Add MLJFlow for integration with MLflow logging platform

Project.toml

@@ -11,13 +11,15 @@ Distributions = "31c24e10-a181-5473-b8eb-7969acd0382f"
 LinearAlgebra = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"
 MLJBase = "a7f614a8-145f-11e9-1d2a-a57a1082229d"
 MLJEnsembles = "50ed68f4-41fd-4504-931a-ed422449fee0"
+MLJFlow = "7b7b8358-b45c-48ea-a8ef-7ca328ad328f"
 MLJIteration = "614be32b-d00c-4edb-bd02-1eb411ab5e55"
 MLJModels = "d491faf4-2d78-11e9-2867-c94bc002c0b7"
 MLJTuning = "03970b2e-30c4-11ea-3135-d1576263f10f"
 OpenML = "8b6db2d4-7670-4922-a472-f9537c81ab66"
 Pkg = "44cfe95a-1eb2-52ea-b672-e2afdf69b78f"
 ProgressMeter = "92933f4c-e287-5a05-a399-4b506db050ca"
 Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
+Reexport = "189a3867-3050-52da-a836-e630ba90ab69"
 ScientificTypes = "321657f4-b219-11e9-178b-2701a2544e81"
 Statistics = "10745b16-79ce-11e8-11f9-7d13ad32a3b2"
 StatsBase = "2913bbd2-ae8a-5f71-8c99-4fb6c76f3a91"
@@ -29,11 +31,13 @@ ComputationalResources = "0.3"
 Distributions = "0.21,0.22,0.23, 0.24, 0.25"
 MLJBase = "0.21.3"
 MLJEnsembles = "0.3"
+MLJFlow = "0.1"
 MLJIteration = "0.5"
 MLJModels = "0.16"
 MLJTuning = "0.7"
 OpenML = "0.2,0.3"
 ProgressMeter = "1.1"
+Reexport = "1.2"
 ScientificTypes = "3"
 StatsBase = "0.32,0.33, 0.34"
 Tables = "0.2,1.0"

docs/make.jl

@@ -74,6 +74,7 @@ pages = [
     "Learning Networks" => "learning_networks.md",
     "Controlling Iterative Models" => "controlling_iterative_models.md",
     "Generating Synthetic Data" => "generating_synthetic_data.md",
+    "Logging Workflows" => "logging_workflows.md",
     "OpenML Integration" => "openml_integration.md",
     "Acceleration and Parallelism" => "acceleration_and_parallelism.md",
     "Simple User Defined Models" => "simple_user_defined_models.md",

docs/src/evaluating_model_performance.md

@@ -9,8 +9,11 @@ In addition to hold-out and cross-validation, the user can specify
 an explicit list of train/test pairs of row indices for resampling, or
 define new resampling strategies.
 
-For simultaneously evaluating *multiple* models and/or data
-sets, see [Benchmarking](benchmarking.md).
+For simultaneously evaluating *multiple* models, see [Comparing models of different type
+and nested cross-validation](@ref).
+
+For externally logging the outcomes of performance evaluation experiments, see [Logging
+Workflows](@ref)
 
 ## Evaluating against a single measure
 

docs/src/logging_workflows.md

@@ -0,0 +1,12 @@
+# Logging Workflows
+
+## MLflow integration
+
+[MLflow](https://mlflow.org) is a popular, language-agnostic, tool for externally logging
+the outcomes of machine learning experiments, including those carried out using MLJ.
+
+This functionality is provided by the [MLJFlow.jl](https://github.com/JuliaAI/MLJFlow.jl)
+package whose methods are automatically available to MLJ users. Refer to the package's
+documentation for examples.
+
+

src/MLJ.jl

@@ -8,13 +8,16 @@ import Distributed: @distributed, nworkers, pmap
 import Pkg
 import Pkg.TOML
 
+using Reexport
+
 # from the MLJ universe:
 using MLJBase
 import MLJBase.save
 using MLJEnsembles
 using MLJTuning
 using MLJModels
 using OpenML
+@reexport using MLJFlow
 using MLJIteration
 import MLJIteration.IterationControl
 
@@ -89,7 +92,7 @@ export nrows, color_off, color_on,
     @load_boston, @load_ames, @load_iris, @load_reduced_ames, @load_crabs,
     load_boston, load_ames, load_iris, load_reduced_ames, load_crabs,
     Machine, machine, AbstractNode, @node,
-    source, node, fit!, freeze!, thaw!, Node, sources, origins,
+
     machines, sources, anonymize!, @from_network, fitresults,
     @pipeline, Stack, Pipeline, TransformedTargetModel,
     ResamplingStrategy, Holdout, CV, TimeSeriesCV,

test/exported_names.jl

@@ -22,4 +22,9 @@ Save()
 
 @test OpenML.load isa Function
 
+
+# MLJFlow
+
+MLFlowLogger
+
 true