port jbrae/MLJOpenML.jl#arfffiles to this new package OpenML

ablaom · ablaom · commit 155334c089b2 · 2021-08-09T15:07:36.000+12:00
diff --git a/Project.toml b/Project.toml
@@ -1,7 +1,7 @@
-name = "MLJOpenML"
-uuid = "cbea4545-8c96-4583-ad3a-44078d60d369"
-authors = ["Anthony D. Blaom <anthony.blaom@gmail.com>"]
-version = "1.2.0"
+name = "OpenML"
+uuid = "8b6db2d4-7670-4922-a472-f9537c81ab66"
+authors = ["Diego Arenas <darenasc@gmail.com>", "Anthony D. Blaom <anthony.blaom@gmail.com>"]
+version = "0.1.0"
 
 [deps]
 ARFFFiles = "da404889-ca92-49ff-9e8b-0aa6b4d38dc8"
@@ -11,10 +11,10 @@ Markdown = "d6f4376e-aef5-505a-96c1-9c027394607a"
 ScientificTypes = "321657f4-b219-11e9-178b-2701a2544e81"
 
 [compat]
+ARFFFiles = "1.3"
 HTTP = "0.8, 0.9"
 JSON = "0.21"
 ScientificTypes = "2"
-ARFFFiles = "1.3"
 julia = "1"
 
 [extras]
diff --git a/README.md b/README.md
@@ -1,57 +1,21 @@
-# MLJOpenML.jl
+# OpenML.jl
 
-| Linux | Coverage |
-| :-----------: | :------: |
-| [![Build status](https://github.com/JuliaAI/MLJOpenML.jl/workflows/CI/badge.svg)](https://github.com/JuliaAI/MLJOpenML.jl/actions)| [![codecov.io](http://codecov.io/github/JuliaAI/MLJOpenML.jl/coverage.svg?branch=master)](http://codecov.io/github/JuliaAI/MLJOpenML.jl?branch=master) |
+| Linux | Coverage | Documentation |
+| :-----------: | :------: | :-------: |
+| [![Build status](https://github.com/JuliaAI/OpenML.jl/workflows/CI/badge.svg)](https://github.com/JuliaAI/OpenML.jl/actions)| [![codecov.io](http://codecov.io/github/JuliaAI/OpenML.jl/coverage.svg?branch=master)](http://codecov.io/github/JuliaAI/OpenML.jl?branch=master) |  [![](https://img.shields.io/badge/docs-dev-blue.svg)](https://JuliaAI.github.io/OpenML.jl/dev) |
 
-A package providing integration of [OpenML](https://www.openml.org) with the
-[MLJ](https://alan-turing-institute.github.io/MLJ.jl/dev/) machine
-learning framework.
+Partial implementation of the [OpenML](https://www.openml.org) API for
+Julia. At present this package allows querying and
+downloading of OpenML datasets. 
 
-Based entirely on Diego Arenas' original code contribution to MLJBase.jl.
+For further integration with the
+[MLJ](https://JuliaAI.github.io/MLJ.jl/dev/) machine
+learning framework (such as uploading MLJ runs) see
+[MLJOpenML.jl](https://github.com/JuliaAI/MLJOpenML.jl).
 
 
-## Installation
-
-```julia
-using Pkg
-Pkg.add("MLJOpenML")
-```
-
-## Sample usage
-
-Load the iris data set from OpenML:
-
-```julia
-using MLJOpenML
-table = MLJOpenML.load(61) # a Tables.DictColumnTable
-```
-
-Convert to a `DataFrame`:
-
-```julia
-Pkg.add("DataFrames")
-using DataFrames
-df = DataFrame(table)
-```
-
-Browsing and filtering datasets:
-
-```julia
-using DataFrames
-ds = MLJOpenML.list_datasets(output_format = DataFrame)
-MLJOpenML.describe_dataset(6)
-MLJOpenML.list_tags() # lists valid tags
-ds = MLJOpenML.list_datasets(tag = "OpenML100", 
-                             filter = "number_instances/100..1000/number_features/1..10",
-                             output_format = DataFrame)
-```
-
-## Documentation
-
-Documentation is provided in the [OpenML
-Integration](https://alan-turing-institute.github.io/MLJ.jl/dev/openml_integration/)
-section of the
-[MLJManual](https://alan-turing-institute.github.io/MLJ.jl/dev/)
-
+The code in this repository is based on contributions of Diego Arenas
+to [MLJBase.jl](https://github.com/JuliaAI/MLJBase.jl) which do not
+appear in the commit history of this repository.
 
