MAINT avoid FutureWarning in test

Author: glemaitre

imblearn/over_sampling/_smote/base.py

@@ -548,7 +548,13 @@ def _fit_resample(self, X, y):
             dtype_ohe = X_continuous.dtype
         else:
             dtype_ohe = np.float64
-        self.ohe_ = OneHotEncoder(sparse=True, handle_unknown="ignore", dtype=dtype_ohe)
+
+        self.ohe_ = OneHotEncoder(handle_unknown="ignore", dtype=dtype_ohe)
+        if hasattr(self.ohe_, "sparse_output"):
+            # scikit-learn >= 1.2
+            self.ohe_.set_params(sparse_output=True)
+        else:
+            self.ohe_.set_params(sparse=True)
 
         # the input of the OneHotEncoder needs to be dense
         X_ohe = self.ohe_.fit_transform(

imblearn/over_sampling/_smote/cluster.py

@@ -125,7 +125,9 @@ class KMeansSMOTE(BaseSMOTE):
     >>> y = np.append(y, 0)
     >>> # Make this a binary classification problem
     >>> y = y == 1
-    >>> sm = KMeansSMOTE(random_state=42)
+    >>> sm = KMeansSMOTE(
+    ...     kmeans_estimator=MiniBatchKMeans(n_init=1, random_state=0), random_state=42
+    ... )
     >>> X_res, y_res = sm.fit_resample(X, y)
     >>> # Find the number of new samples in the middle blob
     >>> n_res_in_middle = ((X_res[:, 0] > -5) & (X_res[:, 0] < 5)).sum()

imblearn/over_sampling/_smote/tests/test_kmeans_smote.py

@@ -38,6 +38,7 @@ def data():
     return X, y
 
 
+@pytest.mark.filterwarnings("ignore:The default value of `n_init` will change")
 def test_kmeans_smote(data):
     X, y = data
     kmeans_smote = KMeansSMOTE(
@@ -59,13 +60,14 @@ def test_kmeans_smote(data):
     assert "batch_size" in kmeans_smote.kmeans_estimator_.get_params()
 
 
+@pytest.mark.filterwarnings("ignore:The default value of `n_init` will change")
 @pytest.mark.parametrize("k_neighbors", [2, NearestNeighbors(n_neighbors=3)])
 @pytest.mark.parametrize(
     "kmeans_estimator",
     [
         3,
-        KMeans(n_clusters=3, random_state=42),
-        MiniBatchKMeans(n_clusters=3, random_state=42),
+        KMeans(n_clusters=3, n_init=1, random_state=42),
+        MiniBatchKMeans(n_clusters=3, n_init=1, random_state=42),
     ],
 )
 def test_sample_kmeans_custom(data, k_neighbors, kmeans_estimator):
@@ -83,6 +85,7 @@ def test_sample_kmeans_custom(data, k_neighbors, kmeans_estimator):
     assert kmeans_smote.kmeans_estimator_.n_clusters == 3
 
 
+@pytest.mark.filterwarnings("ignore:The default value of `n_init` will change")
 def test_sample_kmeans_not_enough_clusters(data):
     X, y = data
     smote = KMeansSMOTE(cluster_balance_threshold=10, random_state=42)
@@ -97,6 +100,7 @@ def test_sample_kmeans_density_estimation(density_exponent, cluster_balance_thre
         n_samples=10_000, n_classes=2, weights=[0.3, 0.7], random_state=42
     )
     smote = KMeansSMOTE(
+        kmeans_estimator=MiniBatchKMeans(n_init=1, random_state=42),
         random_state=0,
         density_exponent=density_exponent,
         cluster_balance_threshold=cluster_balance_threshold,

imblearn/over_sampling/tests/test_common.py

@@ -2,6 +2,7 @@
 
 import numpy as np
 import pytest
+from sklearn.cluster import MiniBatchKMeans
 
 from imblearn.over_sampling import (
     ADASYN,
@@ -67,7 +68,13 @@ def test_smote_m_neighbors(numerical_data, smote):
     [
         (ADASYN(random_state=0), "n_neighbors"),
         (BorderlineSMOTE(random_state=0), "k_neighbors"),
-        (KMeansSMOTE(random_state=1), "k_neighbors"),
+        (
+            KMeansSMOTE(
+                kmeans_estimator=MiniBatchKMeans(n_init=1, random_state=0),
+                random_state=1,
+            ),
+            "k_neighbors",
+        ),
         (SMOTE(random_state=0), "k_neighbors"),
         (SVMSMOTE(random_state=0), "k_neighbors"),
     ],

imblearn/under_sampling/_prototype_generation/_cluster_centroids.py

@@ -93,17 +93,18 @@ class ClusterCentroids(BaseUnderSampler):
 
