updated to MLJ 0.14

Project.toml

@@ -1,7 +1,7 @@
 name = "JLBoostMLJ"
 uuid = "8b86df2c-1bc3-481d-95df-1c4d5a20ed95"
 authors = ["Dai ZJ <[email protected]>"]
-version = "0.1.8"
+version = "0.1.9"
 
 [deps]
 DataFrames = "a93c6f00-e57d-5684-b7b6-d8193f3e46c0"
@@ -15,7 +15,7 @@ ScientificTypes = "321657f4-b219-11e9-178b-2701a2544e81"
 DataFrames = "0.21"
 JLBoost = "^0.1.8"
 LossFunctions = "0.5, 0.6"
-MLJ = "0.10, 0.11, 0.12, 0.13"
+MLJ = "0.10, 0.11, 0.12, 0.13, 0.14"
 MLJBase = "0.12, 0.13, 0.14, 0.15"
 ScientificTypes = "0.7, 0.8, 1.0"
 julia = "1"

README.md

@@ -21,15 +21,15 @@ model = JLBoostClassifier()
 
 ````
 JLBoostClassifier(
-    loss = LogitLogLoss(),
+    loss = JLBoost.LogitLogLoss(),
     nrounds = 1,
     subsample = 1.0,
     eta = 1.0,
     max_depth = 6,
     min_child_weight = 1.0,
     lambda = 0.0,
     gamma = 0.0,
-    colsample_bytree = 1) @810
+    colsample_bytree = 1) @087
 ````
 
 
@@ -47,11 +47,11 @@ mljmachine  = machine(model, X, y)
 
 
 ````
-Machine{JLBoostClassifier} @772 trained 0 times.
+Machine{JLBoostClassifier} @730 trained 0 times.
   args: 
-    1:	Source @097 ⏎ `ScientificTypes.Table{AbstractArray{ScientificTypes.C
+    1:	Source @910 ⏎ `ScientificTypes.Table{AbstractArray{ScientificTypes.C
 ontinuous,1}}`
-    2:	Source @077 ⏎ `AbstractArray{ScientificTypes.Count,1}`
+    2:	Source @954 ⏎ `AbstractArray{ScientificTypes.Count,1}`
 ````
 
 
@@ -81,11 +81,11 @@ Choosing a split on SepalLength
 Choosing a split on SepalWidth
 Choosing a split on PetalLength
 Choosing a split on PetalWidth
-Machine{JLBoostClassifier} @772 trained 1 time.
+Machine{JLBoostClassifier} @730 trained 1 time.
   args: 
-    1:	Source @097 ⏎ `ScientificTypes.Table{AbstractArray{ScientificTypes.C
+    1:	Source @910 ⏎ `ScientificTypes.Table{AbstractArray{ScientificTypes.C
 ontinuous,1}}`
-    2:	Source @077 ⏎ `AbstractArray{ScientificTypes.Count,1}`
+    2:	Source @954 ⏎ `AbstractArray{ScientificTypes.Count,1}`
 ````
 
 
@@ -216,11 +216,11 @@ m = machine(tm, X, y_cate)
 
 
 ````
-Machine{ProbabilisticTunedModel{Grid,…}} @388 trained 0 times.
+Machine{ProbabilisticTunedModel{Grid,…}} @109 trained 0 times.
   args: 
-    1:	Source @578 ⏎ `ScientificTypes.Table{AbstractArray{ScientificTypes.C
+    1:	Source @664 ⏎ `ScientificTypes.Table{AbstractArray{ScientificTypes.C
 ontinuous,1}}`
-    2:	Source @226 ⏎ `AbstractArray{ScientificTypes.Multiclass{2},1}`
+    2:	Source @788 ⏎ `AbstractArray{ScientificTypes.Multiclass{2},1}`
 ````
 
 
@@ -235,11 +235,11 @@ fit!(m)
 
 
 ````
-Machine{ProbabilisticTunedModel{Grid,…}} @388 trained 1 time.
+Machine{ProbabilisticTunedModel{Grid,…}} @109 trained 1 time.
   args: 
-    1:	Source @578 ⏎ `ScientificTypes.Table{AbstractArray{ScientificTypes.C
+    1:	Source @664 ⏎ `ScientificTypes.Table{AbstractArray{ScientificTypes.C
 ontinuous,1}}`
-    2:	Source @226 ⏎ `AbstractArray{ScientificTypes.Multiclass{2},1}`
+    2:	Source @788 ⏎ `AbstractArray{ScientificTypes.Multiclass{2},1}`
 ````
 
 
@@ -287,15 +287,15 @@ Choosing a split on SepalLength
 Choosing a split on SepalWidth
 Choosing a split on PetalLength
 Choosing a split on PetalWidth
-(fitresult = (treemodel = JLBoostTreeModel(AbstractJLBoostTree[eta = 1.0 (t
-ree weight)
+(fitresult = (treemodel = JLBoost.JLBoostTrees.JLBoostTreeModel(JLBoost.JLB
+oostTrees.AbstractJLBoostTree[eta = 1.0 (tree weight)
 
    -- PetalLength <= 1.9
      ---- weight = 2.0
 
    -- PetalLength > 1.9
      ---- weight = -2.0
-], LogitLogLoss(), :__y__),
+], JLBoost.LogitLogLoss(), :__y__),
               target_levels = Bool[0, 1],),
  cache = nothing,
  report = (AUC = 0.16666666666666669,

src/mlj.jl

@@ -1,14 +1,14 @@
 export fit, predict, fitted_params, JLBoostMLJModel, JLBoostClassifier, JLBoostRegressor, JLBoostCount
 
-#using MLJBase
+import MLJBase
 import MLJBase: Probabilistic, Deterministic, clean!, fit, predict, fitted_params, load_path, Table
 import MLJBase: package_name, package_uuid, package_url, is_pure_julia, package_license
 import MLJBase: input_scitype, target_scitype, docstring, UnivariateFinite
 
 using ScientificTypes: Continuous, OrderedFactor, Count, Multiclass, Finite
 
 using LossFunctions: PoissonLoss, L2DistLoss
-using JLBoost: LogitLogLoss, jlboost, AUC, gini, feature_importance, predict
+using JLBoost: LogitLogLoss, jlboost, AUC, gini, feature_importance
 
 using DataFrames: DataFrame, nrow, levels, categorical
 
@@ -194,15 +194,15 @@ fitted_params(model::JLBoostMLJModel, fitresult) = (fitresult = fitresult.treemo
 
 
 #  seehttps://alan-turing-institute.github.io/MLJ.jl/stable/adding_models_for_general_use/#The-predict-method-1
-predict(model::JLBoostClassifier, fitresult, Xnew) = begin
+function MLJBase.predict(model::JLBoostClassifier, fitresult, Xnew)
     res = JLBoost.predict(fitresult.treemodel, Xnew)
     p = 1 ./ (1 .+ exp.(-res))
     levels_cate = categorical(fitresult.target_levels)
     [UnivariateFinite(levels_cate, [p, 1-p]) for p in p]
 end
 
 
-predict(model::JLBoostMLJModel, fitresult, Xnew) = begin
+function MLJBase.predict(model::JLBoostMLJModel, fitresult, Xnew)
     JLBoost.predict(fitresult, Xnew)
 end