fd-update

Author: tlienart

__site/__generated/A-composing-models.tar.gz

@@ -121,9 +121,9 @@ uuid = "8ba89e20-285c-5b6f-9357-94700520ee1b"
 
 [[deps.Distributions]]
 deps = ["ChainRulesCore", "DensityInterface", "FillArrays", "LinearAlgebra", "PDMats", "Printf", "QuadGK", "Random", "SparseArrays", "SpecialFunctions", "Statistics", "StatsBase", "StatsFuns", "Test"]
-git-tree-sha1 = "6a8dc9f82e5ce28279b6e3e2cea9421154f5bd0d"
+git-tree-sha1 = "97e9e9d0b8303bae296f3bdd1c2b0065dcb7e7ef"
 uuid = "31c24e10-a181-5473-b8eb-7969acd0382f"
-version = "0.25.37"
+version = "0.25.38"
 
 [[deps.DocStringExtensions]]
 deps = ["LibGit2"]
@@ -413,9 +413,9 @@ version = "0.12.3"
 
 [[deps.Parsers]]
 deps = ["Dates"]
-git-tree-sha1 = "d7fa6237da8004be601e19bd6666083056649918"
+git-tree-sha1 = "92f91ba9e5941fc781fecf5494ac1da87bdac775"
 uuid = "69de0a69-1ddd-5017-9359-2bf0b02dc9f0"
-version = "2.1.3"
+version = "2.2.0"
 
 [[deps.Pkg]]
 deps = ["Artifacts", "Dates", "Downloads", "LibGit2", "Libdl", "Logging", "Markdown", "Printf", "REPL", "Random", "SHA", "Serialization", "TOML", "Tar", "UUIDs", "p7zip_jll"]

__site/__generated/A-composing-models/Manifest.toml

@@ -10,17 +10,25 @@ height = [178, 194, 165, 173, 168];
 
 scitype(X.age)
 
-pipe = @pipeline(
-    X -> coerce(X, :age=>Continuous),
-    OneHotEncoder(),
-    KNNRegressor(K=3),
-    target = UnivariateStandardizer());
-
-pipe.knn_regressor.K = 2
+pipe = Pipeline(
+    coercer = X -> coerce(X, :age=>Continuous),
+    one_hot_encoder = OneHotEncoder(),
+    transformed_target_model = TransformedTargetModel(
+        model = KNNRegressor(K=3);
+        target=UnivariateStandardizer()
+    )
+)
+
+pipe.transformed_target_model.model.K = 2
 pipe.one_hot_encoder.drop_last = true;
 
-evaluate(pipe, X, height, resampling=Holdout(),
-         measure=rms) |> pprint
+evaluate(
+    pipe,
+    X,
+    height,
+    resampling=Holdout(),
+    measure=rms
+) |> pprint
 
