MAINT remove deprecated and synchronize with scikit-learn master (#617)

Author: glemaitre

.readthedocs.yml

@@ -4,7 +4,7 @@ formats:
   - none
 requirements_file: requirements.txt
 python:
-  version: 3.6
+  version: 3.7
   pip_install: true
   extra_requirements:
     - tests

.travis.yml

@@ -1,5 +1,5 @@
 # make it explicit that we favor the new container-based travis workers
-dist: xenial
+dist: bionic
 sudo: false
 
 language: python
@@ -32,16 +32,12 @@ matrix:
     # Ubuntu 14.04 environment
     - env: DISTRIB="ubuntu"
     # Latest release
-    - env: DISTRIB="conda" PYTHON_VERSION="3.6"
-           NUMPY_VERSION="*" SCIPY_VERSION="*" SKLEARN_VERSION="0.21.2"
-           OPTIONAL_DEPS="true"
     - env: DISTRIB="conda" PYTHON_VERSION="3.7"
-           NUMPY_VERSION="*" SCIPY_VERSION="*" SKLEARN_VERSION="0.21.2"
-           OPTIONAL_DEPS="false"
+           NUMPY_VERSION="*" SCIPY_VERSION="*" SKLEARN_VERSION="master"
+           OPTIONAL_DEPS="keras"
     - env: DISTRIB="conda" PYTHON_VERSION="3.7"
            NUMPY_VERSION="*" SCIPY_VERSION="*" SKLEARN_VERSION="master"
-           OPTIONAL_DEPS="false"
-  allow_failures:
+           OPTIONAL_DEPS="tensorflow"
     - env: DISTRIB="conda" PYTHON_VERSION="3.7"
            NUMPY_VERSION="*" SCIPY_VERSION="*" SKLEARN_VERSION="master"
            OPTIONAL_DEPS="false"

README.rst

@@ -162,10 +162,9 @@ Below is a list of the methods currently implemented in this module.
     2. SMOTE + ENN [11]_
 
 * Ensemble classifier using samplers internally
-    1. EasyEnsemble [13]_
-    2. BalanceCascade [13]_
-    3. Balanced Random Forest [16]_
-    4. Balanced Bagging
+    1. Easy Ensemble classifier [13]_
+    2. Balanced Random Forest [16]_
+    3. Balanced Bagging
 
 The different algorithms are presented in the sphinx-gallery_.
 

appveyor.yml

@@ -10,7 +10,12 @@ environment:
     - PYTHON: "C:\\Miniconda36-x64"
       PYTHON_VERSION: "3.6"
       PYTHON_ARCH: "64"
-      OPTIONAL_DEP: "pandas keras tensorflow"
+      OPTIONAL_DEP: "pandas keras tensorflow=1"
+
+    - PYTHON: "C:\\Miniconda36-x64"
+      PYTHON_VERSION: "3.6"
+      PYTHON_ARCH: "64"
+      OPTIONAL_DEP: "pandas tensorflow"
 
     - PYTHON: "C:\\Miniconda36-x64"
       PYTHON_VERSION: "3.7"
@@ -30,8 +35,8 @@ install:
   - conda update conda -y -q
   - conda create -n testenv --yes python=%PYTHON_VERSION% pip
   - activate testenv
-  - conda install scipy numpy -y -q
-  - conda install scikit-learn -y -q
+  - conda install scipy numpy joblib -y -q
+  - pip install --pre -f https://sklearn-nightly.scdn8.secure.raxcdn.com scikit-learn
   - conda install %OPTIONAL_DEP% -y -q
   - conda install pytest pytest-cov -y -q
   - pip install codecov

build_tools/circle/build_doc.sh

@@ -88,13 +88,14 @@ conda update --yes --quiet conda
 
 # Configure the conda environment and put it in the path using the
 # provided versions
-conda create -n $CONDA_ENV_NAME --yes --quiet python=3.6
+conda create -n $CONDA_ENV_NAME --yes --quiet python=3.7
 source activate $CONDA_ENV_NAME
 
