Merge pull request #192 from alan-turing-institute/dev

Add EvoTrees. For 0.8.1 release

Author: ablaom

Project.toml

@@ -1,7 +1,7 @@
 name = "MLJModels"
 uuid = "d491faf4-2d78-11e9-2867-c94bc002c0b7"
 authors = ["Anthony D. Blaom <[email protected]>"]
-version = "0.8.0"
+version = "0.8.1"
 
 [deps]
 CategoricalArrays = "324d7699-5711-5eae-9e2f-1d82baa6b597"

README.md

@@ -6,6 +6,10 @@ Repository of selected
 [MLJ](https://github.com/alan-turing-institute/MLJ.jl) model
 interfaces, and home of the MLJ model registry.
 
+For instructions on integrating a new model with MLJ visit
+[here](https://alan-turing-institute.github.io/MLJ.jl/dev/adding_models_for_general_use/)
+
+
 ### Contents
 
  - [Who is this repo for?](#who-is-this-repo-for)
@@ -64,25 +68,30 @@ MLJModels contains:
 ## Instructions for updating the MLJ model registry
 
 Generally model registration is performed by administrators. If you
-have an interface you would like registered, open an issue on this repo. 
+have an interface you would like registered, open an issue
+[here](https://github.com/alan-turing-institute/MLJ.jl/issues).
 
 To register all the models in GreatNewPackage with MLJ:
 
 - In the dev branch of a clone of the dev branch of MLJModels, change
   to the `/src/registry/` directory and, in Julia, activate the
   environment specified by the Project.toml there, after checking the
-  [compat] conditions there are up to date.
+  [compat] conditions there are up to date. **Do not use** `Revise`.
 
 - Add `GreatNewPackage` to the environment.
 
-- In some environment in which your MLJModels clone has been added
-  using `Pkg.dev`, execute `using MLJModels; @update`. This updates
+- In some environment to which your MLJModels clone has been added
+  (using `Pkg.dev`) execute `using MLJModels; @update`. This updates
   `src/registry/Metadata.toml` and `src/registry/Models.toml` (the
   latter is generated for convenience and not used by MLJ).
 
 - Test that interfaces load with `MLJModels.check_registry()`
 
--  Quit your REPL session, whose namespace is now polluted.
+- Quit your REPL session, whose namespace is now polluted.
+
+- *Note.* that your local MLJModels will not immediately adopt the
+  updated registry because that requires pre-compilation; for
+  technical reasons the registry is not loaded in `__init__`()`.
 
 - Push your changes to an appropriate branch of MLJModels to make
   the updated metadata available to users of the next MLJModels tagged

src/model_search.jl

@@ -185,7 +185,7 @@ List all models whose names are in the namespace of the specified
 module `modl`, or meeting the `conditions`, if specified. Here a
 *condition* is a `Bool`-valued function on models.
 
-See also [models](@ref)
+See also [`models`](@ref)
 
 """
 function localmodels(args...; modl=Main)

src/registry/Metadata.toml

@@ -7,6 +7,7 @@ MultivariateStats = ["ICA", "BayesianLDA", "BayesianSubspaceLDA", "RidgeRegresso
 DecisionTree = ["AdaBoostStumpClassifier", "DecisionTreeRegressor", "DecisionTreeClassifier", "RandomForestRegressor", "RandomForestClassifier"]
 Clustering = ["KMeans", "KMedoids"]
 XGBoost = ["XGBoostCount", "XGBoostRegressor", "XGBoostClassifier"]
+EvoTrees = ["EvoTreeClassifier", "EvoTreeGaussian", "EvoTreeRegressor", "EvoTreeCount"]
 MLJModels = ["ConstantClassifier", "Standardizer", "DeterministicConstantClassifier", "OneHotEncoder", "UnivariateBoxCoxTransformer", "ConstantRegressor", "FeatureSelector", "UnivariateDiscretizer", "FillImputer", "DeterministicConstantRegressor", "UnivariateStandardizer"]
 LIBSVM = ["EpsilonSVR", "LinearSVC", "NuSVR", "NuSVC", "SVC", "OneClassSVM"]
 GLM = ["LinearBinaryClassifier", "LinearCountRegressor", "LinearRegressor"]

src/registry/Models.toml

@@ -1,6 +1,7 @@
 [deps]
 Clustering = "aaaa29a8-35af-508c-8bc3-b662a17a0fe5"
 DecisionTree = "7806a523-6efd-50cb-b5f6-3fa6f1930dbb"
+EvoTrees = "f6006082-12f8-11e9-0c9c-0d5d367ab1e5"
 GLM = "38e38edf-8417-5370-95a0-9cbb8c7f171a"
 InteractiveUtils = "b77e0a4c-d291-57a0-90e8-8db25a27a240"
 LIBSVM = "b1bec4e5-fd48-53fe-b0cb-9723c09d164b"
@@ -17,7 +18,7 @@ XGBoost = "009559a3-9522-5dbb-924b-0b6ed2b22bb9"
 
 [compat]
 MLJBase = "^0.11"
-MLJModelInterface = "^0.1"
+MLJModelInterface = "^0.1.7"
 
 [extras]
 Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"

src/registry/Project.toml

@@ -125,8 +125,9 @@ function _update(mod, test_env_only)
     mod.eval(program1)
     test_env_only || mod.eval(program2)
 
-    println("If you have called @update from the REPL then your namespace "*
-            "is now polluted. Restart your REPL. ")
+    println("You can now check the registry by running "*
+            "`MLJModels.check_registry().\n\n"*
+            "You can safely ignore \"conflicting import\" warnings. ")
 
     true
 end

src/registry/src/Registry.jl

@@ -20,3 +20,4 @@ function check_registry()
         end
     end
 end
+

src/registry/src/check_registry.jl

@@ -4,6 +4,8 @@ using Test
 import CategoricalArrays
 import CategoricalArrays.categorical
 using MLJBase
+using Random
+Random.seed!(1234)
 
 # load code to be tested:
 import MLJModels