Merge branch 'master' into gaugup/SerializeDeserializeExplainers

Author: gaugup

dice_ml/explainer_interfaces/dice_tensorflow1.py

@@ -225,7 +225,7 @@ def do_cf_initializations(self, total_CFs, algorithm, features_to_vary):
     def predict_fn(self, input_instance):
         """prediction function"""
         temp_preds = self.dice_sess.run(self.output_tensor, feed_dict={self.input_tensor: input_instance})
-        return np.array([preds[(self.num_ouput_nodes-1):] for preds in temp_preds])
+        return np.array([preds[(self.num_output_nodes-1):] for preds in temp_preds])
 
     def predict_fn_for_sparsity(self, input_instance):
         """prediction function for sparsity correction"""
@@ -239,22 +239,22 @@ def compute_yloss(self, method):
             if method == "l2_loss":
                 temp_loss = tf.square(tf.subtract(
                     self.model.get_output(self.cfs_frozen[i]), self.target_cf))
-                temp_loss = temp_loss[:, (self.num_ouput_nodes-1):][0][0]
+                temp_loss = temp_loss[:, (self.num_output_nodes-1):][0][0]
             elif method == "log_loss":
                 temp_logits = tf.log(
                     tf.divide(
                         tf.abs(tf.subtract(self.model.get_output(self.cfs_frozen[i]), 0.000001)),
                         tf.subtract(1.0, tf.abs(tf.subtract(self.model.get_output(
                             self.cfs_frozen[i]), 0.000001)))))
-                temp_logits = temp_logits[:, (self.num_ouput_nodes-1):]
+                temp_logits = temp_logits[:, (self.num_output_nodes-1):]
                 temp_loss = tf.nn.sigmoid_cross_entropy_with_logits(
                     logits=temp_logits, labels=self.target_cf)[0][0]
             elif method == "hinge_loss":
                 temp_logits = tf.log(
                     tf.divide(
                         tf.abs(tf.subtract(self.model.get_output(self.cfs_frozen[i]), 0.000001)),
                         tf.subtract(1.0, tf.abs(tf.subtract(self.model.get_output(self.cfs_frozen[i]), 0.000001)))))
-                temp_logits = temp_logits[:, (self.num_ouput_nodes-1):]
+                temp_logits = temp_logits[:, (self.num_output_nodes-1):]
                 temp_loss = tf.losses.hinge_loss(
                     logits=temp_logits, labels=self.target_cf)
 

requirements-linting.txt

@@ -5,5 +5,6 @@ flake8-breakpoint
 flake8-builtins==1.5.3
 flake8-logging-format==0.6.0
 flake8-nb==0.4.0
+flake8-pytest-style
 isort
 packaging

tests/conftest.py

@@ -9,7 +9,7 @@
 from dice_ml.utils import helpers
 
 
-@pytest.fixture
+@pytest.fixture()
 def binary_classification_exp_object(method="random"):
     backend = 'sklearn'
     dataset = helpers.load_custom_testing_dataset_binary()
@@ -20,7 +20,7 @@ def binary_classification_exp_object(method="random"):
     return exp
 
 
-@pytest.fixture
+@pytest.fixture()
 def binary_classification_exp_object_out_of_order(method="random"):
     backend = 'sklearn'
     dataset = helpers.load_outcome_not_last_column_dataset()
@@ -31,7 +31,7 @@ def binary_classification_exp_object_out_of_order(method="random"):
     return exp
 
 
-@pytest.fixture
+@pytest.fixture()
 def multi_classification_exp_object(method="random"):
     backend = 'sklearn'
     dataset = helpers.load_custom_testing_dataset_multiclass()
@@ -42,7 +42,7 @@ def multi_classification_exp_object(method="random"):
     return exp
 
 
-@pytest.fixture
+@pytest.fixture()
 def regression_exp_object(method="random"):
     backend = 'sklearn'
     dataset = helpers.load_custom_testing_dataset_regression()
@@ -81,7 +81,7 @@ def sklearn_regression_model_interface():
     return m
 
 
-@pytest.fixture
+@pytest.fixture()
 def public_data_object():
     """
     Returns a public data object for the adult income dataset
@@ -90,7 +90,7 @@ def public_data_object():
     return dice_ml.Data(dataframe=dataset, continuous_features=['age', 'hours_per_week'], outcome_name='income')
 
 
-@pytest.fixture
+@pytest.fixture()
 def private_data_object():
     """
     Returns a private data object containing meta information about the adult income dataset.
@@ -110,7 +110,7 @@ def private_data_object():
     return dice_ml.Data(features=features_dict, outcome_name='income')
 