-conda install --yes pip numpy scipy pillow matplotlib sphinx \
-      sphinx_rtd_theme numpydoc pandas keras
-pip install --pre scikit-learn
+conda install --yes pip numpy scipy joblib pillow matplotlib sphinx \
+      memory_profiler sphinx_rtd_theme pandas keras tensorflow=1
+pip install --pre -f https://sklearn-nightly.scdn8.secure.raxcdn.com scikit-learn
 pip install -U git+https://github.com/sphinx-gallery/sphinx-gallery.git
+pip install -U git+https://github.com/numpy/numpydoc.git
 
 # Build and install imbalanced-learn in dev mode
 ls -l

build_tools/travis/install.sh

@@ -28,27 +28,24 @@ if [[ "$DISTRIB" == "conda" ]]; then
     MINICONDA_PATH=/home/travis/miniconda
     chmod +x miniconda.sh && ./miniconda.sh -b -p $MINICONDA_PATH
     export PATH=$MINICONDA_PATH/bin:$PATH
-    conda config --set always_yes yes --set changeps1 no
-    conda install conda=4.6
-    conda update -q conda
-    conda info -a
 
     # Configure the conda environment and put it in the path using the
     # provided versions
     conda create -n testenv --yes python=$PYTHON_VERSION pip
     source activate testenv
     conda install --yes numpy=$NUMPY_VERSION scipy=$SCIPY_VERSION
 
-    if [[ "$OPTIONAL_DEPS" == "true" ]]; then
-        conda install --yes pandas keras tensorflow
+    if [[ "$OPTIONAL_DEPS" == "keras" ]]; then
+        conda install --yes pandas keras tensorflow=1
         KERAS_BACKEND=tensorflow
         python -c "import keras.backend"
         sed -i -e 's/"backend":[[:space:]]*"[^"]*/"backend":\ "'$KERAS_BACKEND'/g' ~/.keras/keras.json;
+    elif [[ "$OPTIONAL_DEPS" == "tensorflow" ]]; then
+        conda install --yes pandas tensorflow
     fi
 
     if [[ "$SKLEARN_VERSION" == "master" ]]; then
-        conda install --yes cython
-        pip install -U git+https://github.com/scikit-learn/scikit-learn.git
+        pip install --pre -f https://sklearn-nightly.scdn8.secure.raxcdn.com scikit-learn
     else
         conda install --yes scikit-learn=$SKLEARN_VERSION
     fi
@@ -65,8 +62,8 @@ elif [[ "$DISTRIB" == "ubuntu" ]]; then
     virtualenv --system-site-packages --python=python3 testvenv
     source testvenv/bin/activate
 
-    pip3 install scikit-learn
-    pip3 install pandas keras tensorflow
+    pip install --pre -f https://sklearn-nightly.scdn8.secure.raxcdn.com scikit-learn
+    pip3 install pandas
     pip3 install pytest pytest-cov codecov sphinx numpydoc
 
 fi

doc/api.rst

@@ -111,10 +111,8 @@ Prototype selection
    :toctree: generated/
    :template: class.rst
 
-   ensemble.BalanceCascade
    ensemble.BalancedBaggingClassifier
    ensemble.BalancedRandomForestClassifier
-   ensemble.EasyEnsemble
    ensemble.EasyEnsembleClassifier
    ensemble.RUSBoostClassifier
 
@@ -251,5 +249,4 @@ Imbalance-learn provides some fast-prototyping tools.
 
