Merge pull request #206 from JuliaAI/dev

Generate new docs.

docs/make.jl

@@ -10,26 +10,30 @@ makedocs(;
              "Quick-start guide" => "quick_start_guide.md",
              "The model type hierarchy" => "the_model_type_hierarchy.md",
              "New model type declarations" => "type_declarations.md",
-             "Supervised models" => "supervised_models.md",
-             "Summary of methods" => "summary_of_methods.md",
-             "The form of data for fitting and predicting" => "form_of_data.md",
-             "The fit method" => "the_fit_method.md",
-             "The fitted_params method" => "the_fitted_params_method.md",
-             "The predict method" => "the_predict_method.md",
-             "The predict_joint method" => "the_predict_joint_method.md",
-             "Training losses" => "training_losses.md",
-             "Feature importances" =>  "feature_importances.md",
-             "Trait declarations" => "trait_declarations.md",
-             "Iterative models and the update! method" => "iterative_models.md",
-             "Implementing a data front end" => "implementing_a_data_front_end.md",
-             "Supervised models with a transform method" =>
-                 "supervised_models_with_transform.md",
-             "Models that learn a probability distribution" => "fitting_distributions.md",
-             "Serialization" => "serialization.md",
-             "Document strings" => "document_strings.md",
+             "Supervised models" => [
+                 "Introduction" => "supervised_models.md",
+                 "Summary of methods" => "summary_of_methods.md",
+                 "The form of data for fitting and predicting" => "form_of_data.md",
+                 "The fit method" => "the_fit_method.md",
+                 "The fitted_params method" => "the_fitted_params_method.md",
+                 "The predict method" => "the_predict_method.md",
+                 "The predict_joint method" => "the_predict_joint_method.md",
+                 "Training losses" => "training_losses.md",
+                 "Feature importances" =>  "feature_importances.md",
+                 "Trait declarations" => "trait_declarations.md",
+                 "Iterative models and the update! method" => "iterative_models.md",
+                 "Implementing a data front end" => "implementing_a_data_front_end.md",
+                 "Supervised models with a transform method" =>
+                     "supervised_models_with_transform.md",
+                 "Models that learn a probability distribution" =>
+                     "fitting_distributions.md",
+             ],
              "Unsupervised models" => "unsupervised_models.md",
              "Static models" => "static_models.md",
              "Outlier detection models" => "outlier_detection_models.md",
+             "Model wrappers" => "model_wrappers.md",
+             "Serialization" => "serialization.md",
+             "Document strings" => "document_strings.md",
              "Convenience methods" => "convenience_methods.md",
              "Where to place code implementing new models" => "where_to_put_code.md",
              "How to add models to the MLJ Model Registry" => "how_to_register.md",

docs/src/index.md

@@ -8,7 +8,7 @@ API defined there, as outlined in this document.
 
 !!! tip
 
-    This is a reference document, which has become rather sprawling over the evolution of the MLJ project. We recommend starting with [Quick start guide](@ref), which covers the main points relevant to most new model implementations.
+    This is a reference document, which has become rather sprawling over the evolution of the MLJ project. We recommend starting with [Quick start guide](@ref), which covers the main points relevant to most new model implementations. Most topics are only detailed for `Supervised` models, so if you are implementing another kind of model, you may still need to refer to the [Supervised models](@ref) section. 
 
 Interface code can be hosted by the package providing the core machine learning algorithm,
 or by a stand-alone "interface-only" package, using the template

docs/src/model_wrappers.md

@@ -0,0 +1,29 @@
+# Model wrappers
+
+A model that can have one or more other models as hyper-parameters should overload the trait `is_wrapper`, as in this example:
+
+```julia
+MLJModelInterface.target_in_fit(::Type{<:MyWrapper}) = true
+```
+
+The constructor for such a model does not need provide default values for the model-valued
+hyper-parameters. If only a single model is wrapped, then the hyper-parameter should have
+the name `:model` and this should be an optional positional argument, as well as a keyword
+argument.
+
+For example, `EnsembleModel` is a model wrapper, and we can construct an instance like this:
+
+```julia
+using MLJ
+atom = ConstantClassfier()
+EnsembleModel(tree, n=100)
+```
+
+but also like this:
+
+```julia
+EnsembleModel(model=tree, n=100)
+```
+
+This is the only case in MLJ where positional arguments in a model constructor are
+allowed.

