Merge pull request #701 from JuliaAI/dev

For a 0.19 release

.github/workflows/ci.yml

@@ -4,6 +4,7 @@ on:
     branches:
       - master
       - dev
+      - for-0-point-19-release
   push:
     branches:
       - master

Project.toml

@@ -1,10 +1,11 @@
 name = "MLJBase"
 uuid = "a7f614a8-145f-11e9-1d2a-a57a1082229d"
 authors = ["Anthony D. Blaom <[email protected]>"]
-version = "0.18.26"
+version = "0.19"
 
 [deps]
 CategoricalArrays = "324d7699-5711-5eae-9e2f-1d82baa6b597"
+CategoricalDistributions = "af321ab8-2d2e-40a6-b165-3d674595d28e"
 ComputationalResources = "ed09eef8-17a6-5b46-8889-db040fac31e3"
 Dates = "ade2ca70-3891-5945-98fb-dc099432e06a"
 DelimitedFiles = "8bb1440f-4735-579b-a4ab-409b98df4dab"
@@ -29,6 +30,7 @@ Tables = "bd369af6-aec1-5ad0-b16a-f7cc5008161c"
 
 [compat]
 CategoricalArrays = "0.9, 0.10"
+CategoricalDistributions = "0.1"
 ComputationalResources = "0.3"
 Distributions = "0.25.3"
 InvertedIndices = "1"
@@ -39,8 +41,8 @@ OrderedCollections = "1.1"
 Parameters = "0.12"
 PrettyTables = "1"
 ProgressMeter = "1.7.1"
-ScientificTypes = "2.1"
-StatisticalTraits = "2.1"
+ScientificTypes = "3"
+StatisticalTraits = "3"
 StatsBase = "0.32, 0.33"
 Tables = "0.2, 1.0"
 julia = "1.3"

README.md

@@ -38,11 +38,6 @@ repository provides core functionality for MLJ, including:
 
 - basic utilities for **manipulating data**
 
-- an extension to
-  [Distributions.jl](https://github.com/JuliaStats/Distributions.jl)
-  called `UnivariateFinite` for randomly sampling *labeled*
-  categorical data
-
 - a [small interface](https://alan-turing-institute.github.io/MLJ.jl/dev/evaluating_model_performance/#Custom-resampling-strategies-1) for **resampling strategies** and implementations, including `CV()`, `StratifiedCV` and `Holdout`
 
 - methods for **performance evaluation**, based on those resampling strategies
@@ -59,9 +54,3 @@ repository provides core functionality for MLJ, including:
   library, user-defined measures, as well as about forty natively
   defined measures.
 
-
-Previously MLJBase provided the model interface for integrating third
-party machine learning models into MLJ. That role has now shifted to
-the light-weight
-[MLJModelInterface](https://github.com/JuliaAI/MLJModelInterface.jl)
-package.

src/MLJBase.jl

@@ -7,21 +7,46 @@ import Base: ==, precision, getindex, setindex!
 import Base.+, Base.*
 
 # Scitype
-import ScientificTypes.ScientificTypesBase: TRAIT_FUNCTION_GIVEN_NAME
 using ScientificTypes
 
 # Traits for models and measures (which are being overloaded):
 using StatisticalTraits
+
 for trait in StatisticalTraits.TRAITS
     eval(:(import StatisticalTraits.$trait))
 end
+
 import Base.instances # considered a trait for measures
 import StatisticalTraits.snakecase
+import StatisticalTraits.info
 
 # Interface
-using MLJModelInterface
+
+# HACK: When https://github.com/JuliaAI/MLJModelInterface.jl/issues/124
+# is resolved:
+# Uncomment next line and delete "Hack Block"
+# using MLJModelInterface
+#####################
+# Hack Block begins #
+#####################
+exported_names(m::Module) =
+    filter!(x -> Base.isexported(m, x),
+            Base.names(m; all=true, imported=true))
+import MLJModelInterface
+for name in exported_names(MLJModelInterface)
+    name in [:UnivariateFinite,
+             :augmented_transform,
+             :info] && continue
+    quote
+        import MLJModelInterface.$name
+    end |> eval
+end
+###################
+# Hack Block ends #
+###################
+
 import MLJModelInterface: fit, update, update_data, transform,
-    inverse_transform, fitted_params, predict, predict_mode, 
+    inverse_transform, fitted_params, predict, predict_mode,
     predict_mean, predict_median, predict_joint,
     evaluate, clean!, is_same_except,
     save, restore, is_same_except, istransparent,
@@ -55,13 +80,13 @@ import StatsBase: fit!, mode, countmap
 import Missings: levels
 using Missings
 import Distributions
+using CategoricalDistributions
 import Distributions: pdf, logpdf, sampler
 const Dist = Distributions
 
 # from Standard Library:
 using Statistics, LinearAlgebra, Random, InteractiveUtils
 
-
 # ===================================================================
 ## EXPORTS
 
@@ -73,9 +98,11 @@ using Statistics, LinearAlgebra, Random, InteractiveUtils
 
 # traits for measures and models:
 using StatisticalTraits
+
 for trait in StatisticalTraits.TRAITS
     eval(:(export $trait))
 end
+
 export implemented_methods # defined here and not in StatisticalTraits
 
 export UnivariateFinite
@@ -88,7 +115,7 @@ export @mlj_model, metadata_pkg, metadata_model
 