 # This file was generated using Literate.jl, https://github.com/fredrikekre/Literate.jl
 

__site/__generated/A-composing-models/tutorial-raw.jl

@@ -102,28 +102,31 @@
    "source": [
     "A typical workflow for such data is to one-hot-encode the categorical data and then apply some regression model on the data.\n",
     "Let's say that we want to apply the following steps:\n",
-    "1. standardize the target variable (`:height`)\n",
-    "1. one hot encode the categorical data\n",
-    "1. train a KNN regression model"
+    "1. One hot encode the categorical features in `X`\n",
+    "1. Standardize the target variable (`:height`)\n",
+    "1. Train a KNN regression model on the one hot encoded data and the Standardized target."
    ],
    "metadata": {}
   },
   {
    "cell_type": "markdown",
    "source": [
-    "The `@pipeline` macro helps you define such a simple (non-branching) pipeline of steps to be applied in order:"
+    "The `Pipeline` constructor helps you define such a simple (non-branching) pipeline of steps to be applied in order:"
    ],
    "metadata": {}
   },
   {
    "outputs": [],
    "cell_type": "code",
    "source": [
-    "pipe = @pipeline(\n",
-    "    X -> coerce(X, :age=>Continuous),\n",
-    "    OneHotEncoder(),\n",
-    "    KNNRegressor(K=3),\n",
-    "    target = UnivariateStandardizer());"
+    "pipe = Pipeline(\n",
+    "    coercer = X -> coerce(X, :age=>Continuous),\n",
+    "    one_hot_encoder = OneHotEncoder(),\n",
+    "    transformed_target_model = TransformedTargetModel(\n",
+    "        model = KNNRegressor(K=3);\n",
+    "        target=UnivariateStandardizer()\n",
+    "    )\n",
+    ")"
    ],
    "metadata": {},
    "execution_count": null
@@ -132,7 +135,8 @@
    "cell_type": "markdown",
    "source": [
     "Note the coercion of the `:age` variable to Continuous since `KNNRegressor` expects `Continuous` input.\n",
-    "Note also the `target` keyword where you can specify a transformation of the target variable."
+    "Note also the `TransformedTargetModel` which allows one to learn a transformation (in this case Standardization) of the\n",
+    "target variable to be passed to the `KNNRegressor`."
    ],
    "metadata": {}
   },
@@ -147,7 +151,7 @@
    "outputs": [],
    "cell_type": "code",
    "source": [
-    "pipe.knn_regressor.K = 2\n",
+    "pipe.transformed_target_model.model.K = 2\n",
     "pipe.one_hot_encoder.drop_last = true;"
    ],
    "metadata": {},
@@ -164,8 +168,13 @@
    "outputs": [],
    "cell_type": "code",
    "source": [
-    "evaluate(pipe, X, height, resampling=Holdout(),\n",
-    "         measure=rms) |> pprint"
+    "evaluate(\n",
+    "    pipe,\n",
+    "    X,\n",
+    "    height,\n",
+    "    resampling=Holdout(),\n",
+    "    measure=rms\n",
+    ") |> pprint"
    ],
    "metadata": {},
    "execution_count": null

__site/__generated/A-composing-models/tutorial.ipynb

@@ -34,30 +34,39 @@ scitype(X.age)
 
 # A typical workflow for such data is to one-hot-encode the categorical data and then apply some regression model on the data.
 # Let's say that we want to apply the following steps:
-# 1. standardize the target variable (`:height`)
-# 1. one hot encode the categorical data
-# 1. train a KNN regression model
+# 1. One hot encode the categorical features in `X`
+# 1. Standardize the target variable (`:height`)
+# 1. Train a KNN regression model on the one hot encoded data and the Standardized target.
 
-# The `@pipeline` macro helps you define such a simple (non-branching) pipeline of steps to be applied in order:
+# The `Pipeline` constructor helps you define such a simple (non-branching) pipeline of steps to be applied in order:
 
-pipe = @pipeline(
-    X -> coerce(X, :age=>Continuous),
-    OneHotEncoder(),
-    KNNRegressor(K=3),
-    target = UnivariateStandardizer());
+pipe = Pipeline(
+    coercer = X -> coerce(X, :age=>Continuous),
+    one_hot_encoder = OneHotEncoder(),
+    transformed_target_model = TransformedTargetModel(
+        model = KNNRegressor(K=3);
+        target=UnivariateStandardizer()
+    )
+)
 
 # Note the coercion of the `:age` variable to Continuous since `KNNRegressor` expects `Continuous` input.
-# Note also the `target` keyword where you can specify a transformation of the target variable.
+# Note also the `TransformedTargetModel` which allows one to learn a transformation (in this case Standardization) of the
+# target variable to be passed to the `KNNRegressor`.
 
