Remove MLJ Model Registry tools

.github/workflows/ci.yml

@@ -41,6 +41,9 @@ jobs:
             ${{ runner.os }}-
       - uses: julia-actions/julia-buildpkg@v1
       - uses: julia-actions/julia-runtest@v1
+        env:
+          # This environment variable enables the integration tests:
+          MLJ_TEST_INTEGRATION: '1'
       - uses: julia-actions/julia-processcoverage@v1
       - uses: codecov/codecov-action@v4
         with:

Project.toml

@@ -31,33 +31,37 @@ Tables = "bd369af6-aec1-5ad0-b16a-f7cc5008161c"
 CategoricalArrays = "0.9, 0.10"
 CategoricalDistributions = "0.1"
 Combinatorics = "1.0"
-Dates = "<0.0.1, 1"
+Dates = "1"
 Distances = "0.9,0.10"
+Distributed = "1"
 Distributions = "0.25"
-InteractiveUtils = "<0.0.1, 1"
-LinearAlgebra = "<0.0.1, 1"
-Markdown = "<0.0.1, 1"
+InteractiveUtils = "1"
+LinearAlgebra = "1"
 MLJModelInterface = "1.10"
+Markdown = "1"
 OrderedCollections = "1.1"
 Parameters = "0.12"
-Pkg = "<0.0.1, 1"
+Pkg = "1"
 PrettyPrinting = "0.3, 0.4"
-Random = "<0.0.1, 1"
+Random = "1"
 RelocatableFolders = "0.3, 1"
 ScientificTypes = "3"
 StatisticalTraits = "3"
-Statistics = "<0.0.1, 1"
+Statistics = "1"
 StatsBase = "0.32,0.33, 0.34"
+Suppressor = "0.2.8"
 Tables = "0.2,1.0"
-julia = "1.6"
+julia = "1.10"
 
 [extras]
+Distributed = "8ba89e20-285c-5b6f-9357-94700520ee1b"
 MLJBase = "a7f614a8-145f-11e9-1d2a-a57a1082229d"
 MLJDecisionTreeInterface = "c6f25543-311c-4c74-83dc-3ea6d1015661"
 MLJMultivariateStatsInterface = "1b6a4a23-ba22-4f51-9698-8599985d3728"
 Pkg = "44cfe95a-1eb2-52ea-b672-e2afdf69b78f"
 StableRNGs = "860ef19b-820b-49d6-a774-d7a799459cd3"
+Suppressor = "fd094767-a336-5f1f-9728-57cf17d0bbfb"
 Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"
 
 [targets]
-test = ["MLJBase", "MLJDecisionTreeInterface", "MLJMultivariateStatsInterface", "Pkg", "StableRNGs", "Test"]
+test = ["Distributed", "MLJBase", "MLJDecisionTreeInterface", "MLJMultivariateStatsInterface", "Pkg", "StableRNGs", "Suppressor", "Test"]

src/MLJModels.jl

@@ -48,7 +48,6 @@ export UnivariateDiscretizer,
     OneHotEncoder, ContinuousEncoder, FillImputer, UnivariateFillImputer,
     UnivariateTimeTypeToContinuous, InteractionTransformer
 
-const srcdir = dirname(@__FILE__) # the directory containing this file
 const MMI = MLJModelInterface
 
 if VERSION < v"1.3"
@@ -64,28 +63,29 @@ include("builtins/Constant.jl")
 include("builtins/Transformers.jl")
 include("builtins/ThresholdPredictors.jl")
 
-Handle = NamedTuple{(:name, :pkg), Tuple{String,String}}
-(::Type{Handle})(name,string) = NamedTuple{(:name, :pkg)}((name, string))
+# declare paths to the metadata and associated project file:
+const REGISTRY_PROJECT = @path joinpath(@__DIR__, "registry", "Project.toml")
+const REGISTRY_METADATA = @path joinpath(@__DIR__, "registry", "Metadata.toml")
+Base.include_dependency(REGISTRY_PROJECT)
+Base.include_dependency(REGISTRY_METADATA)
 
 # load utilities for reading model metadata from file:
 include("metadata.jl")
 
-# read in the metadata:
-metadata_file = joinpath(srcdir, "registry", "Metadata.toml")
-Base.include_dependency(metadata_file)
-const INFO_GIVEN_HANDLE = info_given_handle(metadata_file)
+# read in metadata:
+const INFO_GIVEN_HANDLE = info_given_handle(REGISTRY_METADATA)
 const PKGS_GIVEN_NAME = pkgs_given_name(INFO_GIVEN_HANDLE)
 const AMBIGUOUS_NAMES = ambiguous_names(INFO_GIVEN_HANDLE)
 const NAMES = model_names(INFO_GIVEN_HANDLE)
 const MODEL_TRAITS_IN_REGISTRY = model_traits_in_registry(INFO_GIVEN_HANDLE)
 
-# model search and registry code:
+# include tools to search the model registry:
 include("model_search.jl")
+
+# include tools to load model code:
 include("loading.jl")
-include("registry/src/Registry.jl")
-using .Registry
 