 
-@pytest.fixture
+@pytest.fixture()
 def sample_adultincome_query():
     """
     Returns a sample query instance for adult income dataset
@@ -119,63 +119,63 @@ def sample_adultincome_query():
             'race': 'White', 'gender': 'Female', 'hours_per_week': 45}
 
 
-@pytest.fixture
+@pytest.fixture()
 def sample_custom_query_1():
     """
     Returns a sample query instance for the custom dataset
     """
     return pd.DataFrame({'Categorical': ['a'], 'Numerical': [25]})
 
 
-@pytest.fixture
+@pytest.fixture()
 def sample_custom_query_2():
     """
     Returns a sample query instance for the custom dataset
     """
     return pd.DataFrame({'Categorical': ['b'], 'Numerical': [25]})
 
 
-@pytest.fixture
+@pytest.fixture()
 def sample_custom_query_3():
     """
     Returns a sample query instance for the custom dataset
     """
     return pd.DataFrame({'Categorical': ['d'], 'Numerical': [1000000]})
 
 
-@pytest.fixture
+@pytest.fixture()
 def sample_custom_query_4():
     """
     Returns a sample query instance for the custom dataset
     """
     return pd.DataFrame({'Categorical': ['c'], 'Numerical': [13]})
 
 
-@pytest.fixture
+@pytest.fixture()
 def sample_custom_query_5():
     """
     Returns a sample query instance for the custom dataset
     """
     return pd.DataFrame({'X': ['d'], 'Numerical': [25]})
 
 
-@pytest.fixture
+@pytest.fixture()
 def sample_custom_query_6():
     """
     Returns a sample query instance for the custom dataset including Outcome
     """
     return pd.DataFrame({'Categorical': ['c'], 'Numerical': [13], 'Outcome': 0})
 
 
-@pytest.fixture
+@pytest.fixture()
 def sample_custom_query_index():
     """
     Returns a sample query instance for the custom dataset
     """
     return pd.DataFrame({'Categorical': ['a'], 'Numerical': [88]})
 
 
-@pytest.fixture
+@pytest.fixture()
 def sample_custom_query_10():
     """
     Returns a sample query instance for the custom dataset
@@ -188,7 +188,7 @@ def sample_custom_query_10():
     )
 
 
-@pytest.fixture
+@pytest.fixture()
 def sample_counterfactual_example_dummy():
     """
     Returns a sample counterfactual example
@@ -208,7 +208,7 @@ def sample_counterfactual_example_dummy():
     )
 
 
-@pytest.fixture
+@pytest.fixture()
 def create_iris_data():
     iris = load_iris()
     x_train, x_test, y_train, y_test = train_test_split(
@@ -218,7 +218,7 @@ def create_iris_data():
     return x_train, x_test, y_train, y_test, feature_names, classes
 
 
-@pytest.fixture
+@pytest.fixture()
 def create_housing_data():
     housing = fetch_california_housing()
     x_train, x_test, y_train, y_test = train_test_split(

tests/test_counterfactual_explanations.py

@@ -109,7 +109,7 @@ def test_sorted_local_importance_counterfactual_explanations(self):
                 list(counterfactual_explanations.local_importance[index].keys())
 
 
-@pytest.fixture
+@pytest.fixture()
 def random_binary_classification_exp_object():
     backend = 'sklearn'
     dataset = helpers.load_custom_testing_dataset()
@@ -150,7 +150,7 @@ def verify_counterfactual_explanations(self, counterfactual_explanations,
             assert counterfactual_explanations.summary_importance is None
 
     @pytest.mark.parametrize("version", ['1.0', '2.0'])
-    @pytest.mark.parametrize("desired_class, total_CFs", [(0, 2)])
+    @pytest.mark.parametrize(("desired_class", "total_CFs"), [(0, 2)])
     def test_counterfactual_explanations_output(self, desired_class,
                                                 sample_custom_query_1, total_CFs,
                                                 version):
@@ -172,7 +172,7 @@ def test_counterfactual_explanations_output(self, desired_class,
         assert recovered_counterfactual_explanations == counterfactual_explanations
 
     @pytest.mark.parametrize("version", ['1.0', '2.0'])
-    @pytest.mark.parametrize("desired_class, total_CFs", [(0, 10)])
+    @pytest.mark.parametrize(("desired_class", "total_CFs"), [(0, 10)])
     def test_local_importance_output(self, desired_class, sample_custom_query_1,
                                      total_CFs, version):
         counterfactual_explanations = self.exp.local_feature_importance(
@@ -195,7 +195,7 @@ def test_local_importance_output(self, desired_class, sample_custom_query_1,
         assert recovered_counterfactual_explanations == counterfactual_explanations
 
