Merge branch 'master' into gaugup/ContinueUnifyTest

.github/workflows/python-package-conda.yml

@@ -0,0 +1,31 @@
+name: Python Package using Conda
+
+on: [push]
+
+jobs:
+  build-linux:
+    runs-on: ubuntu-latest
+    strategy:
+      max-parallel: 5
+
+    steps:
+    - uses: actions/checkout@v2
+    - name: Set up Python 3.8
+      uses: actions/setup-python@v2
+      with:
+        python-version: 3.8
+    - name: Add conda to system path
+      run: |
+        # $CONDA is an environment variable pointing to the root of the miniconda directory
+        echo $CONDA/bin >> $GITHUB_PATH
+    - name: Install core dependencies
+      run: |
+        conda env update --file environment.yml --name base
+    - name: Install deep learning dependencies
+      run: |
+        conda env update --file environment-deeplearning.yml --name base
+    - name: Test with pytest
+      run: |
+        conda install pytest ipython jupyter nbformat pytest-mock
+
+        pytest

dice_ml/data_interfaces/public_data_interface.py

@@ -260,7 +260,7 @@ def get_valid_feature_range(self, feature_range_input, normalized=True):
         """
         feature_range = {}
 
-        for idx, feature_name in enumerate(self.feature_names):
+        for _, feature_name in enumerate(self.feature_names):
             feature_range[feature_name] = []
             if feature_name in self.continuous_feature_names:
                 max_value = self.data_df[feature_name].max()

dice_ml/explainer_interfaces/dice_tensorflow2.py

@@ -177,7 +177,7 @@ def do_cf_initializations(self, total_CFs, algorithm, features_to_vary):
         # CF initialization
         if len(self.cfs) != self.total_CFs:
             self.cfs = []
-            for ix in range(self.total_CFs):
+            for _ in range(self.total_CFs):
                 one_init = [[]]
                 for jx in range(self.minx.shape[1]):
                     one_init[0].append(np.random.uniform(self.minx[0][jx], self.maxx[0][jx]))

dice_ml/explainer_interfaces/explainer_base.py

@@ -354,7 +354,7 @@ def feature_importance(self, query_instances, cf_examples_list=None,
                 continue
 
             per_query_point_cfs = 0
-            for index, row in df.iterrows():
+            for _, row in df.iterrows():
                 per_query_point_cfs += 1
                 for col in self.data_interface.continuous_feature_names:
                     if not np.isclose(org_instance[col].iat[0], row[col]):

dice_ml/explainer_interfaces/feasible_base_vae.py

@@ -186,7 +186,7 @@ def generate_counterfactuals(self, query_instance, total_CFs, desired_class="opp
             curr_cf_pred = []
             curr_test_pred = train_y.numpy()
 
-            for cf_count in range(total_CFs):
+            for _ in range(total_CFs):
                 recon_err, kl_err, x_true, x_pred, cf_label = \
                     self.cf_vae.compute_elbo(train_x, 1.0-train_y, self.pred_model)
                 while(cf_label == train_y):

dice_ml/model_interfaces/base_model.py

@@ -64,7 +64,7 @@ def get_num_output_nodes(self, inp_size):
         temp_input = np.transpose(np.array([np.random.uniform(0, 1) for i in range(inp_size)]).reshape(-1, 1))
         return self.get_output(temp_input).shape[1]
 
-    def get_num_output_nodes2(self, input):
+    def get_num_output_nodes2(self, input_instance):
         if self.model_type == ModelTypes.Regressor:
             raise SystemException('Number of output nodes not supported for regression')
-        return self.get_output(input).shape[1]
+        return self.get_output(input_instance).shape[1]

dice_ml/model_interfaces/keras_tensorflow_model.py

@@ -40,7 +40,7 @@ def get_output(self, input_tensor, training=False, transform_data=False):
         else:
             return self.model(input_tensor)
 
-    def get_gradient(self, input):
+    def get_gradient(self, input_instance):
         # Future Support
         raise NotImplementedError("Future Support")
 

dice_ml/model_interfaces/pytorch_model.py

@@ -38,7 +38,7 @@ def get_output(self, input_tensor, transform_data=False):
     def set_eval_mode(self):
         self.model.eval()
 
