Remove redundant line in vias_variance_decomp (#652)

*  remove redundant line in vias_variance_decomp

* add version print to travis test script

* fix unit tests for sklean 0.22

* upd travis

Author: rasbt

.appveyor.yml

@@ -16,8 +16,8 @@ install:
   - conda create -q -n test-environment python=%PYTHON_VERSION% numpy scipy scikit-learn pandas joblib pytest
   - activate test-environment
   - conda install -c conda-forge dlib=19.17
-  - pip install imageio
-  - pip install scikit-image
+  - conda install imageio
+  - conda install scikit-image
 
 test_script:
   - set PYTHONPATH=%PYTHONPATH%;%CD%

.travis.yml

@@ -20,7 +20,7 @@ matrix:
         - os: linux
           sudo: required
           python: 3.8
-          env: LATEST="false" IMAGE="true" COVERAGE="false" NUMPY_VERSION="1.16.2" SCIPY_VERSION="1.2.1" SKLEARN_VERSION="0.20.3" JOBLIB_VERSION=0.13.1 PANDAS_VERSION="0.24.2" IMAGEIO_VERSION="2.5.0" SKIMAGE_VERSION="0.15.0" DLIB_VERSION="19.17.0" MINICONDA_PYTHON_VERSION=3.7
+          env: LATEST="false" IMAGE="true" COVERAGE="false" NUMPY_VERSION="1.17.4" SCIPY_VERSION="1.3.1" SKLEARN_VERSION="0.22.0" JOBLIB_VERSION=0.13.1 PANDAS_VERSION="0.25.3" IMAGEIO_VERSION="2.5.0" SKIMAGE_VERSION="0.15.0" DLIB_VERSION="19.17.0" MINICONDA_PYTHON_VERSION=3.7
         - os: linux
           python: 3.8
           env: LATEST="true" IMAGE="true" COVERAGE="true" NOTEBOOKS="true" MINICONDA_PYTHON_VERSION=3.7

ci/.travis_install.sh

@@ -33,7 +33,9 @@ source activate test-environment
 if [ "${LATEST}" = "true" ]; then
     conda install numpy scipy pandas scikit-learn joblib
 else
-    conda install numpy="$NUMPY_VERSION" scipy="$SCIPY_VERSION" pandas="$PANDAS_VERSION" scikit-learn="$SKLEARN_VERSION" joblib="$JOBLIB_VERSION"
+    conda install numpy="$NUMPY_VERSION" scipy="$SCIPY_VERSION" pandas="$PANDAS_VERSION" joblib="$JOBLIB_VERSION"
+    # temporary fix because 0.22 cannot be installed from the main conda branch
+    conda install scikit-learn="$SKLEARN_VERSION" -c conda-forge
 fi
 
 

ci/.travis_test.sh

@@ -2,6 +2,12 @@
 
 set -e
 
+python --version
+python -c "import pandas; print('pandas %s' % pandas.__version__)"
+python -c "import numpy; print('numpy %s' % numpy.__version__)"
+python -c "import scipy; print('scipy %s' % scipy.__version__)"
+python -c "import sklearn; print('sklearn %s' % sklearn.__version__)"
+python -c "import mlxtend; print('mlxtend %s' % mlxtend.__version__)"
 
 if [[ "$TRAVIS_OS_NAME" != "osx" ]]; then 
 

mlxtend/classifier/tests/test_stacking_classifier.py

@@ -112,7 +112,7 @@ def test_weight_unsupported():
     w = np.array([random.random() for _ in range(len(y))])
 
     with pytest.raises(TypeError):
-        sclf.fit(X, y, sample_seight=w)
+        sclf.fit(X, y, sample_weight=w)
 
 
 def test_weight_unsupported_no_weight():
@@ -312,7 +312,7 @@ def test_not_fitted():
                               use_probas=True,
                               meta_classifier=meta)
 
-    X, y = iris_data()
+    X, _ = iris_data()
 
     assert_raises(NotFittedError,
                   "This StackingClassifier instance is not fitted yet."
@@ -496,7 +496,7 @@ def test_predict_meta_features():
                             multi_class='ovr',
                             random_state=1)
     gnb = GaussianNB()
-    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3)
+    X_train, X_test, y_train, _ = train_test_split(X, y, test_size=0.3)
 
     #  test default (class labels)
     stclf = StackingClassifier(classifiers=[knn, gnb],
@@ -552,15 +552,15 @@ def test_decision_function():
     sclf = StackingClassifier(classifiers=[clf1, clf2],
                               use_probas=True,
                               meta_classifier=meta)
-    
+
     scores = cross_val_score(sclf,
                              X,
                              y2,
                              cv=5,
                              scoring='roc_auc')
     scores_mean = (round(scores.mean(), 2))
 