     @pytest.mark.parametrize("version", ['1.0', '2.0'])
-    @pytest.mark.parametrize("desired_class, total_CFs", [(0, 10)])
+    @pytest.mark.parametrize(("desired_class", "total_CFs"), [(0, 10)])
     def test_summary_importance_output(self, desired_class, sample_custom_query_10,
                                        total_CFs, version):
         counterfactual_explanations = self.exp.global_feature_importance(
@@ -242,7 +242,7 @@ def test_empty_counterfactual_explanations_object(self, version):
         assert counterfactual_explanations == recovered_counterfactual_explanations
 
     @pytest.mark.parametrize("version", ['1.0', '2.0'])
-    @pytest.mark.parametrize("desired_class, total_CFs", [(0, 2)])
+    @pytest.mark.parametrize(("desired_class", "total_CFs"), [(0, 2)])
     def test_no_counterfactuals_found(self, desired_class,
                                       sample_custom_query_1, total_CFs,
                                       version):
@@ -261,7 +261,7 @@ def test_no_counterfactuals_found(self, desired_class,
         assert counterfactual_explanations == recovered_counterfactual_explanations
 
     @pytest.mark.parametrize("version", ['1.0', '2.0'])
-    @pytest.mark.parametrize("desired_class, total_CFs", [(0, 10)])
+    @pytest.mark.parametrize(("desired_class", "total_CFs"), [(0, 10)])
     def test_no_counterfactuals_found_local_importance(self, desired_class,
                                                        sample_custom_query_1, total_CFs,
                                                        version):
@@ -282,7 +282,7 @@ def test_no_counterfactuals_found_local_importance(self, desired_class,
         assert counterfactual_explanations == recovered_counterfactual_explanations
 
     @pytest.mark.parametrize("version", ['1.0', '2.0'])
-    @pytest.mark.parametrize("desired_class, total_CFs", [(0, 10)])
+    @pytest.mark.parametrize(("desired_class", "total_CFs"), [(0, 10)])
     def test_no_counterfactuals_found_summary_importance(self, desired_class,
                                                          sample_custom_query_10, total_CFs,
                                                          version):

tests/test_data_interface/test_private_data_interface.py

@@ -5,7 +5,7 @@
 import dice_ml
 
 
-@pytest.fixture
+@pytest.fixture()
 def data_object():
     features_dict = OrderedDict(
         [('age', [17, 90]),

tests/test_data_interface/test_public_data_interface.py

@@ -10,7 +10,7 @@
 from dice_ml.utils.exception import UserConfigValidationException
 
 
-@pytest.fixture
+@pytest.fixture()
 def data_object():
     dataset = helpers.load_adult_income_dataset()
     return dice_ml.Data(dataframe=dataset, continuous_features=['age', 'hours_per_week'],
@@ -67,16 +67,19 @@ def test_invalid_continuous_features(self, data_type):
         iris = load_iris(as_frame=True)
         dataset = iris.frame
 
+        import re
         if data_type == DataTypeCombinations.Incorrect:
-            with pytest.raises(ValueError) as ve:
+            with pytest.raises(
+                    ValueError,
+                    match=re.escape("should provide the name(s) of continuous features in the data as a list")):
                 dice_ml.Data(dataframe=dataset, continuous_features=np.array(iris.feature_names),
                              outcome_name='target')
-            assert "should provide the name(s) of continuous features in the data as a list" in str(ve)
         elif data_type == DataTypeCombinations.AsNone:
-            with pytest.raises(ValueError) as ve:
+            with pytest.raises(
+                    ValueError,
+                    match=re.escape("should provide the name(s) of continuous features in the data as a list")):
                 dice_ml.Data(dataframe=dataset, continuous_features=None,
                              outcome_name='target')
-            assert "should provide the name(s) of continuous features in the data as a list" in str(ve)
         else:
             with pytest.raises(
                     ValueError,

tests/test_dice.py

@@ -54,8 +54,8 @@ def test_minimum_query_instances(self):
         pytest.importorskip('sklearn')
         backend = 'sklearn'
         exp = self._get_exp(backend)
+        query_instances = helpers.load_adult_income_dataset().drop("income", axis=1)[0:1]
         with pytest.raises(UserConfigValidationException):
-            query_instances = helpers.load_adult_income_dataset().drop("income", axis=1)[0:1]
             exp.global_feature_importance(query_instances)
 
     def test_unsupported_sampling_strategy(self):