 # model api
 export fit, update, update_data, transform, inverse_transform,
-    fitted_params, predict, predict_mode, predict_mean, 
+    fitted_params, predict, predict_mode, predict_mean,
     predict_median, predict_joint,
     evaluate, clean!, training_losses
 
@@ -103,9 +130,14 @@ export CPU1, CPUProcesses, CPUThreads
 export Unknown, Known, Finite, Infinite,
     OrderedFactor, Multiclass, Count, Continuous, Textual,
     Binary, ColorImage, GrayImage, Image, Table
-export scitype, scitype_union, elscitype, nonmissing, trait
+
+export scitype, scitype_union, elscitype, nonmissing
 export coerce, coerce!, autotype, schema, info
 
+# re-exports from CategoricalDistributions:
+export CategoricalUnivariateFiniteArray, UnivariateFiniteVector
+
+
 # -------------------------------------------------------------------
 # exports from this module, MLJBase
 
@@ -123,19 +155,16 @@ export partition, unpack, complement, restrict, corestrict
 
 # utilities.jl:
 export flat_values, recursive_setproperty!,
-       recursive_getproperty, pretty, unwind
+    recursive_getproperty, pretty, unwind
 
 # show.jl
 export HANDLE_GIVEN_ID, @more, @constant, @bind, color_on, color_off
 
-# univariate_finite/
-export average, UnivariateFiniteArray, UnivariateFiniteVector
-
 # datasets.jl:
 export load_boston, load_ames, load_iris, load_sunspots,
-       load_reduced_ames, load_crabs, load_smarket,
-       @load_boston, @load_ames, @load_iris, @load_sunspots,
-       @load_reduced_ames, @load_crabs, @load_smarket
+    load_reduced_ames, load_crabs, load_smarket,
+    @load_boston, @load_ames, @load_iris, @load_sunspots,
+    @load_reduced_ames, @load_crabs, @load_smarket
 
 # sources.jl:
 export source, Source, CallableReturning
@@ -150,14 +179,18 @@ export make_blobs, make_moons, make_circles, make_regression
 export machines, sources, @from_network, @pipeline, Stack,
     glb, @tuple, node, @node, sources, origins, return!,
     nrows_at_source, machine, rebind!, nodes, freeze!, thaw!,
-    Node, AbstractNode
+    Node, AbstractNode, Pipeline,
+    ProbabilisticPipeline, DeterministicPipeline, UnsupervisedPipeline,
+    StaticPipeline, IntervalPipeline
+
+export TransformedTargetModel
 
 # aliases to the above,  kept for backwards compatibility:
 export  DeterministicNetwork, ProbabilisticNetwork, UnsupervisedNetwork
 
 # resampling.jl:
 export ResamplingStrategy, Holdout, CV, StratifiedCV, TimeSeriesCV,
-       evaluate!, Resampler, PerformanceEvaluation
+    evaluate!, Resampler, PerformanceEvaluation
 