    utils.estimator_checks.check_estimator
    utils.check_neighbors_object
-   utils.check_ratio
    utils.check_sampling_strategy

doc/conf.py

@@ -265,6 +265,7 @@
 sphinx_gallery_conf = {
     'doc_module': 'imblearn',
     'backreferences_dir': os.path.join('generated'),
+    'show_memory': True,
     'reference_url': {
         'imblearn': None}
 }

doc/whats_new/v0.6.rst

@@ -6,3 +6,46 @@ Version 0.6.0 (under-development)
 Changelog
 ---------
 
+Changed models
+..............
+
+The following models might give some different sampling due to changes in
+scikit-learn:
+
+- :class:`imblearn.under_sampling.ClusterCentroid`
+- :class:`imblearn.under_sampling.InstanceHardnessThreshold`
+
+Maintenance
+...........
+
+- Update imports from scikit-learn after that some modules have been privatize.
+  The following import have been changed:
+  :class:`sklearn.ensemble._base._set_random_states`,
+  :class:`sklearn.ensemble._forest._parallel_build_trees`,
+  :class:`sklearn.metrics._classification._check_targets`,
+  :class:`sklearn.metrics._classification._prf_divide`,
+  :class:`sklearn.utils.Bunch`,
+  :class:`sklearn.utils._safe_indexing`,
+  :class:`sklearn.utils._testing.assert_allclose`,
+  :class:`sklearn.utils._testing.assert_array_equal`,
+  :class:`sklearn.utils._testing.SkipTest`.
+  :pr:`617` by :user:`Guillaume Lemaitre <glemaitre>`.
+
+Deprecation
+...........
+
+- The following classes have been removed after 2 deprecation cycles:
+  `ensemble.BalanceCascade` and `ensemble.EasyEnsemble`.
+  :pr:`617` by :user:`Guillaume Lemaitre <glemaitre>`.
+
+- The following functions have been removed after 2 deprecation cycles:
+  `utils.check_ratio`.
+  :pr:`617` by :user:`Guillaume Lemaitre <glemaitre>`.
+
+- The parameter `ratio` and `return_indices` has been removed from all
+  samplers.
+  :pr:`617` by :user:`Guillaume Lemaitre <glemaitre>`.
+
+- The parameters `m_neighbors`, `out_step`, `kind`, `svm_estimator`
+  have been removed from the :class:`imblearn.over_sampling.SMOTE`.
+  :pr:`617` by :user:`Guillaume Lemaitre <glemaitre>`.

examples/ensemble/plot_comparison_ensemble_classifier.py

@@ -111,9 +111,8 @@ def plot_confusion_matrix(cm, classes, ax,
 # will use a bagging classifier and its counter part which internally uses a
 # random under-sampling to balanced each boostrap sample.
 
-bagging = BaggingClassifier(n_estimators=50, random_state=0, n_jobs=-1)
-balanced_bagging = BalancedBaggingClassifier(n_estimators=50, random_state=0,
-                                             n_jobs=-1)
+bagging = BaggingClassifier(n_estimators=50, random_state=0)
+balanced_bagging = BalancedBaggingClassifier(n_estimators=50, random_state=0)
 
 bagging.fit(X_train, y_train)
 balanced_bagging.fit(X_train, y_train)
@@ -149,9 +148,8 @@ def plot_confusion_matrix(cm, classes, ax,
 # outperforming bagging. Here, we used a vanilla random forest and its balanced
 # counterpart in which each bootstrap sample is balanced.
 
-rf = RandomForestClassifier(n_estimators=50, random_state=0, n_jobs=-1)
-brf = BalancedRandomForestClassifier(n_estimators=50, random_state=0,
-                                     n_jobs=-1)
+rf = RandomForestClassifier(n_estimators=50, random_state=0)
+brf = BalancedRandomForestClassifier(n_estimators=50, random_state=0)
 
 rf.fit(X_train, y_train)
 brf.fit(X_train, y_train)
@@ -189,8 +187,7 @@ def plot_confusion_matrix(cm, classes, ax,
 
 base_estimator = AdaBoostClassifier(n_estimators=10)
 eec = EasyEnsembleClassifier(n_estimators=10,
-                             base_estimator=base_estimator,
-                             n_jobs=-1)
+                             base_estimator=base_estimator)
 eec.fit(X_train, y_train)
 y_pred_eec = eec.predict(X_test)
 print('Easy ensemble classifier performance:')