Merge pull request #1192 from JuliaAI/drop-mljflow

Drop MLJFlow as a dependency of MLJ

Author: ablaom

Project.toml

@@ -1,7 +1,7 @@
 name = "MLJ"
 uuid = "add582a8-e3ab-11e8-2d5e-e98b27df1bc7"
 authors = ["Anthony D. Blaom <anthony.blaom@gmail.com>"]
-version = "0.22.0"
+version = "0.23.0"
 
 [deps]
 CategoricalArrays = "324d7699-5711-5eae-9e2f-1d82baa6b597"
@@ -13,7 +13,6 @@ LinearAlgebra = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"
 MLJBalancing = "45f359ea-796d-4f51-95a5-deb1a414c586"
 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"
 MLJTransforms = "23777cdb-d90c-4eb0-a694-7c2b83d5c1d6"
@@ -37,7 +36,6 @@ FeatureSelection = "0.2"
 MLJBalancing = "0.1"
 MLJBase = "1.5"
 MLJEnsembles = "0.4"
-MLJFlow = "0.5"
 MLJIteration = "0.6"
 MLJModels = "0.18"
 MLJTestIntegration = "0.5.0"

docs/src/logging_workflows.md

@@ -1,7 +1,11 @@
+!!! note
+
+	Starting with MLJ 0.23.0, MLJFlow.jl methods are not immediately available, as the package has been removed as a direct dependency. Previous behaviour will require `using MLJFlow`. See also the "Warning" below.
+
 # Logging Workflows
 
-Currently the following workflows can log their outcomes to an external machine learning
-tracking platform, such as [MLflow](https://mlflow.org):
+In principle, the following workflows can log their outcomes to an external machine learning
+tracking platform, such as [mlflow](https://mlflow.org):
 
 - Estimating model performance using [`evaluate`](@ref)/[`evaluate!`](@ref).
 
@@ -21,14 +25,13 @@ documentation.
 
 ### Supported tracking platforms
 
-- [MLflow](https://mlflow.org) is natively supported by MLJ. You will still need to
-  install MLflow itself, and separately launch an MLflow service; see the [MLflow
-  docs](https://mlflow.org) on how to do this. The service can immediately be wrapped
-  to create a `logger` object, as demonstrated in the [MLJFlow.jl
-  documentation](https://github.com/JuliaAI/MLJFlow.jl).
-
-
 !!! warning
 
-    MLJFlow.jl is a new package still under active development and should be regarded as experimental. At this time, breaking changes to MLJFlow.jl will not necessarily trigger new breaking releases of MLJ.jl.
+    Due to issues with the mlflow REST API, the current model for MLJ-mlflow integration is being reassessed. Use the existing tools at your own risk.
 
+- To use [mlflow](https://mlflow.org) with MLJ you will need to add MLJFlow to your
+  package environment and call `using MLJFlow`. You additionally need to install
+  mlflow itself, and separately launch an mlflow service; see the [mlflow
+  docs](https://mlflow.org) on how to do this. The service can immediately be wrapped to
+  create a `logger` object, as demonstrated in the [MLJFlow.jl
+  documentation](https://github.com/JuliaAI/MLJFlow.jl).

src/MLJ.jl

@@ -1,9 +1,9 @@
 """
-   MLJ 
+   MLJ
 
 [`MLJ`](https://juliaai.github.io/MLJ.jl//dev/) is a Machine Learning toolbox for
 Julia. It collects together functionality from separate components listed below, which can
-be loaded individually.
+be loaded individually. 
 
 Actual model code (e.g., code for instantiating a `DecisionTreeClassifier`) must generally
 be loaded explicitly from the model-providing package, using `@load`, for example. However
@@ -41,12 +41,8 @@ core model wrappers and some common transformers are immediately available; do
 - FeatureSelection.jl: Transformers for feature selection, and the supervised model wrapper
   `RecursiveFeatureSelection`.
 
-- MLJFlow.jl: Integration with MLflow workflow tracking
-
 - OpenML.jl: Tool for grabbing datasets from OpenML.org
 
-
-
 """
 module MLJ
 
@@ -68,7 +64,6 @@ using MLJTuning
 using MLJModels
 @reexport using FeatureSelection
 using OpenML
-@reexport using MLJFlow
 @reexport using StatisticalMeasures
 import MLJBalancing
 @reexport using MLJBalancing: BalancedModel

test/exported_names.jl

@@ -35,7 +35,7 @@ Save()
 
 # MLJFlow
 
-MLJFlow.Logger
+# MLJFlow.Logger
 
 # StatisticalMeasures