+Package documentation is [here](https://JuliaAI.github.io/OpenML.jl/dev).
diff --git a/docs/Project.toml b/docs/Project.toml
@@ -1,5 +1,8 @@
 [deps]
+DataFrames = "a93c6f00-e57d-5684-b7b6-d8193f3e46c0"
 Documenter = "e30172f5-a6a5-5a46-863b-614d45cd2de4"
+OpenML = "8b6db2d4-7670-4922-a472-f9537c81ab66"
+ScientificTypes = "321657f4-b219-11e9-178b-2701a2544e81"
 
 [compat]
 Documenter = "~0.26"
diff --git a/docs/make.jl b/docs/make.jl
@@ -1,10 +1,10 @@
-using Documenter, MLJOpenML
+using Documenter, OpenML, DataFrames
 
 makedocs(
-    modules = [MLJOpenML],
-    sitename = "MLJOpenML.jl",
+    modules = [OpenML,],
+    sitename = "OpenML.jl",
 )
 
 deploydocs(
-    repo = "github.com/alan-turing-institute/MLJOpenML.jl.git",
+    repo = "github.com/JuliaAI/OpenML.jl.git",
 )
diff --git a/docs/src/index.md b/docs/src/index.md
@@ -1,8 +1,83 @@
-# MLJOpenML.jl Documentation
+# OpenML.jl Documentation
 
 This is the reference documentation of
-[`MLJOpenML.jl`](https://github.com/alan-turing-institute/MLJOpenML.jl).
+[`OpenML.jl`](https://github.com/JuliaAI/OpenML.jl).
 
-```@autodocs
-Modules = [MLJOpenML]
+The [OpenML platform](https://www.openml.org) provides an integration
+platform for carrying out and comparing machine learning solutions
+across a broad collection of public datasets and software platforms.
+
+Summary of OpenML.jl functionality:
+
+- [`OpenML.list_tags`](@ref)`()`: for listing all dataset tags
+        
+- [`OpenML.list_datasets`](@ref)`(; tag=nothing, filter=nothing, output_format=...)`: for listing available datasets
+
+- [`OpenML.describe_dataset`](@ref)`(id)`: to describe a particular dataset
+
+- [`OpenML.load`](@ref)`(id; parser=:arff)`: to download a dataset
+
+
+## Installation
+
+```julia
+using Pkg
+Pkg.add("OpenML")
+```
+
+If running the demonstration below:
+
+```julia
+Pkg.add("DataFrames") 
+Pkg.add("ScientificTypes")
+```
+
+## Sample usage
+
+```@repl new
+using OpenML # or using MLJ
+using DataFrames
+
+OpenML.list_tags()
 ```
+
+Listing all datasets with the "OpenML100" tag which also have `n`
+instances and `p` features, where `100 < n < 1000` and `1 < p < 10`:
+
+```@repl new
+ds = OpenML.list_datasets(
+          tag = "OpenML100",
+          filter = "number_instances/100..1000/number_features/1..10",
+          output_format = DataFrame)
+```
+
+Describing and loading one of these datasets:
+
+```@repl new
+OpenML.describe_dataset(15)
+table = OpenML.load(15)
+```
+
+Converting to a data frame:
+
+```@repl new
+df = DataFrame(table)
+```
+
+Inspecting it's schema:
+
+```@repl new
+using ScientificTypes
+schema(table)
+```
+
+## Public API
+
+```@docs
+OpenML.list_tags
+OpenML.list_datasets
+OpenML.describe_dataset
+OpenML.load
+```
+
+
diff --git a/src/MLJOpenML.jl b/src/MLJOpenML.jl
diff --git a/src/OpenML.jl b/src/OpenML.jl
@@ -0,0 +1,13 @@
+module OpenML
+
+using HTTP
+using JSON
+import ARFFFiles
+import ScientificTypes: Continuous, Count, Textual, Multiclass, coerce, autotype
+using Markdown
+
+export OpenML
+
+include("data.jl")
+
+end # module
diff --git a/src/data.jl b/src/data.jl
@@ -1,9 +1,3 @@
-using HTTP
-using JSON
-import ARFFFiles
-import ScientificTypes: Continuous, Count, Textual, Multiclass, coerce, autotype
-using Markdown
-
 const API_URL = "https://www.openml.org/api/v1/json"
 
 # Data API
@@ -45,7 +39,7 @@ function load_Dataset_Description(id::Int; api_key::String="")
 end
 
 """
-    MLJOpenML.load(id; parser = :arff)
+    OpenML.load(id; parser = :arff)
 
 Load the OpenML dataset with specified `id`, from those listed by
 [`list_datasets`](@ref) or on the [OpenML site](https://www.openml.org/search?type=data).
@@ -59,7 +53,7 @@ Returns a table.
 
