       - uses: julia-actions/cache@v2
 
+      # 1. Install uv and enable caching for faster runs
+      - name: Install uv
+        uses: astral-sh/setup-uv@v5
+        with:
+          enable-cache: true
+
+      # 2. Install your target Python version
+      - name: Set up Python
+        run: uv python install 3.12
+
+      # 3. Auto-setup the virtual environment for the tests
+      - name: Create Python virtual environment
+        run: |
+          uv venv test/test_python/.venv
+          uv pip install --python=test/test_python/.venv numpy pandas doubleml xgboost scikit-learn
+
       - name: Install dependencies
         run: julia --project=. -e 'using Pkg; Pkg.instantiate()'
 
@@ -43,7 +59,7 @@ jobs:
 
       - uses: julia-actions/setup-julia@v2
         with:
-          version: '1.10'
+          version: '1'
 
       - uses: julia-actions/cache@v2
 
@@ -64,12 +80,14 @@ jobs:
 
       - uses: julia-actions/setup-julia@v2
         with:
-          version: '1.10'
+          version: '1'
 
       - uses: julia-actions/cache@v2
 
       - name: Install dependencies
-        run: julia --project=docs -e 'using Pkg; Pkg.develop(PackageSpec(path=pwd())); Pkg.instantiate()'
+        run: |
+          julia --project=docs -e 'using Pkg; Pkg.develop(Pkg.PackageSpec(path=pwd())); Pkg.instantiate()'
+          julia --project=examples -e 'using Pkg; Pkg.instantiate()' # Add this line
 
       - name: Build and Deploy
         env:

.gitignore

@@ -1,3 +1,4 @@
 /.planning
 */.ignore
 .theme
+/bin 

Project.toml

@@ -23,8 +23,8 @@ StatsBase = "2913bbd2-ae8a-5f71-8c99-4fb6c76f3a91"
 StatsFuns = "4c63d2b9-4356-54db-8cca-17b64c39e42c"
 
 [compat]
-CategoricalArrays = "1.0.2"
-MLJBase = "1.12.1"
+CategoricalArrays = "1.0"
+MLJBase = "1.12"
 PrettyTables = "3.2.3"
 Roots = "2.2.12"
 ScientificTypes = "3.2.0"

README.md

@@ -62,14 +62,8 @@ There are many features and models still not yet implemented in this package. Th
 
 Currently, a variety of tests against the Python package are implemented to ensure similar functionality of the DoubleMLPLR, DoubleMLIRM, and DoubleMLLPLR models.
 
-In early benchmarks, the Julia implementation performs well
+In early benchmarks, the Julia implementation performs well and up to 10x faster than the Python package (see the [benchmark](test/test_python/benchmarks/benchmarks.md))
 
 ## Other packages
 
 Other similar Julia packages include [CausalELM](https://github.com/dscolby/CausalELM.jl), which offers a very lightweight approach to causal machine learning, where all the machine learners take the form of extreme learning machines. In comparison, this package aims to offer more similar features to those of the DoubleML Python package and allow flexibility of the model choice.
-
-## Index
-
-```@index
-Pages = ["api.md"]
-```

docs/Project.toml

@@ -3,6 +3,7 @@ Documenter = "e30172f5-a6a5-5a46-863b-614d45cd2de4"
 DocumenterTools = "35a29f4d-8980-5a13-9543-d66fff28ecb8"
 DoubleML = "77251493-8087-46d5-be33-49d84ce818f6"
 LiveServer = "16fef848-5104-11e9-1b77-fb7a48bbb589"
+MLJ = "add582a8-e3ab-11e8-2d5e-e98b27df1bc7"
 Pluto = "c3e4b0f8-55cb-11ea-2926-15256bba5781"
 PlutoStaticHTML = "359b1769-a58e-495b-9770-312e911026ad"
 StatsFuns = "4c63d2b9-4356-54db-8cca-17b64c39e42c"