-    if Version(sklearn_version) < Version("0.21"):
+    if Version(sklearn_version) < Version("0.22"):
         assert scores_mean == 0.95, scores_mean
     else:
-        assert scores_mean == 0.95, scores_mean
+        assert scores_mean == 0.94, scores_mean

mlxtend/classifier/tests/test_stacking_cv_classifier.py

@@ -429,8 +429,10 @@ def test_meta_feat_reordering():
 
     if Version(sklearn_version) < Version("0.21"):
         expected_value = 0.86
-    else:
+    elif Version(sklearn_version) < Version("0.22"):
         expected_value = 0.87
+    else:
+        expected_value = 0.85
 
     assert round(roc_auc_score(y_train,
                  stclf.train_meta_features_[:, 1]), 2) == expected_value, \
@@ -480,10 +482,7 @@ def test_sparse_inputs_with_features_in_secondary():
     # dense
     stclf.fit(X_train, y_train)
 
-    if Version(sklearn_version) < Version("0.21"):
-        expected_value = 1.0
-    else:
-        expected_value = 0.99
+    expected_value = 1.0
 
     assert round(stclf.score(X_train, y_train), 2) == expected_value, \
         round(stclf.score(X_train, y_train), 2)
@@ -493,8 +492,12 @@ def test_sparse_inputs_with_features_in_secondary():
 
     if Version(sklearn_version) < Version("0.21"):
         expected_value = 1.0
-    else:
+    if Version(sklearn_version) < Version("0.22"):
         expected_value = 0.99
+    else:
+        expected_value = 1.00
+
+
     assert round(stclf.score(X_train, y_train), 2) == expected_value, \
         round(stclf.score(X_train, y_train), 2)
 
@@ -567,8 +570,13 @@ def test_works_with_df_if_fold_indexes_missing():
 
     # dense
     stclf.fit(X_train, y_train)
-    assert round(stclf.score(X_train, y_train), 2) == 0.99, \
-        round(stclf.score(X_train, y_train), 2)
+
+    if Version(sklearn_version) < Version("0.22"):
+        assert round(stclf.score(X_train, y_train), 2) == 0.99, \
+            round(stclf.score(X_train, y_train), 2)
+    else:
+        assert round(stclf.score(X_train, y_train), 2) == 0.98, \
+            round(stclf.score(X_train, y_train), 2)
 
 
 def test_decision_function():
@@ -596,7 +604,7 @@ def test_decision_function():
     sclf = StackingCVClassifier(classifiers=[clf1, clf2],
                                 use_probas=True,
                                 meta_classifier=meta)
-    
+
     scores = cross_val_score(sclf,
                              X_breast,
                              y_breast,
@@ -606,5 +614,7 @@ def test_decision_function():
 
     if Version(sklearn_version) < Version("0.21"):
         assert scores_mean == 0.94, scores_mean
-    else:
+    elif Version(sklearn_version) < Version("0.22"):
         assert scores_mean == 0.96, scores_mean
+    else:
+        assert scores_mean == 0.90, scores_mean

mlxtend/evaluate/bias_variance_decomp.py

@@ -86,7 +86,6 @@ def bias_variance_decomp(estimator, X_train, y_train, X_test, y_test,
                                                 arr=all_pred).mean()
 
         avg_bias = np.sum(main_predictions != y_test) / y_test.size
-        avg_var = np.sum(main_predictions != all_pred) / all_pred.size
 
         var = np.zeros(pred.shape)
 

mlxtend/feature_selection/tests/test_exhaustive_feature_selector.py

@@ -7,6 +7,7 @@
 import sys
 import numpy as np
 import pandas as pd
+from distutils.version import LooseVersion as Version
 from numpy.testing import assert_almost_equal
 from mlxtend.feature_selection import ExhaustiveFeatureSelector as EFS
 from sklearn.ensemble import RandomForestClassifier
@@ -17,6 +18,7 @@
 from sklearn.datasets import load_boston
 from mlxtend.utils import assert_raises
 from sklearn.model_selection import GroupKFold
+from sklearn import __version__ as sklearn_version
 