     >>> from collections import Counter
     >>> from sklearn.datasets import make_classification
-    >>> from imblearn.under_sampling import \
-ClusterCentroids # doctest:
+    >>> from sklearn.cluster import MiniBatchKMeans
+    >>> from imblearn.under_sampling import ClusterCentroids
     >>> X, y = make_classification(n_classes=2, class_sep=2,
     ... weights=[0.1, 0.9], n_informative=3, n_redundant=1, flip_y=0,
     ... n_features=20, n_clusters_per_class=1, n_samples=1000, random_state=10)
     >>> print('Original dataset shape %s' % Counter(y))
     Original dataset shape Counter({{1: 900, 0: 100}})
-    >>> cc = ClusterCentroids(random_state=42)
+    >>> cc = ClusterCentroids(
+    ...     estimator=MiniBatchKMeans(n_init=1, random_state=0), random_state=42
+    ... )
     >>> X_res, y_res = cc.fit_resample(X, y)
     >>> print('Resampled dataset shape %s' % Counter(y_res))
-    ... # doctest:
     Resampled dataset shape Counter({{...}})
     """
 

imblearn/under_sampling/_prototype_generation/tests/test_cluster_centroids.py

@@ -33,12 +33,14 @@
 @pytest.mark.parametrize(
     "X, expected_voting", [(X, "soft"), (sparse.csr_matrix(X), "hard")]
 )
+@pytest.mark.filterwarnings("ignore:The default value of `n_init` will change")
 def test_fit_resample_check_voting(X, expected_voting):
     cc = ClusterCentroids(random_state=RND_SEED)
     cc.fit_resample(X, Y)
     assert cc.voting_ == expected_voting
 
 
+@pytest.mark.filterwarnings("ignore:The default value of `n_init` will change")
 def test_fit_resample_auto():
     sampling_strategy = "auto"
     cc = ClusterCentroids(sampling_strategy=sampling_strategy, random_state=RND_SEED)
@@ -47,6 +49,7 @@ def test_fit_resample_auto():
     assert y_resampled.shape == (6,)
 
 
+@pytest.mark.filterwarnings("ignore:The default value of `n_init` will change")
 def test_fit_resample_half():
     sampling_strategy = {0: 3, 1: 6}
     cc = ClusterCentroids(sampling_strategy=sampling_strategy, random_state=RND_SEED)
@@ -55,6 +58,7 @@ def test_fit_resample_half():
     assert y_resampled.shape == (9,)
 
 
+@pytest.mark.filterwarnings("ignore:The default value of `n_init` will change")
 def test_multiclass_fit_resample():
     y = Y.copy()
     y[5] = 2
@@ -69,7 +73,7 @@ def test_multiclass_fit_resample():
 
 def test_fit_resample_object():
     sampling_strategy = "auto"
-    cluster = KMeans(random_state=RND_SEED)
+    cluster = KMeans(random_state=RND_SEED, n_init=1)
     cc = ClusterCentroids(
         sampling_strategy=sampling_strategy,
         random_state=RND_SEED,
@@ -84,7 +88,7 @@ def test_fit_resample_object():
 def test_fit_hard_voting():
     sampling_strategy = "auto"
     voting = "hard"
-    cluster = KMeans(random_state=RND_SEED)
+    cluster = KMeans(random_state=RND_SEED, n_init=1)
     cc = ClusterCentroids(
         sampling_strategy=sampling_strategy,
         random_state=RND_SEED,
@@ -111,6 +115,7 @@ def test_fit_resample_error(cluster_centroids_params, err_msg):
         cc.fit_resample(X, Y)
 
 
+@pytest.mark.filterwarnings("ignore:The default value of `n_init` will change")
 def test_cluster_centroids_hard_target_class():
     # check that the samples selecting by the hard voting corresponds to the
     # targeted class

imblearn/utils/_show_versions.py

@@ -8,8 +8,7 @@
 # Author: Alexander L. Hayes <[email protected]>
 # License: MIT
 
-import importlib
-import sys
+from .. import __version__
 
 
 def _get_deps_info():
@@ -20,35 +19,30 @@ def _get_deps_info():
         version information on relevant Python libraries
     """
     deps = [
+        "imbalanced-learn",
         "pip",
         "setuptools",
-        "imblearn",
-        "sklearn",
         "numpy",
         "scipy",
+        "scikit-learn",
         "Cython",
         "pandas",
         "keras",
         "tensorflow",
         "joblib",
     ]
 
-    def get_version(module):
-        return module.__version__
+    deps_info = {
+        "imbalanced-learn": __version__,
+    }
 
-    deps_info = {}
+    from importlib.metadata import PackageNotFoundError, version
 
     for modname in deps:
         try:
-            if modname in sys.modules:
-                mod = sys.modules[modname]
-            else:
-                mod = importlib.import_module(modname)
-            ver = get_version(mod)
-            deps_info[modname] = ver
-        except ImportError:
+            deps_info[modname] = version(modname)
+        except PackageNotFoundError:
             deps_info[modname] = None
-
     return deps_info
 
 

imblearn/utils/tests/test_show_versions.py

@@ -9,8 +9,8 @@ def test_get_deps_info():
     _deps_info = _get_deps_info()
     assert "pip" in _deps_info
     assert "setuptools" in _deps_info
-    assert "imblearn" in _deps_info
-    assert "sklearn" in _deps_info
+    assert "imbalanced-learn" in _deps_info
+    assert "scikit-learn" in _deps_info
     assert "numpy" in _deps_info
     assert "scipy" in _deps_info
     assert "Cython" in _deps_info
@@ -26,8 +26,8 @@ def test_show_versions_default(capsys):
     assert "machine" in out
     assert "pip" in out
     assert "setuptools" in out
-    assert "imblearn" in out
-    assert "sklearn" in out
+    assert "imbalanced-learn" in out
+    assert "scikit-learn" in out
     assert "numpy" in out
     assert "scipy" in out
     assert "Cython" in out
@@ -48,8 +48,8 @@ def test_show_versions_github(capsys):
     assert "**Python Dependencies**" in out
     assert "* pip" in out
     assert "* setuptools" in out
-    assert "* imblearn" in out
-    assert "* sklearn" in out
+    assert "* imbalanced-learn" in out
+    assert "* scikit-learn" in out
     assert "* numpy" in out
     assert "* scipy" in out
     assert "* Cython" in out