docs/src/notebooks/irm_introduction.md

@@ -25,16 +25,17 @@
 <!--
     # This information is used for caching.
     [PlutoStaticHTML.State]
-    input_sha = "7bd830a3473d17a6cf11e7e59a67a38ef8bc07443c8e514b1512939325a84087"
+    input_sha = "28053b176c99ae50ce44d8973ba9ab3301e74bdddd0bfdde0d45df0bc5a2af05"
     julia_version = "1.12.4"
 -->
 <pre class='language-julia'><code class='language-julia'>begin
-    using Pkg
-    Pkg.activate("..")
-    Pkg.instantiate()
+    import Pkg; Pkg.develop(path=joinpath(@__DIR__, "../.."))
 end</code></pre>
 
 
+<pre class='language-julia'><code class='language-julia'>Pkg.activate(joinpath(@__DIR__, "../../examples"))</code></pre>
+
+
 <pre class='language-julia'><code class='language-julia'>begin
     using DoubleML
     using StableRNGs
@@ -108,7 +109,7 @@ end</code></pre>
 <pre class="code-output documenter-example-output" id="var-dml_irm_simple">────────────────────────────────────────────────────────────────────
    Estimate  Std. Error  z value  Pr(&gt;|z|)  Lower 95.0%  Upper 95.0%
 ────────────────────────────────────────────────────────────────────
-d   0.67957   0.0189615    35.84    &lt;1e-99     0.642406     0.716733
+d  0.674031   0.0183706    36.69    &lt;1e-99     0.638025     0.710037
 ────────────────────────────────────────────────────────────────────</pre>
 
 <pre class='language-julia'><code class='language-julia'>begin
@@ -117,7 +118,7 @@ d   0.67957   0.0189615    35.84    &lt;1e-99     0.642406     0.716733
     space = Dict(
         :max_depth =&gt; HP.QuantUniform(:max_depth, 2.0, 12.0, 1.0)
     )
-    
+
     tuned_ml_g = TunedModel(
         model = EvoTreeRegressor(),
         tuning = MLJTreeParzenTuning(random_trials = 75, draws = 50),
@@ -144,13 +145,13 @@ d   0.67957   0.0189615    35.84    &lt;1e-99     0.642406     0.716733
 end</code></pre>
 <pre class="code-output documenter-example-output" id="var-tuned_ml_m">DoubleMLIRM{Float32, MLJTuning.DeterministicTunedModel{MLJTreeParzenTuning, EvoTrees.EvoTreeRegressor, Nothing}, MLJTuning.ProbabilisticTunedModel{MLJTreeParzenTuning, EvoTrees.EvoTreeClassifier, Nothing}}
 ==========================
-StatsBase.CoefTable(Any[[0.5067636370658875], [0.020287886261940002], [24.9786319732666], [1.0436473597475374e-137], [0.46700011067004005], [0.5465271634617349]], ["Estimate", "Std. Error", "z value", "Pr(&gt;|z|)", "Lower 95.0%", "Upper 95.0%"], ["d"], 4, 3)</pre>
+StatsBase.CoefTable(Any[[0.5573905110359192], [0.04278203845024109], [13.028610229492188], [8.412598772482777e-39], [0.4735392564882387], [0.6412417655835997]], ["Estimate", "Std. Error", "z value", "Pr(&gt;|z|)", "Lower 95.0%", "Upper 95.0%"], ["d"], 4, 3)</pre>
 
 <pre class='language-julia'><code class='language-julia'>coeftable(dml_irm)</code></pre>
 <pre class="code-output documenter-example-output" id="var-hash997185">────────────────────────────────────────────────────────────────────
    Estimate  Std. Error  z value  Pr(&gt;|z|)  Lower 95.0%  Upper 95.0%
 ────────────────────────────────────────────────────────────────────
-d  0.506764   0.0202879    24.98    &lt;1e-99        0.467     0.546527
+d  0.557391    0.042782    13.03    &lt;1e-38     0.473539     0.641242
 ────────────────────────────────────────────────────────────────────</pre>
 