 ```julia
 using DataFrames
-table = MLJOpenML.load(61);
+table = OpenML.load(61);
 df = DataFrame(table);
 ```
 """
@@ -225,7 +219,7 @@ API](https://www.openml.org/api_docs#!/data/get_data_list_filters).
 ```
 julia> using DataFrames
 
-julia> ds = MLJOpenML.list_datasets(
+julia> ds = OpenML.list_datasets(
                tag = "OpenML100",
                filter = "number_instances/100..1000/number_features/1..10",
                output_format = DataFrame
@@ -244,7 +238,7 @@ function list_datasets(; tag = nothing, filter = "", filters=filter,
             return
         end
     end
-    data = MLJOpenML.load_List_And_Filter(filters; api_key = api_key)
+    data = OpenML.load_List_And_Filter(filters; api_key = api_key)
     datasets = data["data"]["dataset"]
     qualities = Symbol.(union(vcat([vcat(qualitynames.(entry["quality"])...) for entry in datasets]...)))
     result = merge((id = Int[], name = String[], status = String[]),
@@ -292,7 +286,7 @@ Use [`list_datasets`](@ref) to browse available data sets.
 
 # Examples
 ```
-julia> MLJOpenML.describe_dataset(6)
+julia> OpenML.describe_dataset(6)
   Author: David J. Slate Source: UCI
   (https://archive.ics.uci.edu/ml/datasets/Letter+Recognition) - 01-01-1991 Please cite: P.
   W. Frey and D. J. Slate. "Letter Recognition Using Holland-style Adaptive Classifiers".
diff --git a/test/data.jl b/test/data.jl
@@ -2,18 +2,18 @@ module TestOpenml
 
 using Test
 using HTTP
-using MLJOpenML
+using OpenML
 import Tables.istable
 
-response_test = MLJOpenML.load_Dataset_Description(61)
-ntp_test = MLJOpenML.load(61)
+response_test = OpenML.load_Dataset_Description(61)
+ntp_test = OpenML.load(61)
 @test istable(ntp_test)
-dqlist_test = MLJOpenML.load_Data_Qualities_List()
-data_features_test = MLJOpenML.load_Data_Features(61)
-data_qualities_test = MLJOpenML.load_Data_Qualities(61)
+dqlist_test = OpenML.load_Data_Qualities_List()
+data_features_test = OpenML.load_Data_Features(61)
+data_qualities_test = OpenML.load_Data_Qualities(61)
 limit = 5
 offset = 8
-filters_test = MLJOpenML.load_List_And_Filter("limit/$limit/offset/$offset")
+filters_test = OpenML.load_List_And_Filter("limit/$limit/offset/$offset")
 
 @testset "HTTP connection" begin
     @test typeof(response_test) <: Dict
diff --git a/test/runtests.jl b/test/runtests.jl
@@ -1 +1 @@
-include("openml.jl")
+include("data.jl")

Original file line number	Diff line number	Diff line change
`@@ -1,10 +1,10 @@`
`1`		`-using Documenter, MLJOpenML`
	`1`	`+using Documenter, OpenML, DataFrames`
`2`	`2`
`3`	`3`	`makedocs(`
`4`		`- modules = [MLJOpenML],`
`5`		`- sitename = "MLJOpenML.jl",`
	`4`	`+ modules = [OpenML,],`
	`5`	`+ sitename = "OpenML.jl",`
`6`	`6`	`)`
`7`	`7`
`8`	`8`	`deploydocs(`
`9`		`- repo = "github.com/alan-turing-institute/MLJOpenML.jl.git",`
	`9`	`+ repo = "github.com/JuliaAI/OpenML.jl.git",`
`10`	`10`	`)`
Original file line number	Diff line number	Diff line change
`@@ -1 +1 @@`
`1`		`-include("openml.jl")`
	`1`	`+include("data.jl")`