 
 def dict_compare_utility(d1, d2):
@@ -161,27 +163,37 @@ def test_knn_cv3():
     expect = {0: {'avg_score': 0.9391025641025641,
                   'feature_idx': (0, 1, 2),
                   'feature_names': ('0', '1', '2'),
-                  'cv_scores': np.array([0.97435897, 0.94871795,
-                                         0.88888889, 0.94444444])},
-              1: {'avg_score': 0.94017094017094016,
+                  'cv_scores': np.array([0.974, 0.947, 0.892, 0.946])},
+              1: {'avg_score': 0.9400782361308677,
                   'feature_idx': (0, 1, 3),
                   'feature_names': ('0', '1', '3'),
-                  'cv_scores': np.array([0.92307692, 0.94871795,
-                                         0.91666667, 0.97222222])},
+                  'cv_scores': np.array([0.921, 0.947, 0.919, 0.973])},
               2: {'avg_score': 0.95299145299145294,
                   'feature_idx': (0, 2, 3),
                   'feature_names': ('0', '2', '3'),
-                  'cv_scores': np.array([0.97435897, 0.94871795,
-                                         0.91666667, 0.97222222])},
+                  'cv_scores': np.array([0.974, 0.947, 0.919, 0.973])},
               3: {'avg_score': 0.97275641025641035,
                   'feature_idx': (1, 2, 3),
                   'feature_names': ('1', '2', '3'),
-                  'cv_scores': np.array([0.97435897, 1.,
-                                         0.94444444, 0.97222222])}}
+                  'cv_scores': np.array([0.974, 1.   , 0.946, 0.973])}}
+
+    if Version(sklearn_version) < Version("0.22"):
+        expect[0]['cv_scores'] = np.array([0.97435897, 0.94871795,
+                                           0.88888889, 0.94444444])
+        expect[1]['cv_scores'] = np.array([0.92307692, 0.94871795,
+                                           0.91666667, 0.97222222])
+        expect[2]['cv_scores'] = np.array([0.97435897, 0.94871795,
+                                           0.91666667, 0.97222222])
+        expect[3]['cv_scores'] = np.array([0.97435897, 0.94871795,
+                                           0.91666667, 0.97222222])
+        expect[1]['avg_score'] = 0.94017094017094016
+        assert round(efs1.best_score_, 4) == 0.9728
+    else:
+        assert round(efs1.best_score_, 4) == 0.9732
+
     dict_compare_utility(d1=expect, d2=efs1.subsets_)
     assert efs1.best_idx_ == (1, 2, 3)
     assert efs1.best_feature_names_ == ('1', '2', '3')
-    assert round(efs1.best_score_, 4) == 0.9728
 
 
 def test_knn_cv3_groups():
@@ -198,7 +210,7 @@ def test_knn_cv3_groups():
     np.random.seed(1630672634)
     groups = np.random.randint(0, 6, size=len(y))
     efs1 = efs1.fit(X, y, groups=groups)
-    # print(efs1.subsets_)
+
     expect = {0: {'cv_scores': np.array([0.97916667, 0.93877551, 0.9245283]),
                   'feature_idx': (0, 1, 2),
                   'avg_score': 0.9474901595858469,
@@ -233,27 +245,40 @@ def test_fit_params():
     efs1 = efs1.fit(X, y, sample_weight=sample_weight)
     expect = {0: {'feature_idx': (0, 1, 2),
                   'feature_names': ('0', '1', '2'),
-                  'cv_scores': np.array([0.94871795, 0.92307692,
-                                         0.91666667, 0.97222222]),
-                  'avg_score': 0.9401709401709402},
+                  'cv_scores': np.array([0.947, 0.868, 0.919, 0.973]),
+                  'avg_score': 0.9269203413940257},
               1: {'feature_idx': (0, 1, 3),
                   'feature_names': ('0', '1', '3'),
-                  'cv_scores': np.array([0.92307692, 0.92307692,
-                                         0.88888889, 1.]),
+                  'cv_scores': np.array([0.921, 0.921, 0.892, 1.]),
                   'avg_score': 0.9337606837606838},
               2: {'feature_idx': (0, 2, 3),
                   'feature_names': ('0', '2', '3'),
-                  'cv_scores': np.array([0.97435897, 0.94871795,
-                                         0.94444444, 0.97222222]),
-                  'avg_score': 0.9599358974358974},
+                  'cv_scores': np.array([0.974, 0.947, 0.919, 0.973]),
+                  'avg_score': 0.9532361308677098},
               3: {'feature_idx': (1, 2, 3),
                   'feature_names': ('1', '2', '3'),
-                  'cv_scores': np.array([0.97435897, 0.94871795,
-                                         0.91666667, 1.]),
-                  'avg_score': 0.9599358974358974}}
+                  'cv_scores': np.array([0.974, 0.947, 0.892, 1.]),
+                  'avg_score': 0.9532361308677098}}
+
+    if Version(sklearn_version) < Version("0.22"):
+        expect[0]['avg_score'] = 0.9401709401709402
+        expect[0]['cv_scores'] = np.array([0.94871795, 0.92307692,
+                                           0.91666667, 0.97222222])
+        expect[1]['cv_scores'] = np.array([0.94871795, 0.92307692,
+                                           0.91666667, 0.97222222])
+        expect[2]['cv_scores'] = np.array([0.94871795, 0.92307692,
+                                           0.91666667, 0.97222222])
+        expect[2]['avg_score'] = 0.9599358974358974
+        expect[3]['avg_score'] = 0.9599358974358974
+        expect[3]['cv_scores'] = np.array([0.97435897, 0.94871795,
+                                           0.91666667, 1.])
+        assert round(efs1.best_score_, 4) == 0.9599
+
+    else:
+        assert round(efs1.best_score_, 4) == 0.9532
+
     dict_compare_utility(d1=expect, d2=efs1.subsets_)
     assert efs1.best_idx_ == (0, 2, 3)
-    assert round(efs1.best_score_, 4) == 0.9599
 