 <!-- PlutoStaticHTML.End -->

docs/src/notebooks/lplr_introduction.md

@@ -25,15 +25,13 @@
 <!--
     # This information is used for caching.
     [PlutoStaticHTML.State]
-    input_sha = "131c5e789bd07e513a5b6235045baf6c603c1ffeae0f97f9675c26111fa2be71"
+    input_sha = "608f5ab3ff7f9c3c1e0165b0fe879290984a7cff3fbc7bccfa761a3811b5dbb4"
     julia_version = "1.12.4"
 -->
-<pre class='language-julia'><code class='language-julia'>begin
-    using Pkg
-    Pkg.activate("..")
-    Pkg.resolve()
-    Pkg.instantiate()
-end</code></pre>
+<pre class='language-julia'><code class='language-julia'>import Pkg; Pkg.develop(path=joinpath(@__DIR__, "../.."))</code></pre>
+
+
+<pre class='language-julia'><code class='language-julia'>Pkg.activate(joinpath(@__DIR__, "../../examples"))</code></pre>
 
 
 <pre class='language-julia'><code class='language-julia'>begin
@@ -57,7 +55,7 @@ end</code></pre>
 <div class="markdown"><h2 id="Generate-LPLR-data">Generate LPLR data</h2></div>
 
 <pre class='language-julia'><code class='language-julia'>
-data_lplr = make_lplr_LZZ2020(1000, alpha=0.5, rng=StableRNG(42))</code></pre>
+data_lplr = make_lplr_LZZ2020(1000, alpha = 0.5, rng = StableRNG(42))</code></pre>
 <pre class="code-output documenter-example-output" id="var-data_lplr">DoubleMLData{Float32, Vector{Float32}}(Float32[1.0, 1.0, 0.0, 0.0, 0.0, 1.0, 0.0, 1.0, 1.0, 1.0  …  0.0, 0.0, 1.0, 1.0, 0.0, 1.0, 1.0, 0.0, 1.0, 0.0], Float32[1.137392, -1.0203186, -1.2559067, 0.35913706, -0.98865694, 2.5477915, 1.6435643, -0.2593958, 2.7788334, 0.5610351  …  1.3883054, 1.6827291, 1.2748861, 1.5950062, 0.8413675, 0.1603024, 0.6246069, 0.86354524, 1.0373387, -1.603643], Float32[-0.67025167 0.3040378 … 0.9723363 -0.6289253; 2.0 -1.3471706 … -0.49255872 1.9596015; … ; -0.5976384 0.034506414 … 0.3402903 -0.70474565; 0.32168102 0.52544063 … 0.5724332 0.40678954], 1000, 20, :y, :d, [:X1, :X2, :X3, :X4, :X5, :X6, :X7, :X8, :X9, :X10, :X11, :X12, :X13, :X14, :X15, :X16, :X17, :X18, :X19, :X20])</pre>
 
 
@@ -117,13 +115,13 @@ end</code></pre>
 end</code></pre>
 <pre class="code-output documenter-example-output" id="var-ml_t">DoubleMLLPLR{Float32, MLJDecisionTreeInterface.RandomForestClassifier, MLJDecisionTreeInterface.RandomForestRegressor, MLJDecisionTreeInterface.RandomForestRegressor, MLJDecisionTreeInterface.RandomForestRegressor}
 ==========================
-StatsBase.CoefTable(Any[[0.5981336832046509], [0.07487558573484421], [7.988367557525635], [1.3673725833843347e-15], [0.4513802318430149], [0.7448871345662869]], ["Estimate", "Std. Error", "z value", "Pr(&gt;|z|)", "Lower 95.0%", "Upper 95.0%"], ["d"], 4, 3)</pre>
+StatsBase.CoefTable(Any[[0.5567321181297302], [0.0747859850525856], [7.444337368011475], [9.743221856517205e-14], [0.41015428087831135], [0.703309955381149]], ["Estimate", "Std. Error", "z value", "Pr(&gt;|z|)", "Lower 95.0%", "Upper 95.0%"], ["d"], 4, 3)</pre>
 
