Merge pull request #290 from alan-turing-institute/dev

For a 0.13.5 release

Project.toml

@@ -1,7 +1,7 @@
 name = "MLJBase"
 uuid = "a7f614a8-145f-11e9-1d2a-a57a1082229d"
 authors = ["Anthony D. Blaom <[email protected]>"]
-version = "0.13.4"
+version = "0.13.5"
 
 [deps]
 CategoricalArrays = "324d7699-5711-5eae-9e2f-1d82baa6b597"

src/MLJBase.jl

@@ -96,8 +96,7 @@ export coerce, coerce!, autotype, schema, info
 # -------------------------------------------------------------------
 # exports from MLJBase
 
-export DeterministicNetwork, ProbabilisticNetwork, UnsupervisedNetwork,
-       best, @load
+export DeterministicNetwork, ProbabilisticNetwork, UnsupervisedNetwork, @load
 
 # computational_resources.jl:
 export default_resource
@@ -109,8 +108,7 @@ export ParamRange, NumericRange, NominalRange, iterator, scale
 export params # note this is *not* an extension of StatsBase.params
 
 # data.jl:
-export reconstruct, levels_seen, container_type,
-       partition, unpack, complement, restrict, corestrict
+export partition, unpack, complement, restrict, corestrict
 
 # utilities.jl:
 export @set_defaults, flat_values, recursive_setproperty!,
@@ -122,10 +120,6 @@ export HANDLE_GIVEN_ID, @more, @constant, color_on, color_off
 # distributions.jl:
 export average
 
-# tasks.jl:
-export SupervisedTask, UnsupervisedTask, MLJTask,
-       X_and_y, X_, y_, nfeatures
-
 # info_dict.jl:
 export info_dict
 

src/data/data.jl

@@ -28,10 +28,20 @@ function _partition(rows, fractions, ::Nothing)
     return tuple((rows[h:t] for (h, t) in zip(heads, tails))...)
 end
 
+_make_numerical(v::AbstractVector) =
+    throw(ArgumentError("`stratify` must have `Count`, `Continuous` "*
+                        "or `Finite` element scitpye. Consider "*
+                        "`coerce(stratify, Finite)`. "))
+_make_numerical(v::AbstractVector{<:Union{Missing,Real}}) = v
+_make_numerical(v::AbstractVector{<:Union{Missing,CategoricalElement}}) =
+                                int.(v)
+
 # Helper function for partitioning in the stratified case
-function _partition(rows, fractions, stratify::AbstractVector)
+function _partition(rows, fractions, raw_stratify::AbstractVector)
+    stratify = _make_numerical(raw_stratify)
     length(stratify) == length(rows) ||
-        throw(ArgumentError("The stratification vector must have as many entries as " *
+        throw(ArgumentError("The stratification vector must "*
+                            "have as many entries as " *
                             "the rows to partition."))
     uv    = unique(stratify)
     # construct table (n_classes * idx_of_that_class)
@@ -72,7 +82,7 @@ end
 
 """
     partition(rows::AbstractVector{Int}, fractions...;
-              shuffle=nothing, rng=Random.GLOBAL_RNG)
+              shuffle=nothing, rng=Random.GLOBAL_RNG, stratify=nothing)
 
 Splits the vector `rows` into a tuple of vectors whose lengths are
 given by the corresponding `fractions` of `length(rows)` where valid

src/resampling.jl

@@ -630,7 +630,7 @@ function _evaluate!(func, machines, ::CPUThreads, nfolds,verbosity)
 
      @sync begin  
 
-        @sync for parts in Iterators.partition(1:nfolds, max(1,floor(Int, nfolds/n_threads)))    
+        @sync for parts in Iterators.partition(1:nfolds, max(1,cld(nfolds, n_threads)))    
         Threads.@spawn begin
             for k in parts
             id = Threads.threadid()

test/data/data.jl

@@ -44,7 +44,11 @@ import MLJBase: decoder, int, classes, partition, unpack, selectcols, matrix,
     y = ones(Int, 1000)
     y[end-100:end] .= 0; # 90%
 
+    train1, test1 =
+        partition(eachindex(y), 0.8, stratify=categorical(y), rng=34)
     train, test = partition(eachindex(y), 0.8, stratify=y, rng=34)
+    @test train == train1
+    @test test == test1
     @test isapprox(sum(y[train])/length(train), 0.9, rtol=1e-2)
     @test isapprox(sum(y[test])/length(test), 0.9, rtol=1e-2)