 # Hyperparameters of this pipeline can be accessed (and set) using dot syntax:
 
-pipe.knn_regressor.K = 2
+pipe.transformed_target_model.model.K = 2
 pipe.one_hot_encoder.drop_last = true;
 
 # Evaluation for a pipe can be done with the `evaluate!` method; implicitly it will construct machines that will contain the fitted parameters etc:
 
-evaluate(pipe, X, height, resampling=Holdout(),
-         measure=rms) |> pprint
+evaluate(
+    pipe,
+    X,
+    height,
+    resampling=Holdout(),
+    measure=rms
+) |> pprint
 
 # This file was generated using Literate.jl, https://github.com/fredrikekre/Literate.jl
 

__site/__generated/A-composing-models/tutorial.jl

@@ -133,9 +133,9 @@ uuid = "8ba89e20-285c-5b6f-9357-94700520ee1b"
 
 [[deps.Distributions]]
 deps = ["ChainRulesCore", "DensityInterface", "FillArrays", "LinearAlgebra", "PDMats", "Printf", "QuadGK", "Random", "SparseArrays", "SpecialFunctions", "Statistics", "StatsBase", "StatsFuns", "Test"]
-git-tree-sha1 = "6a8dc9f82e5ce28279b6e3e2cea9421154f5bd0d"
+git-tree-sha1 = "97e9e9d0b8303bae296f3bdd1c2b0065dcb7e7ef"
 uuid = "31c24e10-a181-5473-b8eb-7969acd0382f"
-version = "0.25.37"
+version = "0.25.38"
 
 [[deps.DocStringExtensions]]
 deps = ["LibGit2"]
@@ -323,9 +323,9 @@ version = "0.2.0"
 
 [[deps.MLJGLMInterface]]
 deps = ["Distributions", "GLM", "MLJModelInterface", "Parameters", "Tables"]
-git-tree-sha1 = "925ae5a51c5b3cc2c66a714bcefbdb8f5950e491"
+git-tree-sha1 = "f9e26c43458be2285e61e96ea18fe7e13aa62007"
 uuid = "caf8df21-4939-456d-ac9c-5fefbfb04c0c"
-version = "0.1.7"
+version = "0.2.0"
 
 [[deps.MLJIteration]]
 deps = ["IterationControl", "MLJBase", "Random"]
@@ -431,9 +431,9 @@ version = "0.12.3"
 
 [[deps.Parsers]]
 deps = ["Dates"]
-git-tree-sha1 = "d7fa6237da8004be601e19bd6666083056649918"
+git-tree-sha1 = "92f91ba9e5941fc781fecf5494ac1da87bdac775"
 uuid = "69de0a69-1ddd-5017-9359-2bf0b02dc9f0"
-version = "2.1.3"
+version = "2.2.0"
 
 [[deps.Pkg]]
 deps = ["Artifacts", "Dates", "Downloads", "LibGit2", "Libdl", "Logging", "Markdown", "Printf", "REPL", "Random", "SHA", "Serialization", "TOML", "Tar", "UUIDs", "p7zip_jll"]
@@ -530,9 +530,9 @@ version = "3.0.0"
 
 [[deps.SentinelArrays]]
 deps = ["Dates", "Random"]
-git-tree-sha1 = "244586bc07462d22aed0113af9c731f2a518c93e"
+git-tree-sha1 = "15dfe6b103c2a993be24404124b8791a09460983"
 uuid = "91c51154-3ec4-41a3-a24f-3f23e20d615c"
-version = "1.3.10"
+version = "1.3.11"
 
 [[deps.Serialization]]
 uuid = "9e88b42a-f829-5b0c-bbe9-9e923198166b"

__site/__generated/A-ensembles-2.tar.gz

@@ -32,15 +32,17 @@ y = copy(dfY1)
 coerce!(X, autotype(X, :string_to_multiclass))
 yv = Vector(y[:, 1])
 
-LinearRegressorPipe = @pipeline(Standardizer(),
-                                OneHotEncoder(drop_last = true),
-                                LinearRegressor())
+LinearRegressorPipe = Pipeline(
+    Standardizer(),
+    OneHotEncoder(drop_last = true),
+    LinearRegressor()
+)
 