 <pre class='language-julia'><code class='language-julia'>coeftable(dml_lplr_simple)</code></pre>
 <pre class="code-output documenter-example-output" id="var-hash360715">────────────────────────────────────────────────────────────────────
    Estimate  Std. Error  z value  Pr(&gt;|z|)  Lower 95.0%  Upper 95.0%
 ────────────────────────────────────────────────────────────────────
-d  0.598134   0.0748756     7.99    &lt;1e-14      0.45138     0.744887
+d  0.556732    0.074786     7.44    &lt;1e-13     0.410154      0.70331
 ────────────────────────────────────────────────────────────────────</pre>
 
 
@@ -134,11 +132,11 @@ d  0.598134   0.0748756     7.99    &lt;1e-14      0.45138     0.744887
     controls = [
         Step(1),
         Patience(10),
-        NumberLimit(20)
+        NumberLimit(20),
     ]
-    
+
     ml_M_iterated = IteratedModel(
-        EvoTreeClassifier(max_depth = 4, eta = 0.05), 
+        EvoTreeClassifier(max_depth = 4, eta = 0.05),
         resampling = Holdout(),
         measure = cross_entropy,
         iteration_parameter = :nrounds,
@@ -154,7 +152,7 @@ d  0.598134   0.0748756     7.99    &lt;1e-14      0.45138     0.744887
     )
 
     ml_m_iterated = IteratedModel(
-        EvoTreeRegressor(max_depth = 4, eta = 0.05), 
+        EvoTreeRegressor(max_depth = 4, eta = 0.05),
         resampling = Holdout(),
         measure = mae,
         iteration_parameter = :nrounds,
@@ -171,13 +169,13 @@ d  0.598134   0.0748756     7.99    &lt;1e-14      0.45138     0.744887
 end</code></pre>
 <pre class="code-output documenter-example-output" id="var-controls">DoubleMLLPLR{Float32, MLJIteration.ProbabilisticIteratedModel{EvoTrees.EvoTreeClassifier}, MLJIteration.DeterministicIteratedModel{EvoTrees.EvoTreeRegressor}, MLJIteration.DeterministicIteratedModel{EvoTrees.EvoTreeRegressor}, MLJIteration.DeterministicIteratedModel{EvoTrees.EvoTreeRegressor}}
 ==========================
-StatsBase.CoefTable(Any[[0.5652134418487549], [0.07496540993452072], [7.539656639099121], [4.712104968941949e-14], [0.41828393829081284], [0.712142945406697]], ["Estimate", "Std. Error", "z value", "Pr(&gt;|z|)", "Lower 95.0%", "Upper 95.0%"], ["d"], 4, 3)</pre>
+StatsBase.CoefTable(Any[[0.5537656545639038], [0.07333461195230484], [7.551218032836914], [4.3120625571213347e-14], [0.41003245631716545], [0.6974988528106422]], ["Estimate", "Std. Error", "z value", "Pr(&gt;|z|)", "Lower 95.0%", "Upper 95.0%"], ["d"], 4, 3)</pre>
 
 <pre class='language-julia'><code class='language-julia'>coeftable(dml_lplr_iterated)</code></pre>
 <pre class="code-output documenter-example-output" id="var-hash950978">────────────────────────────────────────────────────────────────────
    Estimate  Std. Error  z value  Pr(&gt;|z|)  Lower 95.0%  Upper 95.0%
 ────────────────────────────────────────────────────────────────────
-d  0.565213   0.0749654     7.54    &lt;1e-13     0.418284     0.712143
+d  0.553766   0.0733346     7.55    &lt;1e-13     0.410032     0.697499
 ────────────────────────────────────────────────────────────────────</pre>
 
 <!-- PlutoStaticHTML.End -->