docs/src/unsupervised_models.md

@@ -5,11 +5,10 @@ similar fashion. The main differences are:
 
 - The `fit` method, which still returns `(fitresult, cache, report)` will typically have
   only one training argument `X`, as in `MLJModelInterface.fit(model, verbosity, X)`,
-  although this is not a hard requirement. For example, a feature selection tool (wrapping
-  some supervised model) might also include a target `y` as input. Furthermore, in the
-  case of models that subtype `Static <: Unsupervised` (see [Static
-  models](@ref)) `fit` has no training arguments at all, but does not need to be
-  implemented as a fallback returns `(nothing, nothing, nothing)`.
+  although this is not a hard requirement; see [Transformers requiring a target variable
+  in training](@ref) below.  Furthermore, in the case of models that subtype `Static <:
+  Unsupervised` (see [Static models](@ref)) `fit` has no training arguments at all, but
+  does not need to be implemented as a fallback returns `(nothing, nothing, nothing)`.
 
 - A `transform` and/or `predict` method is implemented, and has the same signature as
   `predict` does in the supervised case, as in `MLJModelInterface.transform(model,
@@ -27,15 +26,43 @@ similar fashion. The main differences are:
   argument, you must overload the trait `fit_data_scitype`, which bounds the allowed
   `data` passed to `fit(model, verbosity, data...)` and will always be a `Tuple` type.
 
-- An `inverse_transform` can be optionally implemented. The signature
-  is the same as `transform`, as in
-  `MLJModelInterface.inverse_transform(model, fitresult, Xout)`, which:
+- An `inverse_transform` can be optionally implemented. The signature is the same as
+  `transform`, as in `MLJModelInterface.inverse_transform(model::MyUnsupervisedModel,
+  fitresult, Xout)`, which:
    - must make sense for any `Xout` for which `scitype(Xout) <:
-     output_scitype(SomeSupervisedModel)` (see below); and
+     output_scitype(MyUnsupervisedModel)`; and
    - must return an object `Xin` satisfying `scitype(Xin) <:
-     input_scitype(SomeSupervisedModel)`.
+     input_scitype(MyUnsupervisedModel)`.
 
-For sample implementatations, see MLJ's [built-in
+For sample implementations, see MLJ's [built-in
 transformers](https://github.com/JuliaAI/MLJModels.jl/blob/dev/src/builtins/Transformers.jl)
 and the clustering models at
 [MLJClusteringInterface.jl](https://github.com/jbrea/MLJClusteringInterface.jl).
+
+## Transformers requiring a target variable in training
+
+An `Unsupervised` model that is not `Static` may include a second argument `y` in it's
+`fit` signature, as in `fit(::MyTransformer, verbosity, X, y)`. For example, some feature
+selection tools require a target variable `y` in training. (Unlike `Supervised` models, an
+`Unsupervised` model is not required to implement `predict`, and in pipelines it is the
+output of `transform`, and not `predict`, that is always propagated to the next model.) Such a
+model should overload the trait `target_in_fit`, as in this example:
+
+```julia
+MLJModelInterface.target_in_fit(::Type{<:MyTransformer}) = true
+```
+
+This ensures that such models can appear in pipelines, and that a target provided to the
+pipeline model is passed on to the model in training. 
+
+If the model implements more than one `fit` signature (e.g., one with a target `y` and one
+without) then `fit_data_scitype` must also be overloaded, as in this example:
+
+```julia
+MLJModelInterface.fit_data_scitype(::Type{<:MyTransformer}) = Union{
+    Tuple{Table(Continuous)},
+	Tuple{Table(Continous), AbstractVector{<:Finite}},
+}
+```
+
+