-    def get_gradient(self, input):
+    def get_gradient(self, input_instance):
         # Future Support
         raise NotImplementedError("Future Support")
 

dice_ml/utils/sample_architecture/vae_model.py

@@ -109,7 +109,7 @@ def forward(self, x, c):
         res['z'] = []
         res['x_pred'] = []
         res['mc_samples'] = mc_samples
-        for i in range(mc_samples):
+        for _ in range(mc_samples):
             z = self.sample_latent_code(em, ev)
             x_pred = self.decoder(torch.cat((z, c), 1))
             res['z'].append(z)
@@ -239,7 +239,7 @@ def forward(self, x):
         res['z'] = []
         res['x_pred'] = []
         res['mc_samples'] = mc_samples
-        for i in range(mc_samples):
+        for _ in range(mc_samples):
             z = self.sample_latent_code(em, ev)
             x_pred = self.decoder(z)
             res['z'].append(z)

docs/source/conf.py

@@ -19,7 +19,7 @@
 # -- Project information -----------------------------------------------------
 
 project = 'DiCE'
-copyright = '2020, Ramaravind, Amit, Chenhao'
+copyright = '2020, Ramaravind, Amit, Chenhao'  # noqa: A001
 author = 'Ramaravind, Amit, Chenhao'
 
 # The full version, including alpha/beta/rc tags

docs/source/notebooks/Benchmarking_different_CF_explanation_methods.ipynb

@@ -270,7 +270,7 @@
     "elapsed_kd = 0\n",
     "elapsed_genetic = 0\n",
     "\n",
-    "for i in range(num_loops):\n",
+    "for _ in range(num_loops):\n",
     "    for q in query_instances:\n",
     "        if q in d.categorical_feature_names:\n",
     "            query_instances.loc[:, q] = \\\n",

environment-deeplearning.yml

@@ -0,0 +1,5 @@
+name: example-environment
+
+dependencies:
+  - tensorflow>=1.13.0-rc1
+  - pytorch

environment.yml

@@ -0,0 +1,9 @@
+name: example-environment
+
+dependencies:
+  - numpy
+  - scikit-learn
+  - pandas
+  - h5py
+  - jsonschema
+  - tqdm

requirements-deeplearning.txt

@@ -1,3 +1,2 @@
 tensorflow>=1.13.0-rc1 
 torch
-torchvision

requirements-linting.txt

@@ -1,3 +1,7 @@
 flake8==3.9.2
+flake8-bugbear==21.11.29
+flake8-blind-except==0.1.1
+flake8-builtins==1.5.3
+flake8-logging-format==0.6.0
 flake8-nb==0.3.0
 isort

tests/test_model_interface/test_keras_tensorflow_model.py

@@ -74,4 +74,6 @@ def test_model_output(self, sample_adultincome_query, public_data_object, predic
             predictval = self.sess.run(output_instance)[0][0]
         else:
             predictval = output_instance.numpy()[0][0]
-        pytest.approx(predictval, abs=1e-3) == prediction
+        assert predictval is not None
+        # TODO: The assert below fails.
+        # assert pytest.approx(predictval, abs=1e-3) == prediction

tests/test_model_interface/test_pytorch_model.py

@@ -57,4 +57,6 @@ def test_model_output(self, sample_adultincome_query, public_data_object, predic
         self.m.transformer.initialize_transform_func()
         output_instance = self.m.get_output(sample_adultincome_query, transform_data=True)
         predictval = output_instance.detach().numpy()[0][0]
-        pytest.approx(predictval, abs=1e-3) == prediction
+        assert predictval is not None
+        # TODO: The assert below fails.
+        # assert pytest.approx(predictval, abs=1e-3) == prediction

tests/test_notebooks.py

@@ -41,7 +41,7 @@ def _check_notebook_cell_outputs(filepath):
     for cell in nb.cells:
         if "outputs" in cell:
             if len(cell['outputs']) > 0:
-                assert False, "Output cell found in notebook. Please clean your notebook"
+                raise AssertionError("Output cell found in notebook. Please clean your notebook")
 
 
 def _notebook_run(filepath):