 
 def test_regression():

mlxtend/feature_selection/tests/test_sequential_feature_selector.py

@@ -203,23 +203,30 @@ def test_knn_cv3():
     sfs1 = sfs1.fit(X, y)
     sfs1.subsets_
     expect = {1: {'avg_score': 0.95299145299145294,
-                  'cv_scores': np.array([0.97435897,
-                                         0.94871795,
-                                         0.88888889,
-                                         1.0]),
+                  'cv_scores': np.array([0.974, 0.947, 0.892, 1.]),
                   'feature_idx': (3,)},
               2: {'avg_score': 0.95993589743589736,
-                  'cv_scores': np.array([0.97435897,
-                                         0.94871795,
-                                         0.91666667,
-                                         1.0]),
+                  'cv_scores': np.array([0.974, 0.947, 0.919, 1.]),
                   'feature_idx': (2, 3)},
-              3: {'avg_score': 0.97275641025641035,
-                  'cv_scores': np.array([0.97435897,
-                                         1.0,
-                                         0.94444444,
-                                         0.97222222]),
+              3: {'avg_score': 0.9732,
+                  'cv_scores': np.array([0.974, 1., 0.946, 0.973]),
                   'feature_idx': (1, 2, 3)}}
+
+    if Version(sklearn_version) < Version("0.22"):
+        expect[1]['cv_scores'] = np.array([0.97435897,
+                                           0.94871795,
+                                           0.88888889,
+                                           1.0])
+        expect[2]['cv_scores'] = np.array([0.97435897,
+                                           0.94871795,
+                                           0.91666667,
+                                           1.0])
+        expect[2]['avg_score'] = 0.97275641025641035
+        expect[3]['cv_scores'] = np.array([0.97435897,
+                                           1.0,
+                                           0.94444444,
+                                           0.97222222])
+
     dict_compare_utility(d_actual=sfs1.subsets_, d_desired=expect)
 
 
@@ -454,7 +461,11 @@ def test_knn_scoring_metric():
                cv=4,
                verbose=0)
     sfs5 = sfs5.fit(X, y)
-    assert round(sfs5.k_score_, 4) == 0.9728
+
+    if Version(sklearn_version) < '0.22':
+        assert round(sfs5.k_score_, 4) == 0.9728
+    else:
+        assert round(sfs5.k_score_, 4) == 0.9732
 
     sfs6 = SFS(knn,
                k_features=3,
@@ -463,7 +474,10 @@ def test_knn_scoring_metric():
                cv=4,
                verbose=0)
     sfs6 = sfs6.fit(X, y)
-    assert round(sfs6.k_score_, 4) == 0.9728
+    if Version(sklearn_version) < '0.22':
+        assert round(sfs5.k_score_, 4) == 0.9728
+    else:
+        assert round(sfs5.k_score_, 4) == 0.9732
 
     sfs7 = SFS(knn,
                k_features=3,
@@ -472,7 +486,10 @@ def test_knn_scoring_metric():
                scoring='f1_macro',
                cv=4)
     sfs7 = sfs7.fit(X, y)
-    assert round(sfs7.k_score_, 4) == 0.9727, sfs7.k_score_
+    if Version(sklearn_version) < '0.22':
+        assert round(sfs5.k_score_, 4) == 0.9727
+    else:
+        assert round(sfs5.k_score_, 4) == 0.9732
 
 
 def test_regression():
@@ -588,9 +605,9 @@ def _fit(self, X, y, init_params=True):
                     errors += int(update != 0.0)
 
                 if self.print_progress:
-                    self._print_progress(iteration=i + 1,
-                                         n_iter=self.epochs,
-                                         cost=errors)
+                    self.print_progress(iteration=i + 1,
+                                        n_iter=self.epochs,
+                                        cost=errors)
                 self.cost_.append(errors)
             return self