-# finalize:
-include("init.jl")
+# include tool for cloning the Model Registry project file:
+include("registry_project.jl")
 
 end # module

src/metadata.jl

@@ -1,25 +1,5 @@
-## UTILITIES FOR ENCODING AND DECODING MODEL METADATA
-# (for serializing/deserializing into TOML format)
-
-# fallback encoding:
-function encode_dic(s)
-    prestring = string("`", s, "`")
-    # hack for objects with gensyms in their string representation:
-    str = replace(prestring, '#'=>'_')
-    return str
-end
-
-encode_dic(s::AbstractString) = string(s)
-encode_dic(s::Symbol) = string(":", s)
-encode_dic(s::Nothing) = "`nothing`"
-encode_dic(v::AbstractVector) = encode_dic.(v)
-function encode_dic(d::AbstractDict)
-    ret = LittleDict{}()
-    for (k, v) in d
-        ret[encode_dic(k)] = encode_dic(v)
-    end
-    return ret
-end
+# # DECODING MODEL METADATA
+# (deserializing TOML dictionary)
 
 function decode_dic(s::String)
     if !isempty(s)
@@ -51,34 +31,11 @@ function decode_dic(d::AbstractDict)
     return ret
 end
 
-# the inverse of a multivalued dictionary is a multivalued
-# dictionary:
-function inverse(d::LittleDict{S,Set{T}}) where {S,T}
-    dinv = LittleDict{T,Set{S}}()
-    for key in keys(d)
-        for val in d[key]
-            if val in keys(dinv)
-                push!(dinv[val], key)
-            else
-                dinv[val] = Set([key,])
-            end
-        end
-    end
-    return dinv
-end
-function inverse(d::Dict{S,Set{T}}) where {S,T}
-    dinv = Dict{T,Set{S}}()
-    for key in keys(d)
-        for val in d[key]
-            if val in keys(dinv)
-                push!(dinv[val], key)
-            else
-                dinv[val] = Set([key,])
-            end
-        end
-    end
-    return dinv
-end
+
+# # MODEL HANDLES
+
+Handle = NamedTuple{(:name, :pkg), Tuple{String,String}}
+(::Type{Handle})(name,string) = NamedTuple{(:name, :pkg)}((name, string))
 
 function Base.isless(h1::Handle, h2::Handle)
     if isless(h1.name, h2.name)
@@ -90,30 +47,39 @@ function Base.isless(h1::Handle, h2::Handle)
     end
 end
 
+function (::Type{Handle})(name::String)
+    if name in AMBIGUOUS_NAMES
+        return Handle(name, missing)
+    else
+        return Handle(name, first(PKGS_GIVEN_NAME[name]))
+    end
+end
+
 
-## FUNCTIONS TO BUILD GLOBAL METADATA CONSTANTS IN MLJMODELS
-## INITIALIZATION
+# # FUNCTIONS TO BUILD GLOBAL METADATA CONSTANTS
 
 # to define INFO_GIVEN_HANDLE
 function info_given_handle(metadata_file)
     metadata = LittleDict(TOML.parsefile(metadata_file))
-    metadata_given_pkg = decode_dic(metadata)
+    metadata_given_api_pkg = decode_dic(metadata)
 
     # build info_given_handle dictionary:
     ret = Dict{Handle}{Any}()
-    packages = keys(metadata_given_pkg)
-    for pkg in packages
-        info_given_name = metadata_given_pkg[pkg]
+    packages = keys(metadata_given_api_pkg)
+    for api_pkg in packages
+        info_given_name = metadata_given_api_pkg[api_pkg]
         for name in keys(info_given_name)
+            info = info_given_name[name]
+            pkg = info[:package_name]
             handle = Handle(name, pkg)
-            ret[handle] = info_given_name[name]
+            ret[handle] = info
         end
     end
     return ret
 
 end
 
-# for use in __init__ to define AMBIGUOUS_NAMES
+# to define AMBIGUOUS_NAMES
 function ambiguous_names(info_given_handle)
     names_with_duplicates = map(keys(info_given_handle) |> collect) do handle
         handle.name
@@ -124,7 +90,7 @@ function ambiguous_names(info_given_handle)
     end
 end
 
-# for use in __init__ to define PKGS_GIVEN_NAME
+# to define PKGS_GIVEN_NAME
 function pkgs_given_name(info_given_handle)
     handles = keys(info_given_handle) |> collect
     ret = Dict{String,Vector{String}}()
@@ -145,14 +111,6 @@ function model_names(info_given_handle)
     return unique(names_allowing_duplicates)
 end
 
-function (::Type{Handle})(name::String)
-    if name in AMBIGUOUS_NAMES
-        return Handle(name, missing)
-    else
-        return Handle(name, first(PKGS_GIVEN_NAME[name]))
-    end
-end
-
 function model_traits_in_registry(info_given_handle)
     first_entry = info_given_handle[Handle("ConstantRegressor")]
     return keys(first_entry) |> collect