iter

Author: glemaitre

azure-pipelines.yml

@@ -20,16 +20,7 @@ jobs:
   parameters:
     name: Linux
     vmImage: ubuntu-18.04
-    dependsOn: [linting]
     matrix:
-      # Linux environment to test that scikit-learn can be built against
-      # versions of numpy, scipy with ATLAS that comes with Ubuntu Bionic 18.04
-      # i.e. numpy 1.13.3 and scipy 0.19
-      py36_ubuntu_atlas:
-        DISTRIB: 'ubuntu'
-        PYTHON_VERSION: '3.6'
-        JOBLIB_VERSION: '*'
-      # Linux environment to test the latest available dependencies and MKL.
       pylatest_pip_openblas_pandas:
         DISTRIB: 'conda-pip-latest'
         PYTHON_VERSION: '3.9'
@@ -38,6 +29,20 @@ jobs:
         TEST_DOCSTRINGS: 'true'
         JOBLIB_VERSION: '*'
         CHECK_WARNINGS: 'true'
+
+- template: build_tools/azure/posix.yml
+  parameters:
+    name: Linux
+    vmImage: ubuntu-18.04
+    dependsOn: [linting]
+    matrix:
+      # Linux environment to test that scikit-learn can be built against
+      # versions of numpy, scipy with ATLAS that comes with Ubuntu Bionic 18.04
+      # i.e. numpy 1.13.3 and scipy 0.19
+      py36_ubuntu_atlas:
+        DISTRIB: 'ubuntu'
+        PYTHON_VERSION: '3.6'
+        JOBLIB_VERSION: '*'
       pylatest_conda_pandas_keras:
         DISTRIB: 'conda'
         PYTHON_VERSION: '3.7'

imblearn/over_sampling/_random_over_sampler.py

@@ -38,17 +38,19 @@ class RandomOverSampler(BaseOverSampler):
     {random_state}
 
     smoothed_bootstrap : bool, default=False
-        Whether or not to generate smoothed bootstrap samples.
+        Whether or not to generate smoothed bootstrap samples. When this option
+        is triggered, be aware that the data to be resampled needs to be
+        numerical data since a Gaussian perturbation will be generated and
+        added to the bootstrap.
 
         .. versionadded:: 0.7
 
     shrinkage : float or dict, default=1.0
-        Factor used to shrink the covariance matrix used to generate the
-        smoothed bootstrap. If a float is given, the same factor is applied to
-        generate the bootstrap samples for the classes provided in
-        `sampling_strategy`. If a dictionary is given, different factors will
-        be used to generate the bootstrap samples. The key of the dictionary
-        corresponds to the class and the value to the shrinkage factor.
+        Factor to shrink the covariance matrix used to generate the
+        smoothed bootstrap. A factor could be shared by all classes by
+        providing a floating number or different for each class over-sampled
+        by providing a dictionary where the key are the class targeted and the
+        value is the shrinkage factor.
 
         .. versionadded:: 0.7
 
@@ -86,7 +88,7 @@ class RandomOverSampler(BaseOverSampler):
     Supports heterogeneous data as object array containing string and numeric
     data.
 
-    When generating smoothed bootstrap, this method is also known as Random
+    When generating a smoothed bootstrap, this method is also known as Random
     Over-Sampling Examples (ROSE) [1]_.
 
     .. warning::

imblearn/over_sampling/tests/test_random_over_sampler.py

@@ -93,7 +93,9 @@ def test_ros_fit_resample(X_type, data, params):
 def test_ros_fit_resample_half(data, params):
     X, Y = data
     sampling_strategy = {0: 3, 1: 7}
-    ros = RandomOverSampler(**params, sampling_strategy=sampling_strategy, random_state=RND_SEED)
+    ros = RandomOverSampler(
+        **params, sampling_strategy=sampling_strategy, random_state=RND_SEED
+    )
     X_resampled, y_resampled = ros.fit_resample(X, Y)
     X_gt = np.array(
         [