 LinearModel = machine(LinearRegressorPipe, X, yv)
 fit!(LinearModel)
 fp = fitted_params(LinearModel)
 
-ŷ = MLJ.predict(LinearModel, Xm)
+ŷ = MLJ.predict(LinearModel, X)
 yhatResponse = [ŷ[i,1].μ for i in 1:nrow(y)]
 residuals = y .- yhatResponse
 r = report(LinearModel)
@@ -62,9 +64,11 @@ coerce!(X, autotype(X, :string_to_multiclass))
 yc = CategoricalArray(y[:, 1])
 yc = coerce(yc, OrderedFactor)
 
-LinearBinaryClassifierPipe = @pipeline(Standardizer(),
-                                       OneHotEncoder(drop_last = true),
-                                       LinearBinaryClassifier())
+LinearBinaryClassifierPipe = Pipeline(
+    Standardizer(),
+    OneHotEncoder(drop_last = true),
+    LinearBinaryClassifier()
+)
 
 LogisticModel = machine(LinearBinaryClassifierPipe, X, yc)
 fit!(LogisticModel)

__site/__generated/A-ensembles-3.tar.gz

@@ -151,9 +151,11 @@
     "coerce!(X, autotype(X, :string_to_multiclass))\n",
     "yv = Vector(y[:, 1])\n",
     "\n",
-    "LinearRegressorPipe = @pipeline(Standardizer(),\n",
-    "                                OneHotEncoder(drop_last = true),\n",
-    "                                LinearRegressor())\n",
+    "LinearRegressorPipe = Pipeline(\n",
+    "    Standardizer(),\n",
+    "    OneHotEncoder(drop_last = true),\n",
+    "    LinearRegressor()\n",
+    ")\n",
     "\n",
     "LinearModel = machine(LinearRegressorPipe, X, yv)\n",
     "fit!(LinearModel)\n",
@@ -175,7 +177,7 @@
    "outputs": [],
    "cell_type": "code",
    "source": [
-    "ŷ = MLJ.predict(LinearModel, Xm)\n",
+    "ŷ = MLJ.predict(LinearModel, X)\n",
     "yhatResponse = [ŷ[i,1].μ for i in 1:nrow(y)]\n",
     "residuals = y .- yhatResponse\n",
     "r = report(LinearModel)\n",
@@ -225,9 +227,11 @@
     "yc = CategoricalArray(y[:, 1])\n",
     "yc = coerce(yc, OrderedFactor)\n",
     "\n",
-    "LinearBinaryClassifierPipe = @pipeline(Standardizer(),\n",
-    "                                       OneHotEncoder(drop_last = true),\n",
-    "                                       LinearBinaryClassifier())\n",
+    "LinearBinaryClassifierPipe = Pipeline(\n",
+    "    Standardizer(),\n",
+    "    OneHotEncoder(drop_last = true),\n",
+    "    LinearBinaryClassifier()\n",
+    ")\n",
     "\n",
     "LogisticModel = machine(LinearBinaryClassifierPipe, X, yc)\n",
     "fit!(LogisticModel)\n",

__site/__generated/A-ensembles.tar.gz

@@ -63,9 +63,11 @@ y = copy(dfY1)
 coerce!(X, autotype(X, :string_to_multiclass))
 yv = Vector(y[:, 1])
 
-LinearRegressorPipe = @pipeline(Standardizer(),
-                                OneHotEncoder(drop_last = true),
-                                LinearRegressor())
+LinearRegressorPipe = Pipeline(
+    Standardizer(),
+    OneHotEncoder(drop_last = true),
+    LinearRegressor()
+)
 
 LinearModel = machine(LinearRegressorPipe, X, yv)
 fit!(LinearModel)
@@ -75,7 +77,7 @@ fp = fitted_params(LinearModel)
 #
 # We can quickly read the results of our models in MLJ.  Remember to compute the accuracy of the linear model.
 
