Merge pull request #11 from alan-turing-institute/dev

ablaom · web-flow · commit 7adad1a0c6ac · 2020-02-12T21:56:12.000+13:00
Update to MLJBase 0.11.1
diff --git a/Project.toml b/Project.toml
@@ -1,7 +1,7 @@
 name = "MLJTuning"
 uuid = "03970b2e-30c4-11ea-3135-d1576263f10f"
 authors = ["Anthony D. Blaom <anthony.blaom@gmail.com>"]
-version = "0.1.2"
+version = "0.1.3"
 
 [deps]
 ComputationalResources = "ed09eef8-17a6-5b46-8889-db040fac31e3"
@@ -11,17 +11,18 @@ Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
 RecipesBase = "3cdcf5f2-1ef4-517c-9805-6587b60abb01"
 
 [compat]
-julia = "^1"
-MLJBase = "^0.10.1"
 ComputationalResources = "^0.3"
-RecipesBase = "^0.7"
+MLJBase = "^0.11.1"
+RecipesBase = "^0.8"
+julia = "^1"
 
 [extras]
 CategoricalArrays = "324d7699-5711-5eae-9e2f-1d82baa6b597"
 DecisionTree = "7806a523-6efd-50cb-b5f6-3fa6f1930dbb"
 Distances = "b4f34e82-e78d-54a5-968a-f98e89d6e8f7"
 Distributions = "31c24e10-a181-5473-b8eb-7969acd0382f"
 LinearAlgebra = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"
+MLJModelInterface = "e80e1ace-859a-464e-9ed9-23947d8ae3ea"
 MultivariateStats = "6f286f6a-111f-5878-ab1e-185364afe411"
 NearestNeighbors = "b8a86587-4115-5ab1-83bc-aa920d37bbce"
 ScientificTypes = "321657f4-b219-11e9-178b-2701a2544e81"
@@ -31,4 +32,4 @@ Tables = "bd369af6-aec1-5ad0-b16a-f7cc5008161c"
 Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"
 
 [targets]
-test = ["CategoricalArrays", "DecisionTree", "Distances", "Distributions", "LinearAlgebra", "MultivariateStats", "NearestNeighbors", "ScientificTypes", "Statistics", "StatsBase", "Tables", "Test"]
+test = ["CategoricalArrays", "DecisionTree", "Distances", "Distributions", "LinearAlgebra", "MLJModelInterface", "MultivariateStats", "NearestNeighbors", "ScientificTypes", "Statistics", "StatsBase", "Tables", "Test"]
diff --git a/README.md b/README.md
@@ -10,7 +10,7 @@ learning models.
 ### Contents
 
  - [Who is this repo for?](#who-is-this-repo-for)
- - [What's provided here?](#what's-provided-here)
+ - [What's provided here?](#what-is-provided-here)
  - [How do I implement a new tuning strategy?](#How-do-I-implement-a-new-tuning-strategy)
 
 *Note:* This component of the [MLJ
@@ -46,7 +46,7 @@ MLJTuning is a component of the [MLJ
   there.
 
 
-## What's provided here?
+## What is provided here?
 
 This repository contains:
 
diff --git a/src/strategies/explicit.jl b/src/strategies/explicit.jl
@@ -1,4 +1,4 @@
-mutable struct Explicit <: TuningStrategy end
+mutable struct Explicit <: TuningStrategy end 
 
 # models! returns all available models in the range at once:
 MLJTuning.models!(tuning::Explicit, model, history::Nothing,
diff --git a/test/models.jl b/test/models.jl
@@ -7,6 +7,8 @@
 module Models
 
 using MLJBase
+import MLJModelInterface: @mlj_model, metadata_model, metadata_pkg
+import MLJModelInterface
 
 include("models/Constant.jl")
 include("models/DecisionTree.jl")
diff --git a/test/models/Constant.jl b/test/models/Constant.jl
@@ -5,7 +5,6 @@ export ConstantClassifier, ConstantRegressor,
     ProbabilisticConstantClassifer
 
 import MLJBase
-import MLJBase: metadata_pkg, metadata_model
 import Distributions
 
 """
diff --git a/test/models/DecisionTree.jl b/test/models/DecisionTree.jl
@@ -1,8 +1,5 @@
 export DecisionTreeClassifier, DecisionTreeRegressor
 
-import MLJBase
-import MLJBase: @mlj_model, metadata_pkg, metadata_model
-
 using ScientificTypes
 
 using CategoricalArrays
diff --git a/test/models/MultivariateStats.jl b/test/models/MultivariateStats.jl
@@ -1,7 +1,5 @@
 export RidgeRegressor, PCA
 
-import MLJBase
-import MLJBase: @mlj_model, metadata_model, metadata_pkg
 # using Distances
 using LinearAlgebra
 using Tables, ScientificTypes
diff --git a/test/models/NearestNeighbors.jl b/test/models/NearestNeighbors.jl
@@ -1,6 +1,5 @@
 export KNNRegressor, KNNClassifier
 
-import MLJBase: @mlj_model, metadata_model, metadata_pkg
 using Distances
 
 import NearestNeighbors
diff --git a/test/models/Transformers.jl b/test/models/Transformers.jl
@@ -5,6 +5,8 @@ export FeatureSelector,
     UnivariateBoxCoxTransformer,
     OneHotEncoder, UnivariateDiscretizer
 
+import MLJModelInterface: @mlj_model, metadata_model, metadata_pkg
+
 using Statistics
 
 const N_VALUES_THRESH = 16 # for BoxCoxTransformation
@@ -158,15 +160,15 @@ function MLJBase.fit(transformer::UnivariateDiscretizer, verbosity::Int,X)
     return res, cache, report
 end
 
+
 # acts on scalars:
 function transform_to_int(
-    result::UnivariateDiscretizerResult{<:CategoricalElement{R}},
-    r::Real) where R
-
-    k = R(1)
-    for q in result.odd_quantiles
+            result::UnivariateDiscretizerResult{<:CategoricalValue{R}},
+            r::Real) where R
+    k = oneR = R(1)
+    @inbounds for q in result.odd_quantiles
         if r > q
-            k += R(1)
+            k += oneR
         end
     end
     return k

Original file line number	Diff line number	Diff line change
`@@ -1,4 +1,4 @@`
`1`		`-mutable struct Explicit <: TuningStrategy end`
	`1`	`+mutable struct Explicit <: TuningStrategy end`
`2`	`2`
`3`	`3`	`# models! returns all available models in the range at once:`
`4`	`4`	`MLJTuning.models!(tuning::Explicit, model, history::Nothing,`