 # -------------------------------------------------------------------
 # exports from MLJBase specific to Measure (these may go in their
@@ -223,33 +256,33 @@ export TruePositive, TrueNegative, FalsePositive, FalseNegative,
 # measures/finite.jl -- Finite{N} - multiclass generalizations of
 # above OrderedFactor{2} measures (but order independent):
 export MulticlassTruePositive, MulticlassTrueNegative, MulticlassFalsePositive,
-      MulticlassFalseNegative, MulticlassTruePositiveRate,
-      MulticlassTrueNegativeRate, MulticlassFalsePositiveRate,
-      MulticlassFalseNegativeRate, MulticlassFalseDiscoveryRate,
-      MulticlassPrecision, MulticlassNegativePredictiveValue, MulticlassFScore,
-      # standard synonyms
-      MTPR, MTNR, MFPR, MFNR, MFDR, MPPV,
-      MulticlassRecall, MulticlassSpecificity,
-      # instances and their synonyms
-      multiclass_truepositive, multiclass_truenegative,
-      multiclass_falsepositive,
-      multiclass_falsenegative, multiclass_true_positive,
-      multiclass_true_negative, multiclass_false_positive,
-      multiclass_false_negative, multiclass_truepositive_rate,
-      multiclass_truenegative_rate, multiclass_falsepositive_rate,
-      multiclass_true_positive_rate, multiclass_true_negative_rate,
-      multiclass_false_positive_rate, multiclass_falsenegative_rate,
-      multiclass_negativepredictive_value, multiclass_false_negative_rate,
-      multiclass_negative_predictive_value, multiclass_positivepredictive_value,
-      multiclass_positive_predictive_value, multiclass_tpr, multiclass_tnr,
-      multiclass_fpr, multiclass_fnr, multiclass_falsediscovery_rate,
-      multiclass_false_discovery_rate, multiclass_fdr, multiclass_npv,
-      multiclass_ppv, multiclass_recall, multiclass_sensitivity,
-      multiclass_hit_rate, multiclass_miss_rate, multiclass_specificity,
-      multiclass_selectivity, macro_f1score, micro_f1score,
-      multiclass_f1score, multiclass_fallout, multiclass_precision,
-      # averaging modes
-      no_avg, macro_avg, micro_avg
+    MulticlassFalseNegative, MulticlassTruePositiveRate,
+    MulticlassTrueNegativeRate, MulticlassFalsePositiveRate,
+    MulticlassFalseNegativeRate, MulticlassFalseDiscoveryRate,
+    MulticlassPrecision, MulticlassNegativePredictiveValue, MulticlassFScore,
+    # standard synonyms
+    MTPR, MTNR, MFPR, MFNR, MFDR, MPPV,
+    MulticlassRecall, MulticlassSpecificity,
+    # instances and their synonyms
+    multiclass_truepositive, multiclass_truenegative,
+    multiclass_falsepositive,
+    multiclass_falsenegative, multiclass_true_positive,
+    multiclass_true_negative, multiclass_false_positive,
+    multiclass_false_negative, multiclass_truepositive_rate,
+    multiclass_truenegative_rate, multiclass_falsepositive_rate,
+    multiclass_true_positive_rate, multiclass_true_negative_rate,
+    multiclass_false_positive_rate, multiclass_falsenegative_rate,
+    multiclass_negativepredictive_value, multiclass_false_negative_rate,
+    multiclass_negative_predictive_value, multiclass_positivepredictive_value,
+    multiclass_positive_predictive_value, multiclass_tpr, multiclass_tnr,
+    multiclass_fpr, multiclass_fnr, multiclass_falsediscovery_rate,
+    multiclass_false_discovery_rate, multiclass_fdr, multiclass_npv,
+    multiclass_ppv, multiclass_recall, multiclass_sensitivity,
+    multiclass_hit_rate, multiclass_miss_rate, multiclass_specificity,
+    multiclass_selectivity, macro_f1score, micro_f1score,
+    multiclass_f1score, multiclass_fallout, multiclass_precision,
+    # averaging modes
+    no_avg, macro_avg, micro_avg
 
 # measures/loss_functions_interface.jl
 export dwd_margin_loss, exp_loss, l1_hinge_loss, l2_hinge_loss, l2_margin_loss,
@@ -273,6 +306,7 @@ const MARGIN_LOSSES = [
     :SmoothedL1HingeLoss,
     :ZeroOneLoss
 ]
+
 const DISTANCE_LOSSES = [
     :HuberLoss,
     :L1EpsilonInsLoss,
@@ -282,6 +316,7 @@ const DISTANCE_LOSSES = [
     :PeriodicLoss,
     :QuantileLoss
 ]
+
 const WITH_PARAMETERS = [
     :DWDMarginLoss,
     :SmoothedL1HingeLoss,
@@ -291,6 +326,7 @@ const WITH_PARAMETERS = [
     :LPDistLoss,
     :QuantileLoss,
 ]
+
 const LOSS_FUNCTIONS = vcat(MARGIN_LOSSES, DISTANCE_LOSSES)
 
 for Loss in LOSS_FUNCTIONS
@@ -301,12 +337,20 @@ end
 # re-export from Random, StatsBase, Statistics, Distributions,
 # OrderedCollections, CategoricalArrays, InvertedIndices:
 export pdf, sampler, mode, median, mean, shuffle!, categorical, shuffle,
-       levels, levels!, std, Not, support, logpdf, LittleDict
+   levels, levels!, std, Not, support, logpdf, LittleDict
 
 
 # ===================================================================
 ## CONSTANTS
 
+const PREDICT_OPERATIONS = (:predict,
+                            :predict_mean,
+                            :predict_mode,
+                            :predict_median,
+                            :predict_joint)
+
+const OPERATIONS = (PREDICT_OPERATIONS..., :transform, :inverse_transform)
+
 # the directory containing this file: (.../src/)
 const MODULE_DIR = dirname(@__FILE__)
 
@@ -345,26 +389,20 @@ include("show.jl")
 include("interface/data_utils.jl")
 include("interface/model_api.jl")
 
-include("univariate_finite/types.jl")
-include("univariate_finite/methods.jl")
-include("univariate_finite/arrays.jl")
-
 include("sources.jl")
 include("machines.jl")
 
 include("composition/abstract_types.jl")
 include("composition/learning_networks/nodes.jl")
 include("composition/learning_networks/inspection.jl")
 include("composition/learning_networks/machines.jl")
-@static if VERSION ≥ v"1.3.0-"
-    include("composition/learning_networks/arrows.jl")
-end
 
 include("composition/models/methods.jl")
 include("composition/models/from_network.jl")
 include("composition/models/inspection.jl")
+include("composition/models/deprecated.jl")
 include("composition/models/pipelines.jl")
-include("composition/models/_wrapped_function.jl")
+include("composition/models/transformed_target_model.jl")
 
 include("operations.jl")
 include("resampling.jl")