-ŷ = MLJ.predict(LinearModel, Xm)
+ŷ = MLJ.predict(LinearModel, X)
 yhatResponse = [ŷ[i,1].μ for i in 1:nrow(y)]
 residuals = y .- yhatResponse
 r = report(LinearModel)
@@ -101,9 +103,11 @@ coerce!(X, autotype(X, :string_to_multiclass))
 yc = CategoricalArray(y[:, 1])
 yc = coerce(yc, OrderedFactor)
 
-LinearBinaryClassifierPipe = @pipeline(Standardizer(),
-                                       OneHotEncoder(drop_last = true),
-                                       LinearBinaryClassifier())
+LinearBinaryClassifierPipe = Pipeline(
+    Standardizer(),
+    OneHotEncoder(drop_last = true),
+    LinearBinaryClassifier()
+)
 
 LogisticModel = machine(LinearBinaryClassifierPipe, X, yc)
 fit!(LogisticModel)

__site/__generated/A-fit-predict.tar.gz

@@ -14,9 +14,9 @@ uuid = "0dad84c5-d112-42e6-8d28-ef12dabb789f"
 
 [[deps.ArrayInterface]]
 deps = ["Compat", "IfElse", "LinearAlgebra", "Requires", "SparseArrays", "Static"]
-git-tree-sha1 = "d0d82f1c0b651173a4f839d84f662d03f3417740"
+git-tree-sha1 = "ffc6588e17bcfcaa79dfa5b4f417025e755f83fc"
 uuid = "4fba245c-0d91-5ea0-9b3e-6abc04ee57a9"
-version = "4.0.0"
+version = "4.0.1"
 
 [[deps.Artifacts]]
 uuid = "56f22d72-fd6d-98f1-02f0-08ddc0907c33"
@@ -157,9 +157,9 @@ uuid = "8ba89e20-285c-5b6f-9357-94700520ee1b"
 
 [[deps.Distributions]]
 deps = ["ChainRulesCore", "DensityInterface", "FillArrays", "LinearAlgebra", "PDMats", "Printf", "QuadGK", "Random", "SparseArrays", "SpecialFunctions", "Statistics", "StatsBase", "StatsFuns", "Test"]
-git-tree-sha1 = "6a8dc9f82e5ce28279b6e3e2cea9421154f5bd0d"
+git-tree-sha1 = "97e9e9d0b8303bae296f3bdd1c2b0065dcb7e7ef"
 uuid = "31c24e10-a181-5473-b8eb-7969acd0382f"
-version = "0.25.37"
+version = "0.25.38"
 
 [[deps.DocStringExtensions]]
 deps = ["LibGit2"]
@@ -513,9 +513,9 @@ version = "0.12.3"
 
 [[deps.Parsers]]
 deps = ["Dates"]
-git-tree-sha1 = "d7fa6237da8004be601e19bd6666083056649918"
+git-tree-sha1 = "92f91ba9e5941fc781fecf5494ac1da87bdac775"
 uuid = "69de0a69-1ddd-5017-9359-2bf0b02dc9f0"
-version = "2.1.3"
+version = "2.2.0"
 
 [[deps.Pkg]]
 deps = ["Artifacts", "Dates", "Downloads", "LibGit2", "Libdl", "Logging", "Markdown", "Printf", "REPL", "Random", "SHA", "Serialization", "TOML", "Tar", "UUIDs", "p7zip_jll"]
@@ -618,9 +618,9 @@ version = "3.0.0"
 
 [[deps.SentinelArrays]]
 deps = ["Dates", "Random"]
-git-tree-sha1 = "244586bc07462d22aed0113af9c731f2a518c93e"
+git-tree-sha1 = "15dfe6b103c2a993be24404124b8791a09460983"
 uuid = "91c51154-3ec4-41a3-a24f-3f23e20d615c"
-version = "1.3.10"
+version = "1.3.11"
 
 [[deps.Serialization]]
 uuid = "9e88b42a-f829-5b0c-bbe9-9e923198166b"